EDITORIALS

G ITAL DERMATOL VENEREOL 2005;140:301

Guidelines in dermoscopy
A. W. KOPF

D r. Chimenti and his colleagues are to be congrat-

ulated in the recommendation to develop guide-
lines for this important addition to the clinical diagnosis
Department of Dermatology
New York University School of Medicine
New York, NY, USA
of pigmented lesions of the skin with emphasis on
malignant melanoma. The authors summarize that der-
moscopy, in the hands of those experienced with the
technique, has proven to enhance the in vivo diagnos- Table I and Table II present succinct and meaning-
tic accuracy of malignant melanoma. Furthermore, it ful definitions for melanoma-specific dermoscopic
has been shown that dermoscopy raises the confidence criteria and the main dermoscopic features of the most
of the physician in differentiating malignant melanoma
difficult pigmented lesions that must be differentiated
from other pigmented lesions of the skin.
from melanoma.
The innovative algorithm for the management of
In toto, this comprehensive guideline and its accom-
pigmented skin lesions (Figure 2) is a useful road map
panying rich bibliography contain the essential aspects
for the application of dermoscopy which allows the
of dermoscopy not only for the beginner but also for
physician to decide whether to biopsy or to follow the
the expert.
lesion. Importantly, using dermoscopy, the clinician can
It is becoming abundantly clear that any physician
decide that biopsy is not needed which leads to cost
who accepts the responsibility for the differential diag-
containment of medical care.
nosis of pigmented lesions of the skin needs to learn
IN THIS ISSUE SEE PAGE 329
dermoscopy. The Guidelines by Chimenti et al. is an
excellent source for the current understanding of this
Address reprint request to: A.W. Kopf, MD, Clinical Professor of Der- most useful diagnostic tool, the primary conclusion
matology, New York University School of Medicine, 350 Fifth Avenue, Sui-
te 7805, New York, NY 10118-0189, USA. of which is whether a lesion should be biopsied or
E-mail akopf@compuserve.com not!

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 301


What’s new in hereditary epidermolysis bullosa
E pidermolysis bullosa (EB) is a group of geneti-
cally determined disorders characterized by var-
ious degrees of cutaneous and mucosal fragility. It is
found worldwide. In the last few decades, much has
been learnt about their genetic basis and clinical fea-
tures. As recently as the early 1980’s, the classifica-
tion of EB was based entirely on clinical features.
Many subtypes were recognized, often identified
with confusing eponyms. In an attempt to better
understand these disorders, investigators have estab-
lished national registries of patients in various coun-
tries around the world. These registries allowed inves-
tigators to identify and study large numbers of patients
with this rare disease. The results have been impres-
sive. Investigators have identified at least 10 mutations
underlying all major forms of EB. They have refined
and simplified the classification of major forms of
EB, and increased our understanding of their clinical
features. Based on our knowledge of the newly iden-
tified mutations, prenatal diagnosis is no longer based
on histological study of fetal skin biopsy specimens,
but has been transformed into a molecular based tech-
nique. Perhaps most excitingly, researchers have
begun to study gene therapy as a potential cure for
these devastating diseases. In this issue of Giornale
Italiano di Dermatologia e Venereologia, Tadini et
IN THIS ISSUE SEE PAGE 358
Address reprint requests to: A. N. Lin, MD, Associate Professor, Divi-
sion of Dermatology and Cutaneous Sciences, University of Alberta,
Edmonton, AB, Canada
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:303-8
A. N. LIN
Division of Dermatology and Cutaneous Sciences
University of Alberta, Edmonton, AB, Canada
al. present the data gathered by the Italian EB Reg-
istry, further enhancing our understanding of these
diseases.
Identification of mutations in EB
EB simplex
Classically, EB has been classified into simplex,
junctional, and dystrophic forms. EB simplex (EBS)
is caused by mutations in the genes encoding for ker-
atin 5 (KRT5) and Keratin 14 (KRT14). In most
patients, it is transmitted as an autosomal dominant
form. In the 2000 revised classificatioin system, based
primarily on data gathered by the American EB Reg-
istry, 4 subtypes are recognized.1 Weber-Cockayne
EBS is generally considerd to be the most common
type of EBS.2 It primarily affects the hands and feet,
making it one of the most readily recognizable types.
Patients may not even realize they have EB until they
begin vigorous physical activity, such as joining the
army. Extracutaneous involvement tends to be mild
or absent. Tadini et al. present the interesting finding
that they found 67 cases of Weber-Cockayne EBS,
slightly fewer than their 75 cases of Koebner EBS. In
303
LIN
Koebner EBS, blistering is generalized with relative
sparing of palms and soles. The most severe form is
EBS-Dowling Meara, with generalized blistering pre-
sent at or shortly after birth, progressive palmoplantar
keratoderma, and mucous membrane involvement.
Blisters often are grouped, resulting in “herpetiform”
arrangement. A distinctive feature on electron
microscopy is clumping of tonofilaments. The mot-
tled pigmentation subtype features hypopigmented
reticulated macules, and corresponds with an increased
number of melanosomes within basal keratinocytes,
dermal macrohages, and Scwann cells.3 It is a very
rare type, and was not observed in the data presented
by Tadini et al. This raises the possibility that there
may be regional variations in its prevalence, although
it may simply reflect the extremely uncommon preva-
lence of this subtype.
Tadini et al. reported 1 case of EBS with muscular
dystrophy, another rare subtype. These patients present
with blisters at birth, followed by late onset muscular
dystrophy, and abnormalities of the nail and teeth,
sometimes associated with laryngeal webs and ure-
thral strictures. It is caused by mutations in the gene
encoding plectin (PLEC1), an intermediate filament
inter-acting protein. Plectin is present not only in the
hemidesmosome, but also in sarcolemma of the mus-
cle, findings that may explain the association with
muscular dystrophy. Because of the close association
of plectin with hemidesmosome, it has been proposed
that EBS with muscular dystrophy should be reclas-
sified into a new type, called hemidesdmosomal type.3
In contrast to most forms of EBS, EBS with muscular
dystrophy is inherited as an autosomal recessive dis-
order. EBS-Ogna is another rare form of EB, previously
classified as EBS, which has now been shown to be due
to a missense mutation in PLEC1,3 and which is now
classified as another form of hemidesmosomal EB.3
The plectin mutation has been demonstrated in the
original kindred from the Norwegian village of Ogna,
and also in a German family with similar phenotype.3
These patients present with hemorrhagic blisters of
the skin, but do not have muscular dystrophy.3
The vast majority of EBS is transmitted as an auto-
somal dominant trait, but autosomal recessive muta-
tions have been identified in KRT14 in 7 families.3
Junctional EB
Junctional EB is transmitted invariably as an auto-
somal recessive trait, and is caused by mutations in
304 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
WHAT’S NEW IN HEREDITARY EPIDERMOLYSIS BULLOSA
gene encoding the subunits of laminin 5 (LAMA3,
LAMB3, and LAMC2). The Herlitz type is one of the
most serious types of EB. Patients present with wide-
spread blistering at birth. A characteristic feature is
exuberant granulation tissue on the face, especially
around the mouth. Enamel hypoplasisa is common,
leading to caries and loss of teeth. Laryngeal involve-
ment is a rare but serious complication, and patients
who develop stridor will require immediate ENT con-
sultation to evaluate the airway, and tracheotomy is
sometimes required.2 Because of the widespread blis-
tering, death in the first few years of life is common,
often due to combination of factors such as sepsis,
severe anemia, and malnutrition. In the non-Herlitz
type, patients exhibit a milder phenotype, but gener-
alized blistering is also present at birth, and enamel
hypoplasia is also common.1
It is interesting to note that Shabbir syndrome, also
known as layrngo-onycho-cutaneous syndrome, is a
recessively inherited disorder that features skin fragili-
ty and exuberant granulation tissue around the eyes
and larynx. Investigators have recently uncovered
mutations in LAMA3,3 raising the possibility that this
disorder may in fact be related to Herlitz-JEB.
Hemidesmosomal EB
Generalized atrophic benign EB (GABEB) is a form
of EB originally classified as JEB. In addition to cuta-
neous fragility, patients have nail dystrophy, scarring
alopecia of the scalp, dental abnormalities, loss of eye-
lashes, and patchy hyperpigmentation of the skin. It is
caused by mutations in the gene encoding the 180kDa
BPAG, a transmembrane hemidesmosomal protein
that is also known as type XVII collagen. Because of
the close association of this protein with the
hemidesmosome, this type of EB has been reclassi-
fied as a new type, the hemidesmosomal type,3 along
with the types formerly known as EBS-muscular dys-
trophy and the Ogna variant of EB, discussed previ-
ously in this review.
EB with pyloric atresia (EB-PA) is another form of
EB that was formerly classified as a form of junctional
EB. It features blistering at birth, in association with
pyloric atresia. Often, there is a history of polyhy-
dramnios during the pregnancy, a clue that the newborn
may have some form of gastric outlet obstruction. If this
is suspected, then an upright abdominal film should be
done. If pyloric atresia is present, the X-ray will show
the single bubble sign, reflecting a large bubble of
Agosto 2005
WHAT’S NEW IN HEREDITARY EPIDERMOLYSIS BULLOSA
swallowed air that is trapped in the stomach. Immediate
surgical intervention is then necessary for survival.
With early surgical correction of the pyloric stenosis,
favourable outcome is possible. EB-PA is caused by
one of the genes encoding the subunit polypeptides
of the alpha6-beta4 integrin (ITGA6 and ITGB4).
Because of the close association of this integrin with
hemidesmosomes, EB-PA has also been reclassified as
a hemidesmosomal type of EB.3
Dystrophic EB
Dystrophic EB (DEB) is caused by mutations in the
gene encoding type VII collagen (COL7A1). Over
100 mutations in this gene have been described in var-
ious forms of dystrophic EB. Type VII collagen is the
principal constituent of anchoring fibrils, which nor-
mally ensure cohesion between the epidermis and der-
mis. As a result, separation occurs below the lamina
densa, and electron microscopy shows absence or
reduced numbers of anchoring fibrils.
DEB can be inherited in both autosomal recessive
and dominant forms. The recessively inherited Hallo-
peau-Siemens form (RDEB-HS) is one of the most dev-
astating types of EB. Patients present with widespread
blistering at birth, and soon develop progressive fusion
of the digits in childhood, resulting in considerable
functional disability. Because the esophagus is lined
with stratified squamous epithelium like the skin, patients
can also develop esophageal blisters that can lead to
scarring and stenosis. Dysphagia is then a major clini-
cal problem, leading to anemia and malnutrition. Some
patients may require esophageal dilation, but this may
damage the already fragile mucosa and may cause per-
foration. Bypassing the stenosed portion of the esoph-
agus with an isoperistaltic segment of the colon has
been performed, but this is a major operation. Favourable
results have been observed with gastrostomy, allowing
introduction of nutrients directly into the stomach.2
Patients with the non-Hallopeau Siemens form of
RDEB (RDEB-nonHS) have a milder phenotype. They
still present with generalized blistering at birth, and
may also develop partial fusion of the digits. An unusu-
al variant is the inverse type, in which blisters devel-
op mainly at inverse sites such as the axilla and groin,
and severe oral and esophageal involvement are com-
mon.2
The dominant type of dystrophic EB is also caused
by mutations in the gene encoding type VII collagen.
Patients develop relatively minor blistering, but nail
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
LIN
dystrophy, scarring, and mucosal involvement can also
occur. Some patients present with distinctive white
papules, called albopapuloid lesions.
Clinical observations
By studying large numbers of EB patients identi-
fied through a national registry, investigators have
made important observations about clinical aspects of
EB. These include extracutaneous involvements that
have remained poorly understood until recently, includ-
ing involvement of the kidney, eyes, and prevention of
skin cancer in the dystrophic types.
Ocular involvement
With its stratified squamous epithelium like the skin,
the cornea is subject to erosions and ulcers in EB. In their
study of 3 280 patients enrolled in the American EB
Registry, Fine et al.4 reported an association between
ocular involvement and disease severity. They found
that at least one episode of corneal erosions and blisters
occurred in 74.1% of patients with RDEB-HS, and in
47.5% of patients with JEB-H. Symblepharons and
ectropions were most often seen in the inverse type of
RDEB and JEB-H. Also, blindness resulting from cumu-
lative corneal scarring was seen in 6.46% of patients
with RDEB-HS. All patients with EB who present with
painful tearing eyes should be assessed by an ophthal-
mologist to prevent cumulative corneal scarring.
Renal disease
Individual case reports have suggested that renal
failure may occur in some patients with EB. The caus-
es of the renal failure have included poststreptococcal
glomerulonephritis (presumably secondary to frequent
cutaneous infection), amyloidosis, and chronic mechan-
ical obstruction. Fine et al.5 analyzed data concern-
ing 3 280 EB patients enrolled with the American EB
Registry, and found 9 patients who were reported to
have died of renal failure. Among these 9 patients,
clinical and laboratory data permitted classification
in 7 patients. Five of these patients had RDEB-HS,
one had RDEB-non HS, and one had JEB-non Herlitz.
They found that the cumulative risk for death from
renal failure among patients with RDEB-HS was
12.3% by age 35 years, and recommended surveil-
lance for early kidney involvement should be part of the
routine evaluation of adults with RDEB and JEB.
305
LIN
Squamous cell carcinoma
Squamous cell carcinoma remains a leading cause
of death in patients with recessive dystrophic EB. They
tend to arise in chronically eroded skin, and may pre-
sent as heaped up granulation tissue, or thick keratot-
ic plaques. In contrast to squamous cell carcinomas
that arise from actinically damaged skin in patients
without EB, these cancers are biologically aggressive,
and metastasize widely. Approximately 85% of all
patients with RDEB-HS will have developed at least
one cutaneous squamous cell carcinoma by age 45
years.6 Furthermore, most patients die of metastatic
squamous cell carcinoma within 5 years of diagnosis
of the first tumor.7 In patients with xeroderma pig-
mentosum and renal transplants, studies have shown
that chronic therapy with isotretinoin may prevent
development of squamous cell carcinoma. In 2004,
Fine et al.7 performed an open study of 20 patients
with RDEB (5 had RDEB-HS, 15 had RDEB-nonHS)
who were aged 15 years or older. Each patient was
given isotretinoin daily, starting at 0.1 mg/kg/day, and
the dose was increased monthly by 0.1 mg/kg/day
until either maintenance dose was achieved (0.5
mg/kg/day), or the patient became intolerant of the
next higher dose. Treatment was continued for 8
months in 19 patients. One patient terminated treatment
after 3 months because of hypertriglyceridemia and
transient abdominal pain, consistent with drug-induced
acute pancreatitis, but all symptoms resolved within 3
days. Other patients experienced side effects that were
mild, including skin dryness and fragility, epistaxis, and
pruritus. Interestingly, over half the patients reported
reduced blister formation while taking low-dosage
isotretinoin, but this effect was lost as each patient
approached the target maintenance dosage of 0.5
mg/kg/day. This finding may reflect in vitro data sug-
gesting that isotretinoin may modulate collagenase
synthesis by RDEB fibroblasts.7 These data suggest that
isotretinoin is tolerated at the dosage studied in RDEB
patients, and set the stage for this agent to be studied
as a possible chemopreventive agent against squa-
mous cell carcinoma in RDEB.
Prenatal diagnosis
Because EB can be such a devastating disease, fam-
ilies with pregnancies at risk often request prenatal
diagnosis. In the past, this depended on examination of
306 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
WHAT’S NEW IN HEREDITARY EPIDERMOLYSIS BULLOSA
fetal skin biopsy with electron microscopy and/or
immunohistochemistry. However, fetal skin biopsy
can only be obtained rather late in the pregnancy, often
after 17 weeks, and is associated with a relatively high
rate of miscarriage. Also, this method requires exam-
ination of the biopsy specimen by an expert in ultra-
structure of fetal skin. With identification of the muta-
tions underlying all major forms of EB, investigators
have made impressive progress towards DNA-based
prenatal diagnosis.
Pfendner et al.8 recently reviewed the experience
of DNA-based prenatal diagnosis performed at the
DeBRA Molecular Diagnostics Laboratory at Jeffer-
son Medical College. Since 1993, investigators at that
center performed 144 DNA-based prenatal diagnoses
in 121 families at risk for RDEB (63 pregnancies),
JEB (69 pregnancies), EB-PA (6 pregnancies), and
EBS (6 pregnancies). In most cases with DEB, and in
all cases with JEB and EBS, the diagnosis was con-
firmed by demonstration of specific mutations. Prenatal
testing was done on DNA isolated from chorionic vil-
li or amniocytes obtained in the first or second trimester.
Their overall accuracy was greater than 98% (2 unex-
plained discordant results out of 144 samples submit-
ted), showing that DNA-based prenatal diagnosis is a
very accurate and reliable method to assess pregnan-
cies at risk.
Gene therapy
One of the most exciting advances in EB research is
the prospect of gene therapy. Research on this has
focused on junctional and recessive dystrophic EB, 2
of the most serious forms.
Junctional EB
Important advances in gene therapy for EB have
described in a recent review.9 In early work concern-
ing junctional EB, investigators have isolated ker-
atinocytes from a patient with severe Herlitz JEB, and
showed a homozygous mutation of the LAMB3 gene.
These cells were unable to synthesize lamin-5, and
hence were not able to assemble hemidesmosomes.
Investigators used a retroviral construct expressing
human beta3 cDNA to successfully transduce the JEB
keratinocytes, which were then able to synthesize and
secrete mature heteterotrimeric lamin-5.
The use of viral vectors, however, is associated with
Agosto 2005
WHAT’S NEW IN HEREDITARY EPIDERMOLYSIS BULLOSA
numerous logistical and biosafety concerns. Investi-
gators have therefore developed non-viral approaches
to JEB gene therapy. Using PhiC31 integrase, inves-
tigators have successfully integrated a laminin-5 beta3
expression plasmid into the genome of primary ker-
atinocytes from four unrelated patients with JEB.9
Recently, investigators have used another nonviral
vector (“Sleeping Beauty” transposable element) to
successfully integrate the LAMB3 cDNA into the
genomes of epidermal holoclones from six unrelated
JEB patients.10 These cells also regenerated human
JEB skin on SCID mice that was normalized at the
level of laminin 5 protein expression, hemidesmo-
some formation, and blistering.
Recessive dystrophic EB
Woodley et al.11 have recently reviewed advances
in gene therapy for recessive dystrophic EB. Ker-
atinocytes from patients with inherited dystrophic
EB cannot make type VII collagen, and show various
abnormalities when compared with normal ker-
atinocytes, For example, they are enlarged, elongat-
ed, and attach poorly to extracellular matrix. However,
when these cells are transduced with a lentiviral vec-
tor encoding human gene for type VII collagen, the
cells begin to permanently synthesize and secrete
type VII collagen, and all the abnormal morpholog-
ic features became normal. They showed normal mor-
phology, attached to extracellular matrix, and migrat-
ed in a normal fashion. Investigators have created
two animal models in which 3 kinds of skin equiva-
lents were transplanted onto mice. The first kind con-
sisted of cells obtained from patients with severe
recessive dystrophic EB, the second consisted of the
same cells that had been “gene corrected” by stable
integration of human type VII collagen and are able
to synthesize and secrete type VII collagen, and the
third consisted of normal human keratinocytes and
fibroblasts. The skin equivalents that were not gene
corrected showed features of RDEB, with fragile epi-
dermal-dermal attachment, and virtually absence of
anchoring fibrils. However, the skin equivalents made
with gene-corrected cells showed presence of anchor-
ing fibrils and type VII collagen at the epidermal der-
mal interface.11
These results suggest that it may be possible to per-
form ex vivo gene therapy, a process in which one
takes skin biopsies from patients with RDEB, and then
stably transfect the cells with the human COL7A1
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
LIN
gene, giving them the ability to synthesize and secrete
type VII collagen.11 One can theoretically expand these
cells into large sheets, and transplant them back onto
the patient with RDEB. However, this procedure would
be technically difficult, and the fragile graft would
easily be lost.
In a recent report, Woodley et al.12 injected recom-
binant human type VII collagen into immunocmpe-
tent SKH mice. The injected protein was incorporat-
ed onto the basement membrane zone and remained
stable for at least 6 weeks. Sera from 10 mice were
evaluated for antibodies to type VII collagen, and these
antibodies were found in 6 of the 10 mice. However,
none of the mice lost weight or showed any untoward
effects. Also, the antibodies did not prevent further
incorporation of human type VII collagen in the base-
ment membrane zone when later injected into new
areas of the mouse’s skin.
In the same report, Woodley et al.12 studied RDEB
skin tissues regenerated on immunodeficient mice.
These tissues retained the RDEB phenotype, with
histologic evidence of dermal-epidermal separation
and absence of human type VII collagen staining.
However, intradermal injection of recombinant type
VII collagen into the RDEB skin corrected the blis-
tering and restored type VII collagen expression at
the basement membrane zone. These studies yielded
important insights to the prospect of protein-based
gene therapy.
References
1. Fine J-D, Eady RAJ, Bauer EA, Briggaman RA, Bruckner-Tuder-
man L, Christiano A et al. Revised classification system for inherit-
ed epidermolysis bullosa: Report of the second International Con-
census Meeting on diagnosis and classification of epidermolysis bul-
losa. J Am Acad Dermatol 2000;42:1051-66.
2. Lin AN, Carter DM. Epidermolysis Bullosa: Basic and Clinical
Aspects. New York:Springer-Verlag;1993.
3. Uitto J, Richard G. Progress in epidermolysis bullosa: genetic clas-
sification and clinical implications. Am J Med Genet C Semin Med
Genet 2004;131C:61-74.
4. Fine J-D, Johnson LB, Weiner M, Stein A, Cash S, Deleoz J et al.
Eye involvement in inherited epidermolysis bullosa: experience of
the National Epidermolysis Bullosa Registry. Am J Ophthalmol
2004;138:254-62.
5. Fine J-D, Johnson LB, Weiner M, Stein A, Cash S, Deleoz J et al.
Inherited epidermolysis bullosa and the risk of death from renal dis-
ease: experience of the National Epidermolysis Bullosa Registry. Am
J Kidney Dis 2004;44:651-60.
6. Fine J-D, Bauer EA, McGuire J, Moshell A Epidermolysis Bullosa:
clinical epidemiologic, and laboratory advances and the findings of the
National Epidermolysis Bullosa Registry. Baltimore: Johns Hopkins
University Press;1999.
307
LIN
7. Fine J-D, Johnson LB, Weiner M, Stein A, Suchindran C. Chemo-
prevention of ssquamous cell carcinoma in recessive dystrophic epi-
dermolysis bullosa: results of a phase 1 trial of systemic isotretinoin.
J Am Acad Dermatol 2004;50:563-71.
8. Pfendner EG, Nakano A, Pulkinen L, Christiano A, Uitto J. Prenatal
diagnosis for epidermolysis bullosa: a study of 144 consecutive preg-
nancies at risk. Prenat Diagn 2003;23:447-56.
9. Bauer JW, Laimer M. Gene therapy of epidermolysis bullosa. Expert
Opin Bio Ther 2004;4:1435-43.
308 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
WHAT’S NEW IN HEREDITARY EPIDERMOLYSIS BULLOSA
10. Ortiz-Urda S, Lin Q, Yant SR, Keene D, Kay MA, Khavari PA. Sdus-
tainable correction of junctional epidermolysis bullosa via transposon-
mediated nonviral gene transfer. Gene Ther 2003;10:1099-104.
11. Woodley DT, Chen M. Epidermolysis bullosa: then and now. J Am
Acad Dermatol 2004;51:S55-7.
12. Woodley DT, Keene DR, Atha T, Huang Y, Lipman K, Li W et al.
Injection of recombinant human type VII collagen restores collagen
function in dystrophic epidermolysis bullosa. Nat Med 2004;10:
693-5.
Agosto 2005

Inherited Epidermolysis Bullosa Registries.
Whither Hence?
I n the current issue of the Journal Tadini et al. reports
on the Italian experience with a nationally based
registry of epidermolysis bullosa (EB) patients. This
is an important contribution to our literature, since it
provides an opportunity to compare data generated in
Europe with those produced since 1986 within the
United States by the National EB Registry.1 These
registries differ not only in their geographic origin but
also in the source of data collection, the number of
enrollees, and the duration of follow-up. The Italian
registry is a hospital-based one, involving hospital-
based participating dermatologists throughout Italy.
Data generated in this manner are similar to those
obtained via a hospital-based survey performed in
Japan many years ago.2 The American Registry, based
on data collection by investigators at four regionally
distinct medical schools, has instead used rigorous
epidemiological case finding techniques, in conjunc-
tion with the recruitment advantage of a free, central-
ized, diagnostic laboratory, referrals from outside
physicians, and self-referrals from other patients and
their affected family members, to identify and recruit
its study subjects.3 Given these differences in study
design and methodology, one might expect underre-
porting of milder cases in the Italian registry, since
IN THIS ISSUE SEE PAGE 358
Address reprint reqeusts to: Dr. J. D. Fine, Divisions of Dermatology and
Pediatrics, Vanderbilt University School of Medicine, c/o VMG Derma-
tology, 1900 Patterson Street, Suite 100 Nashville, TN 37203, USA.
E-mail: jo-david.fine@vanderbilt.edu
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:309-13
J. D. FINE 1, 2
1Divisionsof Dermatology and Pediatrics
Vanderbilt University School of Medicine
Nashville, TN, USA
2National Epidermolysis Bullosa Registry
Nashville, TN, USA
few mild cases of EB would likely present for evalu-
ation within a hospital setting. Despite such differ-
ences, however, the overall prevalence (10.1 cases per
million in Italy vs 8.2 in the U.S.) and incidence rates
(20.1 cases per one million live births in Italy vs 19.6
in the U.S.) for inherited EB are nearly identical, sup-
porting the validity of the data already published by the
American EB registry,2 as well as validating the effec-
tiveness of the ongoing registry in Italy in its case find-
ing efforts. Based on the close consistency of these
demographic parameters within both otherwise dis-
tinct study populations, it seems appropriate to sum-
marize the experience of the American EB Registry and
then to suggest how the Italian registry can further
add to our overall understanding of this disease.
What have we already learned from the American
EB registry, based on nearly 19 years of collection of
cross-sectional data on nearly 3 300 patients and lon-
gitudinal data on approximately 425 subjects randomly
selected from the projects’ overall study population?
First, we now have accurate estimates of the preva-
lence and incidence of each of the major EB types and
subtypes, as well as overall rates for EB as a whole.2
309
FINE
Given the rarity of some minor EB subtypes even with-
in a population as large as the United States, howev-
er, it must be stressed that these latter calculated rates,
similar to those reported in this issue of the Journal by
Tadini et al., can be used only as rough estimates,
since statistically these numbers are not as firm as
those derived from the more common EB subtypes.
Second, our data have documented that most of the
cutaneous features in EB, most notably scarring, mil-
ia, and nail dystrophy, can occur not only in dystrophic
EB but also in junctional and simplex patients. For
example, approximately 30%, 11%, and 33% of all
EB simplex patients enrolled in the American registry
had scarring, milia, and nail dystrophy, respectively.4
As a practical correlate, this means that the clinician
cannot reliably use cutaneous morphological features,
either singly or in combinations, to accurately assist in
the diagnosis and classification of EB patients, in the
absence of concurrent laboratory testing. Indeed, we
were unable to achieve sensitivities and specificities of
greater than 90% with any of these findings, singly or
in combinations of up to three findings, even for those
findings which have previously been believed to be
good surrogate markers of different EB subtypes.
Third, we were able to use our extensive database to
develop representative diagrams depicting the rela-
tive frequency of cutaneous disease activity in each
of the major EB subtypes.5 These diagrams, for the
first time, clearly demonstrated that considerable over-
lap exists among all forms of EB, as relates to both
sites and relative frequency of skin involvement, fur-
ther raising concerns over the use of cutaneous findings
as surrogate diagnostic markers for the purpose of
subclassification. These newer observations will
undoubtedly affect any further revisions of the classi-
fication system for EB which we proposed in 1999 as
a result of early data generated on behalf of the Nation-
al EB Registry.6
Fourth, we have been able to determine the fre-
quency with which specific cutaneous findings arise
over time.4 As a result, we now know that some features
commonly used to recognize junctional and dystrophic
EB, to include scarring, nail dystrophy, and acral web-
bing, may not be present during early infancy, the time
during which confirmation of diagnosis is the most
emotionally charged. In contrast, other features, such
as exuberant granulation tissue, may disappear with
increasing age, making them useful hints for diagno-
sis only when they are present. Application of lifetable
analysis technique to our extensive dataset has also
310 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
INHERITED EPIDERMOLYSIS BULLOSA REGISTRIES
allowed us to be able to predict, at any given age of the
patient, where he or she will likely have each of these
cutaneous findings.
Fifth, we have been able to determine the frequen-
cy with which different extracutaneous complications
occur, as stratified by individual EB subtype.7 This
includes quantitation of numerous complications with-
in each of the major organ systems or anatomic sites,
to include the external eye, oral cavity, gastrointestinal
tract, genitourinary tract, lungs, heart, musculoskele-
tal system, and bloodstream. A recently published
example is a summary of the frequency with which
different complications arise within the genitourinary
tract in each of the major EB subtypes.8 It should be
emphasized that the ability to perform detailed analy-
ses, especially lifetable ones, is a reflection of the
power that can be derived from a rare disease registry
which has large numbers of well characterized patients,
since accurate estimates of such frequencies are depen-
dent on the robustness of the size of the study popu-
lation, its generalizability to an entire population of
affected patients, and the length of longitudinal follow-
up which was performed. As one derivative of such
work, we now have quantitative data to demonstrate
that most subtypes of EB are prone to develop at least
several of the extracutaneous complications which
were previously believed to occur in only the most
severe forms of inherited EB.7
Analogous to what we did with cutaneous findings,
we have been able to use lifetable analyses to predict
the cumulative and conditional risk of an EB patient
developing any of these extracutaneous complications
over time, as stratified by EB type and subtype. This
is important, since it provides useful clinical infor-
mation to the physician as to when these patients need
to be screened most carefully for early signs of extra-
cutaneous complications (i.e., early surveillance should
have a beneficial impact on the success of medical or
surgical intervention). These types of data also pro-
vide information on the natural history of EB, as per-
tains to extracutaneous disease activity. Several years
ago, for example, we reported on the cumulative risk
of EB subtypes developing esophageal strictures and
upper airway obstruction.9 In the case of esophageal
strictures, about 12% and 80% of all Hallopeau-
Siemens recessive dystrophic EB (RDEB-HS) patients
could be predicted to develop this by ages 2 and 20,
respectively, as would 25% of all Herlitz JEB (JEB-H)
patients develop tracheolaryngeal stenosis or obstruc-
tion by age 3. Such data argue for meticulous surveil-
Agosto 2005
INHERITED EPIDERMOLYSIS BULLOSA REGISTRIES
lance during infancy for both of these outcomes, since
esophageal strictures will impair nutritional intake
and growth if not corrected by dilatation, and upper air-
way obstruction in the setting of JEB may be life-
threatening. More recently published examples of the
types of analyses possible from a rare disease registry
include estimates for the cumulative risk of EB patients
developing hand and foot deformities,10 and corneal
blisters and scars.11 Not surprisingly, mitten deformi-
ties were primarily seen in RDEB-HS, occurring as
early as the first year of life, and were present in vir-
tually all patients by age 25. In regard to corneal injury,
blisters occurred more frequently and earlier than scar-
ring, and arose primarily in patients with RDEB-HS
and JEB-H.
Sixth, and possibly the most important consequence
of our database, has been our ability to rigorously
address the issue of skin derived cancers in inherited
EB. We made this a major goal of this project from the
onset, since many older case reports and small case
series had suggested that tumors might be prevalent in
at least some of the EB subtypes. As a result of 16
years of systematic data collection by the American reg-
istry, we now know the following about squamous cell
carcinomas (SCCs) and EB.12 First, SCCs are most
commonly seen in RDEB-HS, but also occur in other
forms of RDEB, as well as in junctional EB. In contrast,
there is no increased cumulative risk over lifetime of
SCCs developing in EB simplex or DDEB, when com-
pared to the non-EB population, despite suggestions to
the contrary in some older publications which were
based on case reports. Second, the onset of SCCs is
usually first seen within the latter half of the 2nd decade
of life, and by age 40 nearly 80% of all RDEB-HS
patients will have developed at least one SCC (JD
Fine, unpublished data , 2005). Third, we now know
that multiple primary SCCs are the rule, and that they
usually arise on extremities within areas of chronic
scarring or nonhealing erosions. Fourth, using lifetable
analyses we have demonstrated that about 80% of all
of our RDEB-HS patients die of metastatic SCC with-
in five years of the diagnosis of their first tumors,
despite having undergone aggressive and presumably
successful surgical excision of the primary tumor.
Fifth, there is no increased risk of EB patients devel-
oping internal cancers. Sixth, we have found that there
is a small but significant increased cumulative risk
(about 2%) of malignant melanoma arising by as ear-
ly as age 12 in patients with RDEB-HS, and a higher
than expected cumulative risk of basal cell carcino-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
FINE
mas arising in patients with Dowling-Meara EB sim-
plex by mid-adulthood, when compared to observations
within the overall American population at the same
ages (JD Fine, unpublished data, 2005). These col-
lective findings suggest that tumor surveillance for
malignant melanoma in EB patients is of importance
only in RDEB-HS, and should be done during early
childhood. In contrast, surveillance for SCCs and basal
cell carcinomas is not necessary until at least mid
young adulthood, but must become an integral part of
patient care thereafter, given the risk of mortality from
SCCs in the setting of RDEB.
Data collected on behalf of the American EB Reg-
istry has also permitted determination within each
major EB subtype of the risk of death from a variety
of causes other than cancer, to include failure-to-thrive
and sepsis (each a risk primarily in infants with junc-
tional EB), other infections (pneumonia; other), and
renal failure.13 For example, we now know that the
cumulative risk of death from renal failure in RDEB-
HS patients is approximately 12%,14 and that such
renal disease usually occurs by young to mid adult-
hood.
The collection of such a large patient cohort has
also allowed us the opportunity to study several other
issues pertinent to inherited EB. For example, we have
used a randomly selected subset of our patients to try
to address issues such as: a) the annual cost of care,
stratified by age and EB subtype; b) the relative avail-
ability of insurance reimbursement for hospitalized
and outpatient services, medications, wound healing
products, and nutritional supplements; c) the psycho-
logical impact of this disease on affected children,
adults, parents and their family units;15 and d) the
impact of this disease on basic function (activities of
daily living) and pain.16
Registries also provide the opportunity for the har-
vesting and banking of DNA, cells, and blood and tis-
sue specimens from a large number of well charac-
terized patients. These samples can be then used by
investigators worldwide as new opportunities arise for
basic research. Over the past 10 years, for example, our
patient population has served as a major source for
tissues from which many of the molecular defects in
EB have subsequently been detected.
Finally, rare disease registries can serve as a major
resource from which patients can be recruited for clin-
ical trials. This is invaluable, since it will be other-
wise impossible for any investigator working in one
311
FINE
geographical area to successfully recruit a sufficient
number of well characterized patients to be able to
rigorously test the efficacy of a clinical intervention and
have any hope of achieving statistically interpretable
conclusions. Recently, for example, we used our cohort
to ascertain the clinical efficacy of tetracycline in EB
simplex 17 and to determine whether systemic
isotretinoin might be safe enough to be used as a pos-
sible chemopreventive agent against SCCs in patients
with RDEB.17, 18
How, then, can the Italian EB Registry be used most
effectively, and what questions can be answered as a
result of its well defined and growing patient cohort?
First, it will certainly be very important for the Registry
to attempt to collect data on selected extracutaneous
complications and outcomes of interest, so that their
findings can be compared with the published experi-
ence of its American registry counterpart. This will
allow us to determine whether there are ethnic or
immunogenetic differences within these 2 geograph-
ically distinct populations that might result in signif-
icant differences being observed in the risk for devel-
opment of one or more complications across these 2
different populations. Were such differences noted,
then this would suggest the need to employ different
surveillance strategies in Italy, compared with those
proposed by us for implementation within the United
States.
Second, the Italian registry can rigorously test
whether differences in methods of treatment in Europe
and the United States significantly impact differen-
tially on patient morbidity or mortality. As a corre-
late, it is possible that differences in surveillance or
approach to evaluation and care by physicians in dif-
ferent countries may result in differences in the rela-
tive severity of some extracutaneous complications
seen within these geographically distinct EB popula-
tions.
Third, comparisons may also suggest that inherent
differences in the overall nature of the health care sys-
tems (sources of funding of medical care; relative
access to specialists; other) in these 2 countries might
similarly contribute to differences in clinical outcomes.
Although the latter is less likely to be the case in 2
highly industrialized countries, this is clearly a major
issue elsewhere in the world. For example, when we
attempted to recruit young adult RDEB-HS patients in
South America to participate in an internationally
based clinical trial, we were surprised to learn that
312 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
INHERITED EPIDERMOLYSIS BULLOSA REGISTRIES
very few of these patients survived early infancy,
undoubtedly a reflection of gross differences in the
availability and quality of care afforded to EB patients
in less industrialized parts of the world.
Fourth, the Italian EB Registry will provide a sub-
stantial patient population from which patients can be
recruited to participate in basic research or in clinical
trials. Given the many regulatory constraints on clin-
ical trials currently imposed by the Food and Drug
Administration within the United States, the avail-
ability of patients within the Italian EB Registry might
provide the opportunity to test some hypotheses more
rapidly in Europe than can be presently done in North
America.
The development of a rigorous EB registry in Europe
is long overdue. It is therefore wonderful to see that a
vibrant one is now in place in Italy under the direction
of Professor Tadini. This registry will serve as a major
reference source for future clinical trials in Europe
that require participation of more than only a few
patients with EB. From my perspective as head of the
American EB Registry, I look forward to many fruit-
ful collaborations in the future with my dermatology
counterparts in Italy.
References
1. Fine JD, Johnson LB and Suchindran CM. The National Epidermol-
ysis Bullosa Registry. J Invest Dermatol 1994;102:54S-56S.
2. Fine JD, Johnson LB, Suchindran C, Moshell A, Gedde-Dahl T. The
epidemiology of inherited EB: findings within American, Canadian,
and European study populations. In: Fine JD, Bauer EA, McGuire J,
Moshell A, editors. Epidermolysis bullosa: clinical, epidemiologic, and
laboratory advances, and the findings of the National Epidermolysis
Bullosa Registry. Baltimore: Johns Hopkins University Press; 1999.
p.101-13.
3. Fine JD, Johnson LB, Suchindran C, Carter DM, Moshell A. The
National Epidermolysis Bullosa Registry: organization, goals, method-
ologic approaches, basic demography, and accomplishments. In: Fine
JD, Bauer EA, McGuire J, Moshell A, editors. Epidermolysis bul-
losa: clinical, epidemiologic, and laboratory advances, and the find-
ings of the National Epidermolysis Bullosa Registry. Baltimore: Johns
Hopkins University Press; 1999.p.79-100.
4. Fine JD, Johnson LB, Suchindran C, Bauer EA, Carter DM, McGuire
J et al. Cutaneous and skin-associated musculoskeletal manifesta-
tions of inherited EB: the National Epidermolysis Bullosa Registry
experience. In: Fine JD, Bauer EA, McGuire J, Moshell A, editors. Epi-
dermolysis bullosa: clinical, epidemiologic, and laboratory advances,
and the findings of the National Epidermolysis Bullosa Registry. Bal-
timore: Johns Hopkins University Press; 1999.p.114-46.
5. Devries DT, Johnson LB, Weiner M, Fine J-D. Relative extent of skin
involvement in inherited epidermolysis bullosa (EB): composite
regional anatomic diagrams based on the findings of the National EB
Registry, 1986-2002. J Am Acad Dermatol 2004;50:572-81.
6. Fine J-D, Eady RAJ, Bauer EA, Briggaman RA, Bruckner-Tuder-
man L, Christiano A et al. Revised classification system for inherit-
ed epidermolysis bullosa: report of the Second International Con-
Agosto 2005
INHERITED EPIDERMOLYSIS BULLOSA REGISTRIES
sensus Meeting on diagnosis and classification of epidermolysis bul-
losa. J Am Acad Dermatol 2000;42:1051-66.
7. Fine JD, Johnson LB, Suchindran C, Bauer EA, Carter DM, McGuire
J et al. Extracutaneous features of inherited EB: the National Epi-
dermolysis Bullosa Registry experience. In: Fine JD, Bauer EA,
McGuire J, Moshell A, editors. Epidermolysis bullosa: clinical, epi-
demiologic, and laboratory advances, and the findings of the Nation-
al Epidermolysis Bullosa Registry. Baltimore: Johns Hopkins Uni-
versity Press; 1999.p.147-74.
8. Fine J-D, Johnson LB, Weiner M, Stein A, Cash S, DeLeoz J, et al.
Genitourinary complications of inherited epidermolysis bullosa (EB):
experience of the National EB Registry and review of the literature.
J Urol 2004;172:2040-44.
9. Fine JD, Johnson LB, Moshell A, Suchindran C. The risk of selected
major extracutaneous outcomes in inherited epidermolysis bullosa:
lifetable analyses of the National Epidermolysis Bullosa Registry
study population. In: Fine JD, Bauer EA, McGuire J, Moshell A, edi-
tors. Epidermolysis bullosa: clinical, epidemiologic, and laboratory
advances, and the findings of the National Epidermolysis Bullosa
Registry. Baltimore: Johns Hopkins University Press; 1999.p.193-
205.
10. Fine J-D, Johnson LB, Weiner M, Stein A, Cash S, DeLeoz J, et al.
Pseudosyndactyly and musculoskeletal deformities in inherited epi-
dermolysis bullosa (EB): experience of the National EB Registry,
1986-2002. J Hand Surg (British and European Volume) 2005;30B:
14-22.
11. Fine J-D, Johnson LB, Weiner M, Stein A, Cash S, DeLeoz J, et al. Eye
involvement in inherited epidermolysis bullosa (EB): experience of the
National EB Registry. Am J Ophthalmol 2004;138:254-62.
12. Fine JD, Johnson LB, Suchindran C, Bauer EA, Carter DM, McGuire
J et al. Cancer and inherited epidermolysis bullosa: lifetable analyses
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
FINE
of the National Epidermolysis Bullosa Registry study population. In:
Fine JD, Bauer EA, McGuire J, Moshell A, editors. Epidermolysis bul-
losa: clinical, epidemiologic, and laboratory advances, and the find-
ings of the National Epidermolysis Bullosa Registry. Baltimore: Johns
Hopkins University Press; 1999.p.175-92.
13. Fine JD, Johnson LB, Suchindran C, Bauer EA, Carter DM, McGuire
J et al. Premature death and inherited epidermolysis bullosa: contin-
gency table and lifetable analyses of the National Epidermolysis Bul-
losa Registry study population. In: Fine JD, Bauer EA, McGuire J,
Moshell A, editors. Epidermolysis bullosa: clinical, epidemiologic, and
laboratory advances, and the findings of the National Epidermolysis
Bullosa Registry. Baltimore: Johns Hopkins University Press; 1999.
p.206-24.
14. Fine J-D, Johnson LB, Weiner M, Stein A, Cash S, DeLeoz J et al.
Inherited epidermolysis bullosa (EB) and the risk of death from renal
disease: experience of the National EB Registry. Am J Kidney Dis
2004;44:651-60.
15. Fine J-D, Johnson LB, Weiner M, Suchindran C. Impact of inherited
epidermolysis bullosa on parental interpersonal relationships, mari-
tal status, and family size. Br J Dermatol 2005;152:1009-14.
16. Fine J-D, Johnson LB, Weiner M, Suchindran C. Assessment of mobil-
ity, activities and pain in different subtypes of epidermolysis bullosa.
Clin Exp Dermatol 2004;29:122-27.
17. Weiner M, Stein A, Cash S, DeLeoz J, Fine J-D. Tetracycline and
epidermolysis bullosa simplex: a double-blind, placebo-controlled,
crossover randomized clinical trial. Br J Dermatol 2004;150:613-14.
18. Fine JD, Weiner M, Stein A, Suchindran C, Johnson LB. Systemic
isotretinoin and recessive dystrophic epiderolysis bullosa (RDEB):
results of a Phase 1 clinical trial. J Invest Dermatol 2001;117:543.
313

Cutaneous malignant melanoma risk assessment
H ow does one evaluate risk factors for cutaneous
malignant melanoma? This can be a challenging
question. After all, a goal in life is to minimize risk, an
especially appealing concept for one of the deadliest
of all cancers, melanoma, for which the incidence has
been rising in America and Europe. In this issue Rube-
gni et al.1 of the University of Siena have done an
impressive evaluation of melanoma risk in 140 Italian
natives of Tuscany in describing their results of a case-
control study of melanoma risk factors conducted in
Tuscany during one winter extending from October
2002 to May 2003. They demonstrated a highly sig-
nificant difference between controls and melanoma
patients in nevi number and the presence of atypical
nevi, constitutional skin color and eye color. This study
was unique in that it employed 6 quantitative vari-
ables representing objective skin color, measured with
a Minolta CR-300 colorimeter consisting of a detec-
tor and a microcomputer. They concluded that objec-
tive skin color measurements need to be combined
with phenotypic parameters and sun exposure history
for precise ascertainment of individual melanoma risk.
Their work is consistent with other studies, includ-
ing one by Dabkowski et al.2 of a Polish population in
which an increased number of nevi (especially atypi-
cal ones), fair skin, and blue/green eyes, as well as
IN THIS ISSUE SEE PAGE 373
Address reprint requests to: R. A. Schwartz MD, Professor & Head -Der-
matology, New Jersey Medical School, 185 South Orange Avenue, Newark,
NJ 07103-214, USA. E-mail: roschwar@UMDNJ.EDU
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:315-6
A. R. SCHWARTZ
Department of Dermatology
New Jersey Medical School, Newark, NJ, USA
intense UV exposure and sunburns, were important
risk factors for melanoma development. Others have
shown additional risk factors including freckling, fam-
ily history of melanoma, and certain chromosomal
mutations or polymorphisms.3-10 Identifying these fac-
tors can facilitate recognition of precursor lesions and
the early diagnosis of melanoma, which can save lives.
A critical element of reducing melanoma risk is pre-
vention, particularly in childhood.3 Childhood ultra-
violet exposure may be a critical factor, particularly in
predisposed individuals. Physicians and health edu-
cations need to play an active role in the education
and motivation of children to reduce solar exposure.
Clearly, there are healthful benefits in some ultravio-
let light reaching human skin in order to maintain ade-
quate vitamin D metabolism. However, those at high
risk for skin cancer need to be identified and then
advised on protective measures to lessen the impact of
ultraviolet light on their skin. Efforts such as those of
Rubegni et al.1 and Dabkowski et al.2 are particularly
valuable in this regard.
In addition, there is a need to investigate the role of
specific regulatory proteins, adhesion molecules, and
other factors in the promotion of human melanocytic
neoplasia, in order to facilitate improved understand-
ing of this epidemic and reduce the risk of developing
315
SCHWARTZ
melanoma. Disorders such as oculocutaneous albinism,
xeroderma pigmentosum and congenital neurocuta-
neous melanosis can serve as important model dis-
eases.11-17 Perhaps someday there will be effective
gene therapy, or at least viable options for mending
human genes.18
References
1. Rubegni P, Sbano P, Cevenini G, Risulo M, Stanghellini E, Barbini P
et al. Methological procedure for evaluation of risk factors for cuta-
neous malignant melanoma in a representative sample of the Tuscan
population. G Ital Dermatol Venereol (in press).
2. Dabkowski J, Omulecki A, Zalewska A. Identification of melanoma
risk factors in the Polish population. Dermatol Surg 1997;23:1039-42.
3. Azfar RS, Schwartz RA, Berwick M. Primary melanoma prevention
in children. G Ital Dermatol Venereol 2004;139:267-72.
4. Desmond RA, Soong SJ. Epidemiology of malignant melanoma. Surg
Clin North Am 2003;83:1-29.
5. Rokuhara S, Saida T, Oguchi M, Matsumoto K, Murase S, Oguchi S.
Number of acquired melanocytic nevi in patients with melanoma and
control subjects in Japan: nevus count is a significant risk factor for
nonacral melanoma but not for acral melanoma. J Am Acad Derma-
tol 2004;50:695-700.
6. Youl P, Aitken J, Hayward N, Hogg D, Liu L, Lassam N et al.
Melanoma in adolescents: a case-control study of risk factors in
Queensland, Australia. Int J Cancer 2002;98:92-8.
7. Fargnoli MC, Piccolo D, Altobelli E, Formicone F, Chimenti S, Peris
K. Constitutional and environmental risk factors for cutaneous
316 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CUTANEOUS MALIGNANT MELANOMA RISK ASSESSMENT
melanoma in an Italian population. A case-control study. Melanoma
Res 2004;14:151-7.
8. Lefkowitz A, Schwartz RA, Janniger CK. Melanoma precursors in chil-
dren. Cutis 1999;63:321-4.
9. Kaszuba A, Schwartz RA, Trznadel-Budzko E, Dobrska-Drobnik G,
Seneczko M. Malignant melanoma. Part I - Epidemiology and
etiopathogenesis. Nowa Klinika 2001;8:769-73.
10. Kaszuba A, Seneczko F, Schwartz RA, Trznadel-Budzko E, Kaszuba
A. Malignant melanoma. Part II - Clinical types, diagnostics and con-
temporary methods of treatment. Nowa Klinika 2001;8:773-9.
11. Spicer MS, Stampien TM, Lambert WC, Schwartz RA, Harmon C,
Fitzgerald-Bocarsly P. Severe xeroderma pigmentosum associated
with numerous melanomas, no other skin tumors, high natural killer
cell activity, normal interferon production, and a benign course. J
Cutan Pathol 1997;24:126.
12. Leibowitz E, Janniger CK, Schwartz RA, Lambert WC. Xeroderma
pigmentosum. Cutis 1997;60:79-84.
13. Papadopoulos AJ, Schwartz RA, Sarasin A, Lambert WC. Xeroder-
ma pigmentosum variant in a Greek patient. Int J Dermatol
2001;40:442-5.
14. Cruz MA, Cho ES, Schwartz RA, Janniger CK. Congenital neurocu-
taneous melanosis. Cutis 1997;60:178-81.
15. Okulicz JF, Shah RS, Schwartz RA, Janniger CK. Oculocutaneous
albinism. J Eur Acad Dermatol Venereol 2003;17:251-6.
16. Centurión SA, Schwartz RA. Oculocutaneous albinism type 2. Acta
Dermatovenerol Alp Panonica Adriat 2003;12:32-6.
17. Kraemer KH, Lee MM, Andrews AD, Lambert WC. The role of sun-
light and DNA repair in melanoma and nonmelanoma skin cancer. The
xeroderma pigmentosum paradigm. Arch Dermatol 1994;130:
1018-21.
18. Cleaver JE. Mending human genes: a job for a lifetime. DNA Repair
(Amst) 2005; 4:635-8.
Agosto 2005

Prediction of photodynamic efficacy
P hotodynamic therapy (PDT) refers to light acti-
vation of a tumor-localizing photosensitizer to
generate highly reactive oxygen intermediates, caus-
ing selective tissue injury and necrosis by oxidizing
essential cellular components, vascular damage and/or
inflammatory reaction and immune host response.
Recently, it has been shown that apoptosis is also
involved in tumour cell death after topical photody-
namic therapy with 5-aminolevulinic acid (ALA-PDT)
within 1 day in patients with actinic keratoses (AK).
The easy accessibility of the skin to the light expo-
sure has led to an increasing interest for PDT appli-
cation in dermatology. Photodynamic therapy is main-
ly associated with the treatment of cancer, but is also
being applied to premalignant and benign diseases.
Photodynamic therapy is an alternative treatment
modality for superficial non-melanoma skin tumours
and various inflammatory, viral and other diseases,
with potentially high effectiveness and low morbidity.
Topical ALA-PDT has become a therapeutic option
of growing interest. Its main advantage is the absence
of generalized cutaneous photosensitivity. Topical
ALA-PDT involves photosensitization with endoge-
nous porphyrins and activation with visible light.
Almost all types of cells in the human body are able to
synthesise heme. The principle of ALA-PDT is that 5-
IN THIS ISSUE SEE PAGE 381
Address reprint requests to: Dr. A. D. Tosca, Department of Dermato-
logy, University Hospital of Heraklion, 71110 Heraklion, Crete, Greece.
E-mail:derma@med.uoc.gr
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:317-20
A. D. TOSCA, M. P. STEFANIDOU
Department of Dermatology
University Hospital, Heraklion, Crete, Greece
aminolevulinic acid (ALA) in excess results in a built-
up of intracellular porphyrins and especially proto-
porphyrin IX (PpIX), an extremely potent photosen-
sitizer and fluorescence emitter. In situ conversion of
ALA to PpIX is accomplished in normal and neo-
plastic keratinocytes to a different degree, by enzymes
in the heme pathway resulting to selective accumula-
tion in target-tissue and tissue-specific phototoxic
effects. The relative accumulation of PpIX in diseased
tissue is not specific for neoplastic disease and has
been shown after the application of ALA to benign
proliferative skin conditions, such as viral warts and
condylomata 1 and psoriasis. The number of possible
clinical indications expanded besides oncology and
now encompasses several inflammatory and viral skin
diseases as a consequence of experimental findings
demonstrating that PDT can promote apoptotic cell
death and modulate immune activities of the skin.
Additionally, the application of ALA leads to accu-
mulation of PpIX in hair follicles and sebaceous glands,
suggesting the potential use for disorders of skin
appendages such as acne, alopecia areata, hypertri-
chosis. New applications for ALA-PDT in dermatol-
ogy are presently being investigated such as psoriasis,
plaque-stage cutaneous T-cell lymphoma, Darier’s dis-
317
TOSCA
ease, Kaposi sarcoma, port-wine stains, lichen scle-
rosus, scleroderma, Hailey-Hailey disease.
The literature related to the use of topical ALA-
PDT is extensive for the treatment of actinic keratoses,
basal cell carcinoma and Bowen’s disease.2 Aminole-
vulinic acid, a metabolite of the heme biosynthesis
was the first agent to receive regulatory approval for the
treatment of AK in conjunction with blue light in der-
matology, in 1999 (Levulan, DUSA Pharmaceuticals,
Inc, Valhalla, NY).
Several clinical studies report high response rates in
superficial basal cell carcinoma (BCC) ranging from
79% to 100% and AK from 81% to 100%.2-4 For nodu-
lar BCC the results are more disappointing, with report-
ed response rates ranging from 10% to 42%.
The challenge to the dermatologist remains to opti-
mize outcomes following PDT therapy. The efficacy of
the treatment is dependant on many variables. Small
modifications of the treatment procedure may have
quite a significant impact on the outcome. Controlled
comparative studies or data that sufficiently allow to
compare the different modalities are still missing.
The success of treatment requires an optimal inter-
play among different parameters, such as type and
drug dose, selection of light source, depth of light pen-
etration into tissue and light dose, treatment sched-
ules, criteria of tumour and patient selection.
An efficient photodynamic agent should have sev-
eral properties: selective retention or uptake by the
target tissue, high absorbance in the useful wavelength
range for optimal tissue penetration, high quantum
yield of singlet oxygen, fast clearance from serum and
healthy tissue, short time interval between drug appli-
cation and its accumulation into lesion, high chemical
purity, low systemic toxicity and side effects, lack of
mutagenic potential.
Aminolevulinic acid is the natural precursor of PpIX
which is formed endogenously via the biosynthetic
pathway of heme. Advantages of ALA-PDT are: local-
ized and short-term photosensitivity and rapid photo-
degradation by light illumination. The induction of a
more lipophilic ester-group (ALA- methylester,
methyl-aminolevulinate) seems to enhance the selec-
tivity and deeper penetration of the photosensitizer.
Discomfort and the intensity of pain may be lower
during PDT with methyl-aminolevulinate than with
ALA.
Rossi et al.5 in this issue describe a study of PDT
using topical methyl-aminolevulinate (Metvix, Pho-
toCure ASA, Oslo, Norway) to treat 170 AK of the
318 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
PREDICTION OF PHOTODYNAMIC EFFICACY
face and scalp. A non-coherent red light source emit-
ting at 630 nm was employed with light intensity 70-
100 mW/cm2 and light dose of 37 J/cm2, with complete
response rate of 90% in 6 months.
Photoexitation of the photosensitizer by light cor-
responding to its absorption spectrum is a basic require-
ment in PDT. The light used for PDT can be provided
by incoherent light sources or laser systems. Basic
considerations for the choice of a light source are the
following: the emitted wavelength should match an
absorption peak of the photosensitizer used and the
fact that longer wavelengths penetrate deeper into tis-
sues. The depth of tissue penetration depends on the
wavelength, the absorption by endogenous chro-
mophores and the absorption by the sensitizing drug
(self-shielding), and is limited by optical scattering
within the tissue. It might even be of advantage to use
a broad band light source, since photosensitizer’s pho-
toproducts with other absorption peaks are formed
during PDT. Protoporphyrin IX has its maximum
absorption in the Soret band and additional smaller
peaks in the green and red regions. However, with red
light the interactions with chromophors of the skin
(melanin, hemoglobin) are minimized, leading to a
sufficient penetration of light into skin.
Regarding the light dose in PDT the fluence and
the intensity used are different for oncologic and non-
oncologic indications. For PDT the light source has to
provide sufficient intensity of light, up to 40 mW/cm2
for non-oncologic indications and up to 150 mW/cm2
for oncologic indications. Basically, the fluence of the
light should be less than 150 J/cm2 and the intensity of
the light source less than 200 mW/cm2 in order to
avoid photothermal effects.
The time interval between drug administration and
sufficient photosensitizer concentration determines
the optimal time point of light exposure and has been
estimated for various photosensitizers and route of
administration. Irradiation should be performed when
an optimal ratio of photosensitizer levels in tumor ver-
sus normal tissue is reached.
Some authors emphasized tumour particularities
that might lead to decreased responsiveness. Thick
nodular BCC and hypertrophic AK are rather resis-
tant, although surgical debulking of nodules and curet-
tage of scales, or repetition of the treatment may
improve clearance rates. Morpheiform and pigmented
BCC are almost always resistant to the treatment.
However, different photosensitizers and PDT regi-
mens may result in different cellular response patterns
Agosto 2005
PREDICTION OF PHOTODYNAMIC EFFICACY
even in the same type of tumour, while many specif-
ic responses may be effected only in a narrow win-
dow of time, PDT dose, or both. It becomes clear that
there is a need to monitor parameters other than deliv-
ered light dose and irradiance and to correlate them
with ALA-PDT.
Prediction of PDT efficacy could be made on the
basis of indirect data from fluorescence time course
kinetics and the direct biologic tissue response such as
clinical erythema development after PDT.
The determination of the time course of photosen-
sitizer’s fluorescence in skin lesion is crucial for effec-
tive PDT in order to maximize PDT effectiveness.
Chromophores such as porphyrins have the ability to
absorb energy when excited by light of certain wave-
length. The emission of light following absorption of
incident photons is termed “fluorescence”. Blue light
is mainly applied for fluorescence detection, due to
its reparability from the red fluorescence of PpIX gen-
erated. The penetration of blue light is only a few tenth
of millimeter and the respective fluorescence is gen-
erated only in the superficial part of the skin. This flu-
orescence yields hardly any information from deeper
structures inside the diseased tissue. The light intensity
used for fluorescence excitation is at least twenty-fold
lower as compared to PDT.
Since the PpIX accumulation after ALA-PDT is a
gradual process, fluorescence kinetics could be used to
determine the ideal time point application to initiate the
irradiation. Imaging spectroscopy using digital cam-
eras (CCD) can be used to study the time course kinet-
ics and spatial distribution of fluorescence.6 The equip-
ment needed for fluorescence imaging is a light source
for excitation, a digital camera for detection, a sys-
tem of optical filters and the software for data pro-
cessing. The fluorescence images are recorded in a
dark room. In our set-up the power density of the light
source is very low (approximately 0.5 mW/cm2) and
the exposition time very short, in order to avoid pos-
sible photobleaching of the photosensitizer. Excita-
tion is performed by light of 425±10 nm. For the study
of fluorescence kinetics, serial in vivo fluorescence
images are captured. For any image the medians of
fluorescence intensity are calculated after correcting for
background autofluorescence and the sequential inte-
grated fluorescence signals are then plotted versus
time. Two parameters are evaluated: the maximal flu-
orescence intensity, as compared with the zero time
baseline intensity (photosensitizer accumulation effi-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TOSCA
ciency) and the correlation of fluorescence spatial dis-
tribution with the area of the lesion assessed with visu-
al light (photosensitizer localization efficiency).
Serial, in vivo and real-time measurements of fluo-
rescence intensity are of particular advantage when
an endogenous photosensitizer is used, which might be
synthesized in various amounts over time in different
tumors of the same type, depending of the metabolic
activity of the target cell. Optimal destruction of skin
lesion is therefore achieved while the surrounding skin
is left intact.
In vivo fluorescence kinetics evaluation over time
showed that AK and BCC developed maximum fluo-
rescence emission intensity between 4 to 4.5 h after
application. Despite fluorescence intensity values until
14 h the irradiation starting is chosen to be early 3.5 to
5 h when high localization efficiency was also noted,
ensuring the selectivity of the procedure.6
Some authors, based on the principle that the inten-
sity of the emitted fluorescence is a function of the
amount of the sensitizer present, claimed that the flu-
orescence of skin lesions after ALA application using
a Wood’s lamp or laser-induced fluorescence provid-
ed an estimate for prediction of PDT.7 However, no
direct correlation was found between fluorescence
intensity and clinical response by other authors.8 On the
other hand, the optimal time for PDT could be deduced
from the time-dependent concentration of PpIX, since
fluorescence kinetics is a temporal process. In order to
maximize PDT effectiveness irradiation of the treat-
ment area should occur at the time of sufficiently
increased concentration of the endogenous photosen-
sitizer within the lesion and of optimal ratio of pho-
tosensitizer levels in tumour: normal tissue. Tope et al.9
studied fluorescence kinetics in BCC after oral admin-
istration of ALA and showed a full thickness PpIX
accumulation in all BCC histological subtypes and
maximal tumour: normal skin fluorescence ratios 1
to 3 h after ALA ingestion.
Based on the cumulative evidence of selective ery-
thema during ALA-PDT which might be closely cor-
related with the tissue photo-induced actions, we stud-
ied the in vivo skin color changes, as represented by
erythema development by means of a remote machine
vision system in correlation with the clinical and his-
tological responses of AK and BCC subjected to treat-
ment.3 A remarkable correlation of erythema devel-
opment and effective tissue response to PDT was
found. The skin erythema imaging is a reliable notable
319
TOSCA
marker of the phototoxic effect and, thereafter, of the
PDT effectiveness. It is interesting that contrary to the
UVB-induced erythema, where higher doses provoke
even stronger reaction, during ALA-PDT the pro-
gressive increase of energy leads the skin erythema
to a saturation level. This is probably owed to PpIX
photobleaching due to self reaction with the released
singlet oxygen and free radicals, and this may explain
why further increase of the light dose has no effect on
both the tumour response and the erythema develop-
ment.
Laser Doppler measurements for assessing blood
flow were used. By means of the Laser Doppler Per-
fusion Imager it is possible to record blood flow in
the superficial and reticular dermis over a large surface
and estimate a representative mean perfusion. Wang et
al. have investigated the perfusion in superficial BCC
and found an increased blood perfusion after topical
PDT.
A full picture of the therapeutic effectiveness of
ALA-PDT remains the histological evaluation. Alter-
native methods allow for control of PDT efficacy.
Techniques to measure photosensitizer concentration
such as microdialysis and to evaluate light fluence
within tissue, tumour tissue oxygen consumption and
radical generation are being developed to assist PDT
treatment.
Microdialysis is a technique for investigation of
drug-penetration. A semipermeable dialysis catheter is
placed in the dermis and perfused by a solution.
Depending on the concentration gradient, molecules in
the surrounding extracellular space which are at a
higher concentration diffuse into the dialysis fibre and
the opposite takes place in the case of molecules of
lower concentration. Microdialysis samples before
and after ALA application are analysed with an ion
exchange high performance chromatograph.11 The
penetration of ALA in tumour area is rapid and after 15
minutes the concentration is high and stable. On the
other hand, virtually no ALA penetrates healthy skin.
The availability of oxygen within the tissue under-
going PDT treatment is an important parameter that can
limit direct tumour kill. Since singlet oxygen arises
from ground state oxygen, rapid and substantial reduc-
320 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
PREDICTION OF PHOTODYNAMIC EFFICACY
tion of tissue oxygen tension upon illumination during
PDT were reported through damage of the vascular
system and through O2 consumption in the oxidative
reactions taking place. The rates of singlet oxygen
generation and therefore tissue oxygen consump-
tion/depletion are high when both tissue photosensitizer
levels and the fluence rate of light are high. The fluence
rate must be adjusted downward to slow oxygen con-
sumption sufficiently to facilitate the maintenance of
tissue pO2 levels during treatment.
The potential for PDT in the treatment of several
skin conditions is promising, but rigorous trials must
be performed. Further studies are required to confirm
the optimal treatment parameters and reliable predic-
tors of the therapeutic outcome.
References
1. Stefanaki IM, Georgiou S, Themelis GC, Vazgiouraki EM, Tosca AD.
In vivo fluorescence kinetics and photodynamic therapy in condylo-
mata acuminata. Br J Dermatol 2003;149:972-6.
2. Kalka K, Merk H, Mukhtar H. Photodymamic therapy in deramtology.
J Am Acad Dermatol 2000;42:389-413.
3. Tosca AD, Balas CJ, Stefanidou MP, Katsantonis JC, Georgiou SK,
Tzardi MN. Photodynamic therapy of skin malignancies with aminole-
vulinic acid. Emphasis on anatomical observations and in vivo erythema
visual assessment. Dermatol Surg 1996;22:929-34.
4. Morton CA, MacKie RM, Whitehurst C, Moore JV, McColl JH. Pho-
todynamic therapy for basal cell carcinoma – effect of tumour thick-
ness and duration of photosensitizer application and response. Arch
Dermatol 1998;134:248-9.
5. Rossi R, Mavilia L, Ghersetich I, Lotti TM. Photodynamic therapy of
actinic keratoses with methyl-aminolevulinate (Metvix). G Ital Der-
matol Venereol (in press)
6. Stefanidou M, Tosca A, Themelis G, Vazgiouraki E, Balas K. in vivo
fluorescence kinetics and photodynamic therapy efficacy of δ-aminole-
vulinic acid-induced porphyrins in basal cell carcinomas and actinic
keratoses; implications for optimization of photodynamic therapy.
Eur J Dermatol 2000;10;351-6.
7. Svanberg K, Andersson T, Killander D, Wang I, Stenram U, Anderson-
Engels S et al. Photodynamic theory of non-melanoma malignant
tumors of the skin using topical δ-aminolevulinic acid sensitization and
light irradiation. Br J Dermatol 1994;130:743-51.
8. Fijan S, Hönigsmann H, Ortel B. Photodynamic therapy of epithelial
skin tumours using delta-aminolevulinic acid and desferrioxamine. Br
J Dermatol 1995;133:282-8.
9. Tope WD, Ross EV, Kollias N, Martin A, Gillies R, Anderson RR. Pro-
toporphyrin IX fluorescence induced in basal cell carcinoma by oral
δ-aminolevulinic acid. Photochem Photobiol 1998;67:249-55.
10. Wennberg A, Larko O, Lonnroth P, Larson G, Krogstad A. Delta-
aminolevulinic acid in superficial basal cell carcinomas and normal
skin –a microdialysis and perfusion study. Clin Exp Dermatol
2000;25:317-22.
Agosto 2005

L yme borreliosis (LB) is a systemic infectious dis-
ease caused by Borrelia burgdorferi sensu lato
transmitted by Ixodes ricinus ticks. The incidence of
LB is estimated up to 70-100 cases per 100 000 inhab-
itants in Europe. Skin manifestations of the disease
were described in 1883 by Buchwald, nevertheless
the causal organism was discovered by Burgdorfer
only in 1982. Clinical manifestations of the disease
include mainly cutaneous, nervous, cardiovascular,
locomotor signs and symptoms.
LB might present in three stages: 1st early localized dis-
ease (erythema migrans [EM], borrelial lymphocytoma
[BL] or lymphocytoma borreliensis, and lymphadeni-
tis), 2nd early disseminated disease (flu-like symptoms,
secondary lesions of EM, acute neuroborreliosis, dys-
rhytmias, atrio-ventricular block, myocarditis, acute
arthritis, etc.), and 3rd late disseminated disease (chron-
ic neuroborreliosis, chronic arthritis, acrodermatitis
chronica atrophicans [ACA]).1 Every tick-bite does not
cause skin manifestations, EM develops only in 50% of
infected patients and 30-50 % of patients with EM does
not report any tick or insect bites. Each stage of LB
might have a typical skin manifestation, some of them are
pathognomic for LB, namely, annular EM in all patients
and BL localized on the ear lobe in children.
IN THIS ISSUE SEE PAGE 417
Address reprint requests to: Dr. J. Hercogová, Department of Derma-
tology, 2nd Medical School, Charles University Prague, Czech Republic,
University Hospital Bulovka, Budinova 2, 180 81 Prague 8, Czech Repu-
blic. E-mail: dermatology@fnb.cz
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:321-3
Lyme Borreliosis up-to-date
J. HERCOGOVÁ
Department of Dermatology, 2nd Medical School
Charles University Prague, Czech Republic
EM is the most frequent manifestation of Borrelia
infection, it represents also 85% of all skin manistations
of LB. Three types of EM are recognized based on
the colour of the lesion: annular patch with central
clearing (EM anulare), homogenous patch (EM mac-
ulare) and target-like lesions (EM concentricum). EM
is mostly solitary. The histopathological picture is
non-specific: superficial dermatitis, perivascular and
periadnexal infiltrate predominantly consists from
lymphocytes with some few plasma cells.
Another skin manifestation of the 1st stage is BL. It
is a rare manifestation of LB (only 5% of skin involve-
ment), it is a red-violet papule (BL papulare) or plaque
(BL infiltratum), few mm to 5 cm in diameter. The
lesion is usually solitary, localized predominantly on
the ear lobe, nose tip, areola mammae, scrotum, and
above the bone prominences. Histopathological picture
of BL is characterized by superficial and deep der-
matitis composed of lymphocytic infiltrates localized
periadnexally and perivascullary in the upper and mid-
dle dermis, with the presence of plasma cells in the
inflammatory infiltrate. Both, EM and BL could be
accompanied by regional lymphadenopathy.
Second stage of LB, includes not only skin involve-
ment, but also extracutaneous manifestations. The spiro-
chete can disseminate soon after the tick bite and neu-
321
HERCOGOVÁ
rologic symptoms or arthritis can occur in early or late
in the course.2 Those could be neurologic (lymphocyt-
ic meningitis, cranial neuritis also with peripheral facial
nerve palsy, radiculoneuropathy, rarely encephalo-
myelitis), cardiovascular (carditis), locomotor (arthri-
tis, typically in one or both knees, myositis) and others
(hepatitis, conjunctivitis, incl. general signs and symp-
toms – malaise, headache, fatique etc.). Secondary EM
follows the primary EM after one week, EM lesions
are annular and smaller than the primary one.
Third LB stage of late infection includes involvement
of the skin, joints, nerves, fatique could be a very
important symptom. Characteristic skin manifestation
is ACA which appears late in the course of LB (months
or years after infection). ACA is much less frequent
compared to EM, but its diagnostics could be very dif-
ficult. It represents 10% of all cutaneous manifestations
of LB. Initial oedema and diffuse dark erythema local-
ized predominantly on the protrudent parts of the limbs
changes into an atrophy of the skin and the adnexa. Dis-
tinct clinical types could be recognized in ACA patients
– inflammatory ACA stage is mainly macular (ACA
maculare) in 75% of ACA patients, or oedematous
(ACA oedematosa) in 13% of patients, atrophic stage
of ACA is present in 12 % of our ACA patients and it
could be of various types: ACA teleangiectatica –
teleangiectasias predominate, ACA fibromatosa –
fibrous nodules above the bone prominences, typical-
ly localized above elbow joints, ulnar aspects of the
forarms and above the interdigital and carpophalangeal
joints, and ACA atrophicans sensu stricto – atrophy of
the skin and underlying tissue. Histopahology of ACA
lesions varies on the duration of the skin lesions. At the
beginning, band-like infiltrate of lymphocytes with
plasma cells, histiocytes, oesinophils is seen in the
upper dermis, later inflammation resolves and is promi-
nent around vessels and adnexa, compact orthoker-
atosis of the epidermis and epidermis atrophy follows,
sometimes dilatation of vessels in the upper dermis
(histopathological background of teleangiectasiae).
After some years, atrophy of the dermis, incl. elastic
fibres and adnexa is present. In fibromatous ACA type,
collagen bundles are concentrically composed. Half of
ACA patients suffer also from the joint and peripher-
al nerve involvement.3 The infection heals sponta-
neously in some patients, in the others the disease con-
tinues even after the proper treatment.
Besides those characteristic manifestations some
authors hesitate, if other signs or symptoms also belong
to the clinical picture of LB. Diffuse reversible alope-
322 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
LYME BORRELIOSIS UP-TO-DATE
cia, pseudopelade Brocq, morphoea and lichen scle-
rosus et atrophicus, cutaneous marginal zone B-cell
lymphoma, anetoderma, idiopatic atrophoderma Pasi-
ni-Pierini, progressive facial hemiatrophy, are dis-
cussed in the literature.4-7
The diagnosis of LB should be based on a presence
of three factors together: opportunity for a tick exposure,
a characteristic clinical manifestation (both local and
systemic) and a confirmation of B. burgdorferi infec-
tion, with exception of pathognomic skin manifesta-
tions, i.e. anular erythema migrans (but illness of longer
than 30 days duration is required for IgG immunoblot
positivity) and papular BL on the ear lobe in a child.1
Direct proof of borrelial infection include: 1)
histopathological detection of the microorganisms in
the tissue by the light or the electron microscope, 2) iso-
lation of B. burgdorferi (sensitivity of isolation of the
skin samples in non-treated patients is 50 %), 3) con-
ventional, mainly nested polymerase chain reaction
(PCR) or LighCycler real-time PCR or template DNA
hybridization using TaqMan fluorogenic probe. Those
systems are capable to identify 1-10 cultivated
spirochaete. Isolation of borrelial DNA from clinical
samples is difficult as they contain abundance of host
DNA. Specificity of PCR is evaluated in the Southern
blots using specific probes, restriction enzymes or
sequencing of PCR products.8, 9
Indirect laboratory tests include: 1) indirect immuno-
fluorescence, 2) ELISA assay, using the whole-cell
sonificated antigen or recombinant borrelial antigens.
Serum specimen with a positive test result is further
tested with immunoblotting. However, immunoblotting
still has many problems. True standardisation of an
immunoblotting method for diagnosis of LB would
require agreement on the strains used for antigen prepa-
ration. This approach would not be possible in Europe
due to different local prevalence of species and strains
of Borrelia burgdorferi sensu lato and also hetero-
genecity within those strains.10
LB is both underdiagnosed and overdiagnosed, and
we believe variability of symptoms is in close relation-
ship to different Borrelia serotypes. B. burgdorferi sen-
su lato which has been subdivided into three
genospecies: B. burgdorferi sensu stricto, B. afzelii and
B. garinii. Some studies show that B. afzelii represents
a dominant human skin isolate in Europe, B. garinii is
mainly connected to neuroborreliosis and B. burgdore-
fri sensu stricto appears to be the major pathogen in
Lyme arthritis. The other Borreliae (e.g. B. valaisiana)
were not proved to cause the human disease. Further-
Agosto 2005
LYME BORRELIOSIS UP-TO-DATE
more, ACA is connected with B. afzelii OspA serotype
2, infection with B. garinii OspA serotype 4 correlates
with neuroboreliosis, and B. burgdorferi sensu stricto
appears to play the major role in arthritis.11-13
There are still some unanswered questions con-
cerning LB. We do not know, if co-infections with the
other tick-born pathogens, especially Anaplasma
phagocytophilum are important, and if B. burgdorferi
could be terratogenic while the infection is present
during pregnancy, how to treat and prevent the dis-
ease with its serious consequences.
The agent of human granulocytic ehrlichiosis (HGE),
A. phagocytophilum, was identified in the USA in
1994 and later in Europe. The vectors of HGE are
nymphs and adults of I. ricinus tick.14, 15 Coinfection
may alter the clinical manifestation and response to
treatment of LB and therefore they should be consid-
ered in differential diagnosis when evaluating persons
who are at the risk for tick-borne diseases. Certain
clinical features (e.g. trombocytopenia or leukopenia)
which are not typical for LB should suggest these
coinfections. Transmission of the agents of LB and
HGE by individual ticks is equally efficient and inde-
pendent. Most dually infected ticks are able to trans-
mit both pathogens to a susceptible host.16
Borrelia infections during pregnancy were consid-
ered dangerous, more recent studies have refused some
of these fears. A causal association with Borrelia infec-
tion was not proven in any infant born to 105 mothers
with EM during pregnancy, however, two abortions
and six preterm babies, including one who had car-
diac abnormalities and two who died shortly after
delivery were observed.17
Problematic effect of LB treatment is based on pos-
sible chronic character of infection, non-sufficient
knowledge on pharmacodynamic interactions of
antimicrobial agents with Borreliae, documented fail-
ure of therapy. Currently, beta-lactams, macrolides
and tetracyclines are used. The duration of treatment
(minimum 14 days) and the daily dose depends on the
LB stage and the clinical manifestations. New antimi-
crobial agents (fluoroquinolones, ketolids) are under
evaluation.18 Some studies show that a one dose pro-
phylaxis with doxycycline (200 mg) could decrease the
risk of LB transmission after a tick-bite.19 Recom-
mendations to prevent LB include avoiding exposure
to tick bites by limiting outdoor activities in tick-infest-
ed locations, using tick repellents, tucking in cloth-
ing and frequent skin inspection for early detection
and correct removal of ticks. Antibiotic prophylaxis has
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
HERCOGOVÁ
not been shown to be effective and no vaccination
available in Europe until nowadays.
References
1. McGinley-Smith DE, Tsao SS. Dermatoses from ticks. J Am Acad Der-
matol 2003;49:363-92.
2. Sood SK. Lyme disease. Pediatr Inf Dis J 1999;18:913-25.
3. Hercogová J, Brzonǒvá I. Lyme disease in central Europe. Curr Opin
Infect Dis 2001;14:133-7.
4. Hercogová J. Borrelia burgdorferi: a protagonist in Lyme disease, a
bystander in morphoea? J Eur Acad Dermat Ven 2002;16:98-9.
5. Roggero E, Zucca E, Mainetti C, Bertoni F, Valsangiacomo C, Pedri-
nis E et al. Eradication of Borrelia burgdorferi infection in primary mar-
ginal zone B-cell lymphoma of the skin. Hum Pathol 2000;31:263-8.
6. Trevisan G, Rees DHE, Stinco G. Borrelia burgdorferi and localized
scleroderma. Clin Dermatol 1994; 12: 475-9.
7. Weide B, Walz T, Garbe C. Is morphea caused by Borrelia burgdor-
feri? A review. Br J Dermatol 2000;142:636-44.
8. Xu Y, Bruno JF, Luft BJ. Detection of genetic diversity in linear plas-
mids 28-3 and 36 in Borrelia burgdorferi sensu stricto by subtrac-
tive hybridization. Mikrob Pathol 2003;25:269-78.
9. Zore A, Ruzic-Sabljic E, Maraspin V, Cimperman J, Lotric-Furlan
S, Pikelj A et al. Sensitivity of culture and polymerase chain reaction
for the etiologic diagnosis of erythema migrans. Wien Klin Wochen-
schr 2002;114:606-9.
10. Robertson J, Guy E, Andrews N, Wilske B, Anda P, Granstrom M et
al. A European multicenter study of immunoblotting in serodiagno-
sis of Lyme borreliosis. J Clin Microbiol 2000;38:2097-102.
11. Manconi RT, Hohenberger S, Jauris-Heipke S. Genetic analysis of
Borrelia garinii OspA serotype 4 strains associated with neuroborre-
liosis: evidence for extensive genetic homogeneity. J Clin Microbiol
1999;37:3965-70.
12. Ornstein K, Berglund J, Nilsson I, Norrby R, Bergstrom S. Charac-
terization of Lyme borreliosis isolates from patients with erythema
migrans and neuroborreliosis in southern Sweden. J Clin Microbiol
2001;39:1294-8.
13. Luneman JD, Krause A. Heterogenita von Borrelia burgdorferi:
Atiopathogenetische Relevanz und Klinische Implikationen. Z
Rheumatol 2003;62:148-54.
14. Hulínská D, Votýpka J, Plch J, Vlcek E, Valesova M, Bojar M et al.
Molecular and microscopical evidence of Ehrlichia spp. and Borrelia
burgdorferi sensu lato in patients, animals and ticks in the Czech
Republic. Microbiologica 2002;25:437-48.
15. Santino I, Del Piano M, Sessa R, Favia G, Iori A. Detection of four Bor-
relia burgdorferi genospecies and first report of human granulocytic
ehrlichiosis agent in Ixodes ricicnus ticks collected in central Italy. Epi-
demiol Infect 2002;129:893-97.
16. Levin ML, Fish D. Acquisition of coinfection and simultaneous trans-
mission of Borrelia burgdorferi and Ehrlichia phagocytophila by
Ixodes scapularis ticks. Infect Immun 2000;68:2183-6.
17. Maraspin V, Cimperman J, Lotric-Furlan S, Pleterski-Rigler D, Str-
le F. Erythema migrans in pregnancy. Wien Klin Wochenschr
1999;111:933-40.
18. Hunfeld KP, Kraiczy P, Kekoukh E. Standardised in vitro susceptibililty
testing of Borrelia burgdorferi aganist well-known and newly devel-
oped antimicrobial agents. Possible implications for new therapeutic
approaches to Lyme disease. Int J Med Microbiol 2002;291 Suppl
33:125-37.
19. Nadelman RB, Nowakowski J, Fish D, Falco RC, Freeman K, McKen-
na D et al. Prophylaxis with single-dose doxycycline for the preven-
tion of Lyme disease after an Ixodes scapularis tick bite. N Engl J
Med 2001;345:79-84.
323

Photodynamic therapy. When and how?
Photodynamic therapy
T he term photodynamic therapy (PDT) includes the
presence of a photosensitizer in the tissue which is
then activated by light to produce reactive oxygen species
in particular singlet oxygen which then damage and kill
the cells.1 In Europe the only registered drug for topical
PDT is the methyl aminolevulinate (MAL) which is pre-
sent in a 16% concentration in the cream Metvix®. MAL
is taken-up by the highly active cancer cells which then
produce large amounts of photosensitive porphyrins,
above all protoporphyrin IX, which then makes the cell
photosensitive to red light. By illuminating the tissue
singlet oxygen is produced which then damages and
kills the cells.
Because MAL has a high cancer tissue selectivity it
can also be used for diagnostic purposes, by illumi-
nating the MAL-treated area 3 h after application with
blue light the cancer tissue demonstrates red/pink flu-
orescence and thus clearly delineating the tumor tissue.
This procedure, called fluorescence detection, enables
dermatologists to perform guided biopsies or guided
tumor resections. Using a newly developed technical
system using a camera and digital images one can doc-
ument the findings before treatment and also when
applied after the successful treatment to demonstrate
the efficacy of the treatment.1
IN THIS ISSUE SEE PAGE 381
Address reprint requests to: L. R. Braathen, MD, PhD, MHA, Derma-
tological University Clinic, Inselspital, 3010 Bern, Switzerland.
E-mail: Lasse.r.braathen@insel.ch
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:325-7
L. R. BRAATHEN
Dermatological University Clinic, Inselspital
Bern, Switzerland
Topical photodynamic therapy procedure
Before application of the MAL cream (Metvix®) one
should remove hyperkeratosis or crusts with gentle abra-
sion. Nodular basal cell carcinomas (BCCs) more than
3 mm thick should be debulked. Bleeding can be stopped
by compression with a physiological saline-soaked cloth
or gaze. After the bleeding has stopped the cream is
applied to the lesion under occlusion for 3 h. It is also
possible to leave it on for several hours more. Thereafter,
the occlusive dressing is removed, the cream is wiped
off and the lesion is illuminated with high intensity red
light, which using lamps with light emission diodes
(LED) technology takes 8-10 min. Most patients have
light or moderate pain mainly during the illumination,
a few have stronger pain which need to be treated. Spray-
ing water to the treated area during the illumination
helps as does also an experienced nurse talking calmly
to the patient during the illumination.
Treatment of basal cell carcinoma
MAL-PDT for BCCs has been extensively studied
over the last years. In one study Solèr et al. studied the
long-term effects on MAL-PDT in 59 patients with 350
BCCs. The patients were followed for 2-4 years (mean
325
BRAATHEN
35 months) and had an overall cure rate of 79% with a
recurrence rate of 11% at 35 months and cosmetic out-
come excellent or good in 98% of the completely
responding lesions.2 Another multicenter study inves-
tigated MAL-PDT in patients with superficial and/or
nodular BCCs. Clinical remission rate 3 months after
treatment was 92% for superficial and 87% for nodular
BCCs.3 Compiled data from several trials demonstrat-
ed complete clearance rate of 87% for superficial BCCs
and 71% for nodular BCCs.4 Overall one can conclude
that MAL-PDT is an efficient therapy which can also be
repeated if recurrence occurs. In all studies a good to
excellent cosmetic result was reported.
Actinic keratosis
MAL-PDT of actinic keratosis (AK) has been inves-
tigated in several prospective studies and compared
with cryotherapy. Randomized multicenter prospective
studies of MAL-PDT compared to cryotherapy have
been performed in Europe and Australia. The com-
plete response rates with PDT in these studies were
69% and 91% as compared to 68% and 75% for
cryotherapy. All studies demonstrated an excellent or
good cosmetic outcome of PDT in close to a 100% of
the patients.5, 6 PDT is especially well suited for larg-
er areas with many AKs, i.e. field cancerization areas.
Bowen’s disease and incipient squamous cell
carcinoma
In the largest existing study of treatment of Bowen’s
disease MAL-PDT was compared with cryotherapy
and 5-fluorouracil (5-FU) in a controlled European
multicenter study (40 centers). A total of 225 patients
with 275 lesions were included in the study and MAL-
PDT induced a complete response in 93% of lesion
compared to 86% with cryotherapy and 83% with 5-
FU. After 12 months the overall lesion cure rate was
74% with MAL-PDT, 65% with cryotherapy and 62%
with 5-FU.7
Methyl aminolevulinate-photodynamic therapy
in field cancerization areas
Field cancerization areas are usually sun exposed
areas, e.g. scalp, face, ears, dorsal aspects of hands, and
326 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
PHOTODYNAMIC THERAPY. WHEN AND HOW?
décolleté which have clinical signs of seriously sun
damaged skin and recurrent AKs, BCCs, and spin-
ocellular carcinomas (SCC). These patients have, when
being treated with cryotherapy or surgery, often mul-
tiple scars and white spots from previous treatments.
MAL-PDT is very well suited for treating such larger
areas.
Organ transplanted and immunosuppressed
patients
These patients demonstrates increased skin cancer
frequency. They should be regularly seen by derma-
tologists and their AKs, BCCs, and SCCs should be
treated at an early stage. This is especially important
for the SCCs to prevent metastases. We include all
transplanted patients with immunosuppressive treat-
ment in a skin care program including surveillance of
their skin and treatment of their non-melanoma skin
cancer lesions.
Conclusions
PDT is rapidly evolving to become a routine thera-
py in dermatology. It is well documented through con-
trolled studies and thus evidence based with efficacy
comparable to other commonly used standard treat-
ments. The British Photodermatology Group has
already published guidelines for topical PDT.8 The
advantages of PDT includes simultaneous treatment of
larger areas with multiple lesions; relatively short heal-
ing period and high patient preference because of the
excellent cosmetic outcome. PDT can also be repeat-
ed in the same area if needed.
References
1. Szeimies RM, Karrer S, Abels C, Landthaler M, Elmets CA. Photo-
dynamic therapy in dermatology. In: Krutmann J, Hönigsmann H,
Elmets CA, Bergstresser PR editors. Dermatological phototherapy
and photodiagnostic methods. Berlin: Springer, 2001. p. 209-47.
2. Solèr AM, Warloe T, Berner A, Giercksky KE. A follow-up study of
recurrence and cosmesis in completely responding superficial and
nodular basal cell carcinomas treated with methyl 5-aminolaevuli-
nate-based photodynamic therapy alone and with prior curettage. Br
J Dermatol 2001;145:467-71.
3. Horn M, Wolf P, Wulf HC, Warloe T, Fritsch C, Rhodes LE et al.
Topical methyl aminolevulinate photodynamic therapy in patients
with basal cell carcinoma prone to complications and poor cosmetic
outcome with conventional treatment. Br J Dermatol 2003;149:
1242-9.
Agosto 2005
PHOTODYNAMIC THERAPY. WHEN AND HOW?
4. Zeitouni NC, Oseroff AR, Shieh S. Photodynamic therapy for non-
melanoma skin cancers. Mol Immunol 2003;39:1133-6.
5. Szeimies RM, Karrer S, Radakovic-Fijan S, Tanew A, Calzavara-Pin-
ton PG, Zane C et al. Photodynamic therapy using topical methyl 5-
aminolevulinate compared with cryotherapy for actinic keratosis: a
prospective randomized study. J Am Acad Dermatol 2002;47:258-
62.
6. Freeman M, Vinciullo C, Francis D, Spelman L, Nguyen R, Fergin P
et al. A comparison of photodynamic therapy using topical methyl
aminolevulinate with single cycle cryotherapy in patients with actinic
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
BRAATHEN
keratosis: a prospective, randomized study. J Dermatol Treat
2003;14:99-106.
7. Morton C, Horn M, Leman J, Tack B, Bédane C, Tjioe M et al. A
placebo controlled European study comparing MAL-PDT with
cryotherapy and 5-fluorouracil in patients with Bowen’s disease. J
Eur Acad Dermatol Venereol 2004;18 Suppl 2:415.
8. Morton CA, Brown SB, Collins S, Ibbotson S, Jenkinson H, Kurwa
H et al. Guidelines for topical photodynamic therapy: report of a
workshop of the British Photodermatology Group. Br J Dermatol
2002;146:552-67.
327
 GUIDELINES
G ITAL DERMATOL VENEREOL 2005;140:329-47

Guidelines in dermoscopy
S. CHIMENTI 1, G. ARGENZIANO 2, A. DI STEFANI 1, L. ANDREASSI 3, P. CARLI 4, V. DE GIORGI 4,
G. FERRARA 5, A. FERRARI 6, S. GASPARINI 7, G. L. GIOVENE 8, M. LOMUTO 9, G. MAZZOCCHETTI 10,
G. PELLACANI 11, R. PELLICANO 9, K. PERIS 6, D. PICCOLO 6, M. A. PIZZICHETTA 12, P. RUBEGNI 3,
M. SCALVENZI 13, S. SEIDENARI 11, S. SERRESI 14, I. STANGANELLI 15, B. GIANNOTTI 4

T he guidelines that we propose reflect the state of

the art in dermoscopy at the time the report was
prepared. Results of in-progress study might require
1Department of Dermatology
Università degli Studi di Roma Tor Vergata, Rome
2Department of Dermatology

some changes to the conclusions or recommenda- 2nd University of Naples, Naples

tions reported in the following. Adherence to these
guidelines should always take into consideration care
and conscientiousness in the interpretation of der-
moscopic criteria. The final outcome is to establish
the most accurate preoperative diagnosis and the
most proper management, in light of all the circum-
stances presented by the individual patient. Guide-
lines can never replace individual medical responsi-
bility.
The significance of dermoscopy
Dermoscopy is a non-invasive technique widely
used in daily practice for the early diagnosis of
melanoma.1 It has been reported that clinical exami-
nation alone is 65-80% sensitive in the diagnosis of
melanoma.2 A recent systematic review of the litera-
ture demonstrated that dermoscopy improves the diag-
nostic accuracy of melanoma up to 35% compared to
the naked eye.3 This diagnostic improvement can be
achieved only if the observer has a good degree of
Address reprint requests to: Prof. S. Chimenti, Clinica di Dermatologia,
Università degli Studi di Roma Tor Vergata, PTV-Policlinico di Tor Ver-
gata, Viale Oxford 81, 00133 Roma (Italy).
E-mail: chimenti@dermatologica.it
3Department of Dermatology, University of Siena, Siena
4Department of Dermatology, University of Florence, Florence
5Department of Pathology, Ospedale G. Rummo, Benevento
6Department of Dermatology, University of L’Aquila, L’Aquila
7Private practice, Terni
8Private practice, Perugia
9Department of Dematology, S. Giovanni Rotondo
10Department of Dermatology, P.O. di Lanciano
11Department of Dermatology,
University of Modena e Reggio Emilia
Modena e Reggio Emilia
12Department of Oncology
Centro di Riferimento Oncologico, Aviano
13Divion of Dermatology, University Federico II, Naples
14Divion of Dermatology, I.N.R.C.A. Hospital, Ancona
15Department of Oncology and Dermatology
CPO, Ravenna and Niguarda Hospital, Milan
experience in dermoscopy, otherwise for the untrained
or less experienced examiners diagnostic accuracy
can decrease as compared to the naked eye.4, 5 There-
fore, adequate training is necessary for the effective
application of this technique.6-8 In addition, the value
of formal dermoscopy teaching courses conducted by
a qualified and expert staff should be highlighted. Fur-
thermore, the inclusion of dermoscopy training with-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 329

CHIMENTI
100
90
80
70
60
%
50
40
30
20
10
0
Specificity Sensitivity
Clinical examination Pattern analysis 7-point checklist
ABCD Combined approach
Figure 1.—Specificity and sensitivity for melanoma diagnosis of clinical
examination and of the different diagnostic algorithms.
in teaching programs for residents in dermatology
could be recommended.
Integration of clinical
and dermoscopic examination
Integration of dermoscopy in the context of phys-
ical examination has been shown to improve the pre-
operative diagnosis of melanoma (Figure 1). In 1996
Menzies et al. 9 demonstrated that 9 out of 107 (8%)
melanomas did not have any melanoma-specific der-
moscopic criteria and were reported as featureless
melanomas. Such lesions were excised only on the
basis of recent changes to the lesions as observed
by the patients. Because the great importance of the
clinical evolution of the lesion, the letter E (= evo-
lution of the lesion) has been added to the ABCD
rule of dermoscopy.10 Moreover, the percentage of
correctly diagnosed melanomas is higher for in vivo
dermoscopy (face to face with the patient) compared
with dermoscopy performed on slide images of the
same cases.11 This also implies that some clinical
parameters such as the age and skin phototype of
the patient, number and characteristics of the other
nevi, location and history of the lesion, can be cru-
cial for increasing diagnostic accuracy. In the pres-
ence of any suspicious clinical data the dermatolo-
gist may focus attention on slight or less noticeable
330 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GUIDELINES IN DERMOSCOPY
dermoscopic criteria otherwise not visible with the
naked eye.11
Impact of dermoscopy
in the clinical management
of pigmented skin lesions
The role of dermoscopy in the clinical management
of pigmented skin lesions (PSL), has been recently
evaluated. In order to verify whether this technique
may decrease the number of surgical excisions of
benign PSL, a series of lesions consecutively excised
and histologically diagnosed were retrospectively eval-
uated both clinically and dermoscopically.12 The results
showed that, although the sensitivity for melanoma
was 90%, all malignant skin neoplasms (melanomas
and basal cell carcinomas, BCC) were correctly clas-
sified as equivocal lesions to be excised. In addition,
40% of clinically suspicious (false positives) PSL,
were correctly diagnosed as benign lesions by der-
moscopic examination thus avoiding unnecessary
surgery. Therefore, the use of dermoscopy to estab-
lish whether a PSL should be biopsied or not allows us
to significantly improve the clinical diagnosis and
management of PSL.12 An algorithm for the manage-
ment of PSL showing a combined clinical and der-
moscopic approach is reported in Figure 2. This algo-
rithm was designed on the basis of the mean activity
of a reference center for PSL screening and early diag-
nosis of melanoma.
Two-step procedure for dermoscopic diagnosis
Recently a diagnostic method for the dermoscopic
diagnosis of PSL was validated and proposed as stan-
dard at the Consensus Net Meeting on Dermoscopy
(CNMD). The CNMD, organized in 2000 via the
Internet between 40 experts from 14 countries, had the
objective to investigate some important issues in der-
moscopy such as the better definition and standard-
ization of dermoscopic terminology, and the repro-
ducibility and validity of the different criteria and
diagnostic algorithms.13 This diagnostic method in
dermoscopy is based on a two-step procedure. The
first step aims to differentiate melanocytic and non
melanocytic PSL. The criteria to define a specific
lesion of a melanocytic nature are: pigmented net-
work, brown globules, streaks, homogeneous blue
Agosto 2005
GUIDELINES IN DERMOSCOPY CHIMENTI

~65% PSL = Diagnosis of Follow-up in

CLINICAL EXAMINATION benign lesion selected cases

PSL
(approximately ~35% clinically
DERMOSCOPY
1 melanoma out EQUIVOCAL (Pattern analysis, ABCD, 7-Point)
of 100 patients lesions
observed)

Suspicious for Melanoma

COMBINED APPROACH: 40% of clinically equivocal lesions:
at least 1 diagnosis (clinical diagnosis of benign lesion
or dermoscopic) of melanoma

Follow-up in
EXCISION
selected lesions

Figure 2.—Algorithm for the management of pigmented skin lesions (PLS). Data related to the mean activity of a reference center for melanoma
screening.

pigmentation and parallel pattern.13 If none of those
criteria can be identified, one should recognize the
presence of criteria for the diagnosis of seborrheic
keratosis (milia-like cysts, comedo-like openings, fin-
gerprint-like structures, cerebriforme pattern with fis-
sures and ridges), BCC (arborizing vessels, leaf-like
structures, large blue-gray ovoid nests, multiple blue-
gray globules, spoke-wheel areas and ulceration), and
vascular lesions (red-blue lacunas, red-bluish to red-
dish-black homogeneous areas).14 In the absence of
any of the above mentioned criteria, the pattern is
defined as non-specific and should be considered sus-
picious for the diagnosis of melanocytic lesion.13 The
second step is useful for differentiating benign
melanocytic lesions from melanoma, and includes dif-
ferent diagnostic algorithms: modified pattern analy-
sis,14, 15 ABCD rule,16 Menzies method,9 and seven-
point checklist.17 According to the results of the
CNMD, all diagnostic methods exhibited high sensi-
tivity in the diagnosis of melanoma, although pattern
analysis has shown a significantly higher specificity as
compared to the other algorithms. However, it should
be emphasized that pattern analysis requires a higher
degree of experience in dermoscopy.13 Recently the
three-point checklist has been proposed to allow non-
expert dermoscopists to increase their sensitivity to
melanoma diagnosis (albeit with a decrease in speci-
ficity).18 The three-point checklist is a simplified
method based on the evaluation of only 3 dermoscopic
criteria: asymmetry of the lesion, the presence of an
atypical pigmented network and of blue-white struc-
tures (defined as the presence of any type of blue and/or
white color). The three-point checklist could repre-
sent a dermoscopic method for melanoma screening
also in the hands of non-experts.18
Melanoma-specific dermoscopic criteria
In recent years, several studies have demonstrated
the validity and reproducibility of certain dermoscopic
criteria, significantly associated to melanoma diag-
nosis, and therefore defined as melanoma-specific
criteria 13-15, 19-28 (Table I). Of the various global fea-
tures, the multicomponent pattern, defined as the com-
bination of 3 or more dermoscopic patterns in a giv-
en PSL, was the most predictive for the diagnosis of
melanoma.13 The globular, cobblestone, homoge-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 331

CHIMENTI GUIDELINES IN DERMOSCOPY

TABLE I.—Melanoma-specific dermoscopic criteria showing a statistically predictive value for melanoma diagnosis, and corresponding histo-
pathologic substrates.13-15, 19-28
Association with
Melanoma-specific Dermoscopic description Histopathologic substrates melanoma diagnosis
dermoscopic criteria (Odds ratio) 13

Multicomponent pattern Combination of 3 or more dermoscopic pat- 4.3

terns in a given PSL
Atypical pigment network Black, brown or gray network, with irregular Thickened and irregularly broadened rete rid- 9,3
holes and thick lines, sharply interrupted at ges; a loss of rete ridges may occur with the
the periphery of the lesion progression of the melanoma
Irregular streaks Irregular bulbous or linear structures, irregu- Confluent junctional nests of melanocytes 5.8
larly distributed at the edge of the lesion,
not clearly associated with pigment net-
work lines
Regression structures White scar-like areas and/or blue peppering- Thickened papillary dermis with fibrosis 5.4
like granules. Usually corresponding to a and/or variable amounts of melanopha-
clinically flat part of the lesion ges
Irregular dots/globules Black, brown, round to oval variously sized Pigment aggregates within the stratum cor- 4.8
structures, irregularly distributed within neum (or even a sign of pagetoid invasion
the lesion of the epidermis) / nests of melanocytes
at the dermo-epidermal junction or papil-
lary dermis
Irregular blotches Black, brown and/or gray structureless areas Hyperpigmentation throughout the epidermis 4.1
with irregular shape and asymmetrical and/or upper dermis
distribution within the lesion
Blue-whitish veil Irregular structureless area of a confluent blue Acanthotic epidermis with focal hypergranu- 2.9
pigmentation with an overlying whitish losis above sheets of heavily pigmented
“ground-glass” film, usually correspon- melanocytes in the dermis
ding to a clinically elevated part of the
lesion
Asymmetry Asymmetry in shape of the lesion and in distri- 13.7 (asymmetry on both
bution of colors and structures (as calcu- axes to ABCD rule)
lated by both ABCD rule and Menzies 43.8 (asymmetry ac-cor-
method) ding to Menzies method)

neous, and starburst pattern were highly predictive
of a diagnosis of benign melanocytic lesions.13 Atyp-
ical pigment network, irregular streaks and regres-
sion structures were the local features that showed
the highest association with melanoma, followed by
irregular dots/globules, irregular blotches, and a blue-
whitish veil.13 In contrast, the typical pigment net-
work, regular dots/globules, regular streaks and reg-
ular blotches were associated with benign melanocyt-
ic lesions.13 Structural asymmetry of the lesion, as
assessed by the ABCD rule or Menzies method, was
also significantly associated with melanoma.13 All
melanoma-specific criteria have a well defined
histopathological substrate (Table I) and their presence
should always be investigated in a PSL. The obser-
vation of melanoma-specific criteria is almost always
sufficient to decide to surgically excise and histopatho-
logically examine a given lesion.
The problem of false negatives
and equivocal lesions
False negatives in dermoscopy mainly include
melanomas which are misdiagnosed and, therefore,
not referred for surgical excision. False negative
melanomas may essentially simulate a benign
melanocytic lesion such as Clark nevus and Spitz/Reed
nevus, or may lack any distinctive dermoscopic crite-
ria (featureless melanoma), or shows hypo- or no pig-
mentation.9, 14, 29 From a practical point of view, when
a lesion is clinically suspicious and, at first glance, is
dermoscopically benign, it is crucial to perform an
accurate dermoscopic examination in order to identi-
fy any subtle atypical features. However, the diagno-
sis of melanoma should always be suspected when a
lesion shows a non-specific pattern. In addition, clin-
ical history is an essential integration to dermoscopy

332 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

GUIDELINES IN DERMOSCOPY
when featureless lesions are diagnosed.10 In hypopig-
mented lesions the detection of an atypical vascular pat-
tern (milky-red areas/globules, linear-irregular ves-
sels or a combination of dotted and linear-irregular
vessels) 29, 30 can be highly suggestive of melanoma,
especially if associated with other criteria such as
irregular blotches, dots/globules, regression structures
or a blue-whitish veil.30 In cases of pink or amelanot-
ic lesions, vascular patterns alone may not be suffi-
cient for the diagnosis of melanoma, and should be
integrated with clinical information such as age, sex,
personal or family history of melanoma, number and
sites of lesions, time of onset and description of any
changes over time. A combined approach (dermo-
scopic examination and clinical data) may help in the
early detection of amelanotic melanoma.30
Moreover, a percentage of lesions that are equivo-
cal by clinical and dermoscopic examination, may still
be equivocal after histopathologic examination.31 The
limit between benign and malignant lesions is not clear
for lesions such as junctional Clark nevi (and
melanoma in situ) or atypical Spitz/Reed nevi (and
spitzoid melanomas). Skin lesions within this “gray
zone” should be surgically excised or followed up
closely (1-3 months) by dermoscopy in order to detect
a potential asymmetric enlargement or changes in der-
moscopic features.32, 33 Recently a new dermoscopic
classification of Clark nevi has been proposed to select
specifically those lesions which should be surgically
excised.34
An eccentric peripheral hyperpigmentation or the
presence of 3 different structures in a given lesion
(reticular, globular and homogeneous pattern) were
significantly more frequently found in malignant than
in benign melanocytic lesions and this implies the sur-
gical removal of the lesion.35
A model of possible natural evolution over time has
been described with sequential dermoscopic exami-
nations for spitzoid lesions: from a globular pattern
to a starburst pattern,36 and subsequently the disap-
pearance of streaks at the periphery of the lesion and
the presence of a central homogeneous pigmenta-
tion.37, 38 In a percentage of spitzoid lesions the detec-
tion of a superficial black network (that histopatho-
logically corresponds to focal areas of pigmented
parakeratosis, producing a black reticulated appear-
ance on the horizontal plane) can be useful for the
diagnosis of benignancy.39 The age of the patient is
an important clinical parameter for the management of
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHIMENTI
these lesions: surgical excision should be performed for
any spitzoid lesion occurring in adult patients, while
in a child, a spitzoid lesion showing typical dermo-
scopic features could be dermoscopically monitored
over time.40 Melanocytic skin lesions showing fea-
tures of regression (blue-white structures such as white
scar-like depigmentation and blue pepper-like granules)
may be difficult to classify clinically and dermoscop-
ically. A recent study of dermoscopic-pathologic cor-
relation on clinically equivocal melanocytic lesions
with blue-white structures demonstrated that the major-
ity of nevi with regression exhibited blue areas with a
central distribution and involving <50% of the lesion
surface, while histopathologic equivocal lesions
revealed a combination of white and blue areas with an
irregular distribution and involving >50% of the lesion
surface.41 Based on those results, an algorithm was
proposed that can be applied to the management of
lesions exhibiting dermoscopic features of regression:
lesions showing a low degree (<10%) of blue-white
structures can be dermoscopically monitored over
time; in contrast, lesions with a high degree of regres-
sion (>50%) or with a moderate degree (between 10%
and 50%) of regression along with the presence of a
combination of blue and white areas should be surgi-
cally excised and histopathologically examined.41
Dermoscopic follow-up and modification during
time of pigmented skin lesions
There are 2 main reasons why a patient must be
periodically examined: first, some patients have a high-
er risk of developing a melanoma (i.e. personal or fam-
ily history of melanoma, total nevus count, skin pho-
totype I-II); second, to monitor the possible evolution
of single only moderately atypical lesions (not suspi-
cious for melanoma) over time. In addition, dermo-
scopic follow-up is useful in patients with multiple
nevi, often clinically atypical, which would be practi-
cally impossible to remove simultaneously.1, 42, 43
In a recent study the characteristics of growing
melanocytic nevi were described: in a series of 1 612
common nevi, 5% showed an enlargement in a mean
follow-up period of 12 months.44 Dermoscopy revealed
a peripheral symmetric rim of brown globules in 50%
of enlarging nevi, due to the junctional activity of
melanocytes. Although this phenomenon was more
common in the under-20s, symmetrical enlargement
alone (without any other atypical feature) did not indi-
333
CHIMENTI
cate malignancy, as demonstrated by histopathologi-
cal examination of these lesions.44 However, enlarging
lesions in adults, especially if showing a peripheral
rim of brown globules, should be carefully monitored
over time (3 months dermoscopic follow-up) or sur-
gically excised.45
Other studies demonstrated the efficacy of der-
moscopic follow-up in detecting patterns of modifi-
cation of PSL over time. 46 Atypical nevi showed
focal enlargement without substantial structural der-
moscopic changes. In contrast, melanomas showed
focal enlargement associated with a change in shape
as well as appearance of dermoscopic structures such
as irregular black dots, atypical network, regression
structures, irregular streaks or a blue-whitish veil.47
The detection of changes in dermoscopic criteria
(i.e. extension or loss of pigment network, distribu-
tion or number of black dots and/or hypopigmented
areas or regression structures), in a dermoscopic fol-
low-up examination, should suggest the surgical
excision of the lesion. Similarly, the appearance of
atypical dermoscopic features such as irregular black
dots, atypical network, regression structures, irreg-
ular streaks, a blue-whitish veil and atypical vascu-
lar pattern, are indicative of a suspicious lesion to
be removed. Moreover, Menzies et al.32 demonstrated
that a short term follow-up (median 3 months) of
318 only moderately atypical lesions revealed 7 ear-
ly dermoscopically featureless melanomas. Such
melanomas did not show any atypical dermoscopic
criteria but could be identified only by morpholog-
ic changes over a short period of time.32 Remark-
ably there are some risks in performing a dermo-
scopic follow-up of atypical lesions. In a recent study,
some authors 48 claimed that the uncritical use of
sequential imaging cannot be recommended, since the
usefulness of this technique depends on the experi-
ence in the interpretation of follow-up images and on
the patient’s compliance with a scheduled follow-
up program over time. The selection of patients and
lesions submitted for follow-up examination must
be carefully performed in order to avoid the risk of
leaving a melanoma unexcised.33 Recently Carli et
al.49 demonstrated in a randomized study on 938
subjects, that dermoscopic follow-up of equivocal
lesions is associated with a reduction in the number
of PSL excised for diagnostic verification but also
with a non-negligible occurrence of initial melanomas
left unexcised.
334 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GUIDELINES IN DERMOSCOPY
Furthermore, concerning the dermoscopic modi-
fication of melanocytic lesions, the effects of ultra-
violet irradiation, including increased pigmentation
and irregular distribution of the pigment, increased
dimension of brown globules, decrease of hypopig-
mented areas and less visibility of the pigment net-
work are well known.50-53 Sun-induced morpholog-
ical changes are transient and presumably related to
activating melanocytes.50-53 Several studies emphasize
the need to re-examine the lesions 4-6 weeks after sun
exposure, since the differentiation from melanoma
can be difficult in the period following ultraviolet
irradiation.50-53
In general, dermoscopic follow-up examination over
time should be performed only for moderately atypi-
cal lesions, which are not elevated on the skin surface,
with no melanoma-specific features or history of
changes. Nodular lesions showing atypical features
should never be submitted to a dermoscopic follow-up,
since it is not possible to rule out the diagnosis of
nodular melanoma. In such cases surgical excision is
mandatory.
Main dermoscopic features of the most difficult
pigmented skin lesions for clinical management
and the differentiation with melanoma
The dermoscopic detection of melanoma-specific
criteria, as discussed in a previous section, should
always lead to surgical excision of a given lesion. In
Table II 13, 14, 30, 34-36, 38, 40, 41, 54-70 the main dermoscop-
ic features of the most difficult PSL that commonly rep-
resent false negatives or false positives, and some clues
for an accurate differential diagnosis and a better clin-
ical management are reported.
In Clark nevi, as already mentioned, the detection of
an eccentric peripheral hyperpigmentation or the pres-
ence of reticular, globular and homogeneous struc-
tures in the same lesion is important to recommend
surgical excision, because the same characteristics can
frequently be found in melanoma.34, 35 If a patient
shows multiple lesions with those dermoscopic fea-
tures, it is reasonable to excise the most atypical ones
and to perform a short term digital follow-up in the
others.32, 34 For nevi showing dermoscopic features of
regression a model for the management is reported in
Table II.41
Spitz/Reed nevi (spindle and/or epithelioid cell nevi,
including the pigmented variant, previously consid-
Agosto 2005
GUIDELINES IN DERMOSCOPY CHIMENTI

TABLE II.—Main dermoscopic features of the most difficult PSL (equivocal lesions and false negatives) with clues for differential diagnosis
and clinical management.
Skin lesions Main dermoscopic features and management recommendations

Equivocal lesions
1) Clark nevus with eccentric hyper- Excision if solitary lesion; short term follow-up (3 months) if multiple lesions with this dermoscopic fea-
pigmentation ture in the same patient 32, 34, 35
2) Clark nevus with regression Excision if regression >50% of the lesion surface; follow-up if regression <10%; excision if regression
10-50% together with the a combination of blue and white areas 41
3) Spitz/Reed nevus Excision in adult patients,40 short term digital follow-up if typical dermoscopic features in children 36, 38
False negatives
1) Melanoma simulating the following
PSL:
— Clark nevus Short term follow-up (3 months) if multiple atypical lesions in the same patient;32, 34, 35 excision if solitary
lesion even if slightly atypical. Excision if simultaneous presence of reticular, globular and homogeneous
structures 35
— Spitz/Reed nevus Excision in adult patients,40 short term digital follow-up if typical dermoscopic features in children 36, 38
— Dermal nevus Differential criteria to evaluate:
— cobblestone pattern, comma-like vessels and hair favor a dermal nevus 14
— asymmetry, blue-whitish veil, irregular dots/globules and atypical vascular pattern favor a melanoma, also
with the presence of a cobblestone pattern 13 Dermoscopic follow-up in nodular lesion should be avoi-
ded.
— Pigmented lesions of the face Diagnostic criteria to evaluate:
— annular-granular structures
— asymmetric pigmented follicular openings
— rhomboidal structures 54 surgical excision or incisional biopsy in suspicious (blue-gray) areas in broad
lesions
— Acral nevus Surgical excision of the lesions showing:70
— parallel-ridge pattern
— atypical or multicomponent pattern
— Ungueal lesions Incisional biopsy if:55-57
— brown background and longitudinal brown to black lines, irregular in shape, coloration, thickness and
parallelism.
— micro-Hutchinson sign
Dermoscopic follow-up in doubtful lesions 56
— Labial and genital melanosis Incisional biopsy in lesions with variegated coloration and irregular distribution of the pigment 58
Excision of nodular lesions
— Blue nevus Excision when dermoscopically atypical or clinical history unclear 59 Rule out a melanoma metastasis 60
— Congenital nevus Dermoscopic follow-up by dermoscopic images of:61, 62
— entire lesion if possible
— representative areas of the architectural pattern
— borders
— special interest areas
Excision or incisional biopsy of suspicious lesions
— Recurrent nevus Atypical dermoscopic features (irregular streaks and dots/globules),63 definite clinical history is essential
if re-excise or not the lesion
— Irritated nevi Close follow-up (1-2 weeks); excision in unsolved cases
— Reticulated lentigo Thickened, dark brown to black pigmented network, with irregular and asymmetric mashes.64 Excision in
equivocal lesions
— Basal cell carcinoma Arborizing vessels, leaf-like areas, large blue-gray ovoid nests, multiple blue-gray globules, spoke wheel areas and
ulceration, with the absence of pigment network 65, 66. Surgical excision is the elective treatment
— Seborrheic keratosis — evaluate the number of milia-like cysts and comedo-like openings: often numerous in seborrheic kera-
tosis, few in melanoma 67
— excision in suspicious lesions, showing false pigment network and/or pseudo-globular structures 67-69
— Vascular lesions — Haemangioma: well circumscribed red lacunas (must be differentiated from milky red areas (less defi-
ned) in melanoma 14
— Pyogenic granuloma: excision and histopathologic exam in adults
2) Melanoma with non-specific pattern Excision of the lesion in the absence of specific dermoscopic criteria 14
3) Amelanotic melanoma Suspicious dermoscopic criteria in pink lesions requiring excision 29, 30
— nodular ulcerated lesion
— pigment remnants, especially if blue-gray in color
— milky red areas
— atypical vascular pattern (dotted and/or linear-irregular vessels)
4) Melanoma metastasis — homogeneous bluish pigmentation
— diffuse non-homogeneous brown-bluish pigmentation
— red-bluish globular pattern 60

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 335

CHIMENTI
ered as a distinctive entity named Reed nevus) can
dermoscopically show different dermoscopic patterns:
starburst, globular, reticular, homogeneous, hypopig-
mented and atypical.37 Often these lesions can hardly
be differentiated from a spitzoid melanoma, either
dermoscopically or histopathologically. Consequent-
ly, any spitzoid lesion in adult patients should be sur-
gically excised 40 while in childhood, if the spitzoid
lesion shows typical dermoscopic features, could be
closely followed-up over time, avoiding unnecessary
excisions of benign lesions.36, 38
Dermal nevi (Unna and Miescher nevi) are usually
characterized by a homogeneous pigmentation (that on
the face appears as a pigment pseudonetwork), by a
cobblestone pattern and comma-like vessels.14 In addi-
tion, dermal nevi may show exophytic papillary struc-
tures and irregular crypts.14 The detection of asym-
metry, a blue-whitish veil, irregular dots/globules and
atypical vascular pattern is suspicious of melanoma.13
When a nodular lesion is observed, a follow-up is nev-
er recommended in the presence of even a minimal
diagnostic suspicion.
Pigmented lesions of the face show a characteristic
pigment pseudonetwork, due to the distribution of the
pigment around follicular ostia.14 Criteria to diagnose
a lentigo maligna include: annular-granular structures
(of blue-gray color) and asymmetric pigmented fol-
licular openings.54 Rhomboidal structures and homo-
geneous pigmented areas of follicular invasion are
indeed associated with progression to an invasive
melanoma (lentigo maligna melanoma).54 Surgical
excision should always be performed when these fea-
tures are observed, while an incisional (punch) biop-
sy can be performed in large lesions, preferentially
within the bluish black areas.
Acral nevi (nevi of the palms and soles) show a
characteristic parallel pattern due to the disposition
of the pigment along the sulci. The parallel-furrows pat-
tern, the lattice-like pattern and the fibrillar pattern
are typical of benign acral melanocytic lesions while
the parallel-ridge pattern is associated with acral lentig-
inous melanoma.70 Lesions showing atypical or mul-
ticomponent patterns should be surgically excised and
histopathologically examined.
Ungueal lesions should be distinguished from sub-
ungueal hemorrhage, melanocytic nevi and drug-
induced longitudinal melanonychia. Subungueal
hemorrage is characterized by a clinical history of
trauma and dermoscopically by roundish sharply
336 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GUIDELINES IN DERMOSCOPY
demarked black-reddish areas and by blood spots or
purple-to-black dots. Melanocytic nevi show a brown
coloration of the background and the presence of reg-
ular lines in shape and parallelism, and drug-induced
longitudinal melanonychia reveals a grayish col-
oration of the background and the presence of regu-
lar thin lines.55 When a diagnosis can not be made
with certainty, a close dermoscopic follow-up can be
useful to establish the final diagnosis.56 Subungueal
melanoma dermoscopically shows a brown back-
ground and longitudinal brown to black lines, irreg-
ular in shape, coloration, thickness and parallelism.
Important is the detection of micro-Hutchinson sign
(pigmentation of the periungueal skin and cuticle,
visible only at dermoscopic examination) that,
although rarely detected, is suspicious of melanoma.55,
57 In equivocal lesions, an incisional biopsy, includ-
ing the matrix, is required.
Labial and genital melanosis are benign pigmented
lesions dermoscopically characterized by a diffuse
background pigmentation with a granular or globular
(often with aligned globules) or linear-curvilinear, fre-
quently parallel, intensification of the pigment, light
brown to brown or grayish in coloration.71 Dermo-
scopic follow-up of lesions showing even a slightly
irregular pigmentation is recommended. In equivocal
lesions showing variegated coloration and irregular
distribution of the pigment, dermoscopy can help to
identify the most atypical zone in order to perform an
incisional biopsy at the diagnostic site.58
Blue nevi are easily identified for their typical homo-
geneous blue pigmentation, common locations on the
extremities and without history of changes. In some
cases, characterized by yellow-whitish areas, related
to the presence of fibrosis, the main differential diag-
nosis is melanoma. Because of the important signifi-
cance of blue structures in dermoscopy, a careful exam-
ination of these lesions is always recommended and a
biopsy is suggested when lesions are dermoscopical-
ly atypical or the clinical history is unclear.59 Fur-
thermore, melanoma metastasis may simulate a blue
nevus.60
Dermoscopy can also be useful to further charac-
terize and follow up congenital nevi, although their
dermoscopic features are quite variable. The cobble-
stone pattern is the most common, followed by a mul-
ticomponent pattern consisting of multiple colors,
dots/globules, pigmented network, hypopigmented
areas and homogeneous blue areas.61 Recently, other
Agosto 2005
GUIDELINES IN DERMOSCOPY
peculiar features have been described such as target
network, target globules and target vessels.62 A care-
ful examination of a sequential digital clinical image
of the entire lesion, integrated with a dermoscopic
image representative of the architectural pattern, bor-
ders, and areas of special interest, may make it possible
to identify the appearance of atypical features.61 In
such case, small congenital nevi should be excised,
while in large or giant congenital nevi, an incisional
biopsy in dermoscopically suspicious areas is rec-
ommended.
Recurrent nevi (or persistent nevi) usually exhibit
such bizarre and atypical features to be suspicious for
melanoma, including irregular streaks and irregular
dots/globules close to or in the context of the scar.63
Recurrent nevi following incomplete excision of a
histopathologically non atypical lesion can be moni-
tored over time. When anamnestic data or previous
histopathologic diagnosis are not available or not clear,
the lesion must be excised.
Another example of the usefulness of a dermoscopic
follow-up is the irritated nevi (mostly by trauma or
infections), also named Meyerson nevi:72 a close dig-
ital monitoring (1-2 weeks) after local treatment can
solve the diagnostic doubt.
Reticulated lentigo (or ink spot lentigo) is usually
located in severe sun damaged skin and may mimic a
melanoma in situ. Dermoscopically it is characterized
by a thickened, dark brown to black pigmented net-
work, with irregular and asymmetric mashes which
are uniformly distributed throughout the lesion.64 In
equivocal lesions a biopsy is recommended.
BCC especially if pigmented, may simulate a
melanoma. Dermoscopic hallmarks of BCC are:
arborizing vessels, leaf-like areas, large blue-gray
ovoid nests, multiple blue-gray globules, spoke wheel
areas and ulceration, in the absence of a pigment net-
work.65, 66
Seborrheic keratosis, mainly the acanthotic and pig-
mented variants, may mimic a melanoma. Typical der-
moscopic features include: milia-like cysts, comedo-
like openings, fingerprint-like structures and a cere-
briforme pattern with fissures and ridges. Recently
other dermoscopic criteria have been described such as
sharp demarcation and moth-eaten borders,68 above
all in the early reticulated type of seborrheic keratosis
arising from a solar lentigo. A false pigment network,
especially of the reticulated type, and pseudo-globu-
lar structures can sometimes be observed.68, 69 A sur-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHIMENTI
gical excision is suggested in equivocal lesions to rule
out the possibility, though rare, of a melanoma simu-
lating a seborrheic keratosis.67
Dermoscopic examination of vascular lesions
allows us to detect specific criteria and to differen-
tiate with accuracy vascular lesions from melanoma.
Haemangioma is characterized by a lacunar pattern,
composed of numerous red to red-bluish ovoid,
sharply circumscribed areas, called red lacunas.
These structures must be distinguished from the less
defined milky-red areas, that can be sometimes but
specifically seen in melanoma.14 The lacunar pat-
tern is hardly recognized in a pyogenic granuloma,
therefore a biopsy and histopathologic confirmation
are recommended in adult patients. Angiokeratoma
exhibits red-bluish to black lacunas, associated with
whitish-yellowish keratotic areas.14 Subcorneal hem-
orrhage dermoscopically shows a blackish homoge-
neous area and also a pseudo-parallel or pseudo-
globular pattern.
Technological standard in dermoscopy
The hand-held dermatoscope and digital videoder-
matoscope represent the most widely used instruments
for dermoscopic examination.1-3, 14 In some derma-
tology centers, a stereomicroscope implemented by
digital system such as a high resolution digital camera
(3CCD) are employed.3, 52, 53 The current standard tool
for dermoscopic photographic documentation of PSL
is the Dermaphot (Heine Optotechnik, Herrsching,
Germany), consisting of a special designed lens on a
camera with high resolution power and optimal image
quality.73-75 However, a large number of instruments
and systems for digital videodermoscopy are current-
ly commercially available, achieving an image quali-
ty not always comparable to the standard.76, 77 We hope
that in the near future the companies will conform to
the standard, possibly through the institution of a spe-
cific committee for the evaluation and validation of
videodermoscopy systems.
Video-dermoscopic report controversies
At the moment, there are no precise regulations nor
published papers regarding this topic. In the present
we would suggest a proposal of standardization of
the video-dermoscopic report. Considering together
337
CHIMENTI
medico-legal and deontological matters, and aiming
a full statement of dermoscopy as a second-level
instrumental examination, it is evidently important
to issue a suitable report after providing a specialized
service.
Hence, in every video-dermoscopic report, we sug-
gest including the following points (minimal criteria):
— Symmetry/asymmetry of the lesion.
— Global pattern.
— Local features.
— Diagnostic conclusion.
— Management recommendation.
Although the different video-dermoscopy systems
have some limits in the standard of image quality and
resolution, printing quality etc., we believe that a print-
ed dermoscopic image of the lesion should be given to
the patient at the dermatologist’s discretion, specify-
ing current technological problems.
References
1. Argenziano G, Soyer HP. Dermoscopy of pigmented skin lesions-a
valuable tool for early diagnosis of melanoma. Lancet Oncol
2001;2:443-9.
2. Wolf IH, Smolle J, Soyer HP, Kerl H. Sensitivity in the clinical diag-
nosis of malignant melanoma. Melanoma Res 1998;8:425-9.
3. Kittler H, Pehamberger H, Wolff K, Binder M. Diagnostic accuracy
of dermoscopy. Lancet Oncol 2002;3:159-65.
4. Binder M, Schwarz M, Winkler A, Binder M. Epiluminescence
microscopy. A useful tool for the diagnosis of pigmented skin lesions
for formally trained dermatologists. Arch Dermatol 1995;131:286-91.
5. Piccolo D, Smolle J, Argenziano G, Wolf IH, Braun R, Cerroni L et
al. Teledermoscopy--results of a multicentre study on 43 pigmented
skin lesions. J Telemed Telecare 2000;6:132-7.
6. Binder M, Puespoeck-Schwarz M, Steiner A, Kittler H, Muellner M,
Wolff K et al. Epiluminescence microscopy of small pigmented
skin lesions: short-term formal training improves the diagnostic
performance of dermatologists. J Am Acad Dermatol 1997;36:
197-202.
7. Carli P, Quercioli E, Sestini S, Stante M, Ricci L, Brunasso G et al. Pat-
tern analysis, not simplified algorithms, is the most reliable method
for teaching dermoscopy for melanoma diagnosis to residents in der-
matology. Br J Dermatol 2003;148:981-4.
8. Pagnanelli G, Soyer HP, Argenziano G, Talamini R, Barbati R, Bianchi
L et al. Diagnosis of pigmented skin lesions by dermoscopy: web-based
training improves diagnostic performance of non-experts. Br J Der-
matol 2003;148:698-702.
9. Menzies SW, Ingvar C, Crotty KA, McCarthy WH. Frequency and
morphologic characteristics of invasive melanomas lacking specific
surface microscopic features. Arch Dermatol 1996;132:1178-82.
10. Kittler H, Seltenheim M, Dawid M, Pehamberger H, Wolff K, Binder
M. Morphologic changes of pigmented skin lesions: a useful exten-
sion of the ABCD rule for dermatoscopy. J Am Acad Dermatol
1999;40:558-62.
11. Carli P, De Giorgi V, Giannotti B. Dermoscopy as a second step in
the diagnosis of doubtful pigmented skin lesions: how great is the risk
of missing a melanoma? J Eur Acad Dermatol Venereol 2001;15:
24-6.
338 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GUIDELINES IN DERMOSCOPY
12. Argenziano G, Soyer HP, Chimenti S, Ruocco V. Impact of der-
moscopy on the clinical management of pigmented skin lesions. Clin
Dermatol 2002;20:200-2.
13. Argenziano G, Soyer HP, Chimenti S, Talamini R, Corona R, Sera F
et al. Dermoscopy of pigmented skin lesions: results of a consensus
meeting via the Internet. J Am Acad Dermatol 2003;48:679-93.
14. Argenziano G, Soyer HP, De Giorgi V, Piccolo D, Carli P, Delfino M
et al. Interactive atlas of dermoscopy. Milan: Edra Medical Publish-
ing and New Media; 2000.
15. Pehamberger H, Binder M, Steiner A, Wolff K. In vivo epiluminescence
microscopy: improvement of early diagnosis of melanoma. J Invest
Dermatol 1993;100:356S-62S.
16. Nachbar F, Stolz W, Merkle T, Cognetta AB, Vogt T, Landthaler M et
al. The ABCD rule of dermatoscopy. High prospective value in the
diagnosis of doubtful melanocytic skin lesions. J Am Acad Dermatol
1994;30:551-9.
17. Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Sammarco E,
Delfino M. Epiluminescence microscopy for the diagnosis of doubt-
ful melanocytic skin lesions. Comparison of the ABCD rule of der-
matoscopy and a new 7-point checklist based on pattern analysis.
Arch Dermatol 1998;134:1563-70.
18. Soyer HP, Argenziano G, Zalaudek I, Corona R, Sera F, Talamini R et
al. Three-point checklist of dermoscopy. A new screening method
for early detection of melanoma. Dermatology 2004;208:27-31.
19. Nilles M, Boedeker RH, Schill WB. Surface microscopy of naevi and
melanomas--clues to melanoma. Br J Dermatol 1994;130:349-55.
20. Soyer HP, Smolle J, Leitinger G, Rieger E, Kerl H. Diagnostic relia-
bility of dermoscopic criteria for detecting malignant melanoma. Der-
matology 1995;190:25-30.
21. Steiner A, Pehamberger H, Wolff K. In vivo epiluminescence
microscopy of pigmented skin lesions. II. Diagnosis of small pig-
mented skin lesions and early detection of malignant melanoma. J
Am Acad Dermatol 1987;17:584-91.
22. Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Delfino M. Epi-
luminescence microscopy: criteria of cutaneous melanoma progres-
sion. J Am Acad Dermatol 1997;37:68-74.
23. Pizzichetta MA, Argenziano G, Talamini R, Piccolo D, Gatti A, Tre-
visan G et al. Dermoscopic criteria for melanoma in situ are similar
to those for early invasive melanoma. Cancer 2001;91: 992-7.
24. Soyer HP, Kenet RO, Wolf IH, Kenet BJ, Cerroni L. Clinicopatho-
logical correlation of pigmented skin lesions using dermoscopy. Eur
J Dermatol 2000;10:22-8.
25. Massi D, De Giorgi V, Soyer HP. Histopathologic correlates of der-
moscopic criteria. Dermatol Clin 2001;19:259-68, vii.
26. Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Delfino M. Clin-
ical and dermatoscopic criteria for the preoperative evaluation of cuta-
neous melanoma thickness. J Am Acad Dermatol 1999;40:61-8.
27. De Giorgi V, Carli P. Dermoscopy and preoperative evaluation of
melanoma thickness. Clin Dermatol 2002;20:305-8.
28. Stante M, Carli P, Massi D, De Giorgi V. Dermoscopic features of
naevus-associated melanoma. Clin Exp Dermatol 2003;28:476-80.
29. Carli P, Massi D, De Giorgi V, Giannotti B. Clinically and dermo-
scopically featureless melanoma: when prevention fails. J Am Acad
Dermatol 2002;46:957-9.
30. Pizzichetta MA, Talamini R, Stanganelli I, Puddu P, Bono R, Argen-
ziano G et al. Amelanotic/hypomelanotic melanoma: clinical and
dermoscopic features. Br J Dermatol 2004;150:1117-24.
31. Ferrara G, Argenziano G, Soyer HP, Corona R, Sera F, Brunetti B et
al. Dermoscopic and histopathologic diagnosis of equivocal melanocyt-
ic skin lesions: an interdisciplinary study on 107 cases. Cancer
2002;95:1094-100.
32. Menzies SW, Gutenev A, Avramidis M, Batrac A, McCarthy WH.
Short-term digital surface microscopic monitoring of atypical or
changing melanocytic lesions. Arch Dermatol 2001;137:1583-9.
33. Carli P, De Giorgi V, Giannotti B. Dermoscopy and early diagnosis
of melanoma: the light and the dark. Arch Dermatol 2001;137:
1641-4.
Agosto 2005
GUIDELINES IN DERMOSCOPY
34. Hofmann-Wellenhof R, Blum A, Wolf IH, Piccolo D, Kerl H, Garbe
C et al. Dermoscopic classification of atypical melanocytic nevi (Clark
nevi). Arch Dermatol 2001;137:1575-80.
35. Blum A, Soyer HP, Garbe C, Kerl H, Rassner G, Hofmann-Wellenhof
R. The dermoscopic classification of atypical melanocytic naevi (Clark
naevi) is useful to discriminate benign from malignant melanocytic
lesions. Br J Dermatol 2003;149:1159-64.
36. Pizzichetta MA, Argenziano G, Grandi G, De Giacomi C, Trevisan G,
Soyer HP. Morphologic changes of a pigmented Spitz nevus assessed
by dermoscopy. J Am Acad Dermatol 2002;47:137-9.
37. Peris K, Ferrari A, Argenziano G, Soyer HP, Chimenti S. Dermo-
scopic classification of Spitz/Reed nevi. Clin Dermatol 2002;20:
259-62.
38. Piccolo D, Ferrari A, Peris K. Sequential dermoscopic evolution of pig-
mented Spitz nevus in childhood. J Am Acad Dermatol 2003;49:
556-8.
39. Argenziano G, Soyer HP, Ferrara G, Piccolo D, Hofmann-Wellen-
hof R, Peris K et al. Superficial black network: an additional dermo-
scopic clue for the diagnosis of pigmented spindle and/or epithelioid
cell nevus. Dermatology 2001;203:333-5.
40. Argenziano G, Scalvenzi M, Staibano S, Brunetti B, Piccolo D,
Delfino M et al. Dermatoscopic pitfalls in differentiating pigment-
ed Spitz naevi from cutaneous melanomas. Br J Dermatol 1999;141:
788-93.
41. Zalaudek I, Argenziano G, Ferrara G, Soyer HP, Corona R, Sera F et
al. Clinically equivocal melanocytic skin lesions with features of
regression: a dermoscopic-pathological study. Br J Dermatol
2004;150:64-71.
42. Malvehy J, Puig S. Follow-up of melanocytic skin lesions with digi-
tal total-body photography and digital dermoscopy: a two-step method.
Clin Dermatol 2002;20:297-304.
43. Robinson JK, Nickoloff BJ. Digital epiluminescence microscopy
monitoring of high-risk patients. Arch Dermatol 2004;140: 49-56.
44. Kittler H, Seltenheim M, Dawid M, Pehamberger H, Wolff K, Binder
M. Frequency and characteristics of enlarging common melanocytic
nevi. Arch Dermatol 2000;136:316-20.
45. Rhodes AR. Common acquired nevomelanocytic nevi and the fourth
dimension. Arch Dermatol 2000;136:400-5.
46. Braun RP, Lemonnier E, Guillod J, Skaria A, Salomon D, Saurat JH.
Two types of pattern modification detected on the follow-up of benign
melanocytic skin lesions by digitized epiluminescence microscopy.
Melanoma Res 1998;8:431-7.
47. Kittler H, Pehamberger H, Wolff K, Binder M. Follow-up of
melanocytic skin lesions with digital epiluminescence microscopy: pat-
terns of modifications observed in early melanoma, atypical nevi,
and common nevi. J Am Acad Dermatol 2000;43:467-76.
48. Kittler H, Binder M. Risks and benefits of sequential imaging of
melanocytic skin lesions in patients with multiple atypical nevi. Arch
Dermatol 2001;137:1590-5.
49. Carli P, De Giorgi V, Chiarugi A, Nardini P, Weinstock MA, Crocetti
E et al. Addition of dermoscopy to conventional naked-eye examination
in melanoma screening: a randomized study. J Am Acad Dermatol
2004;50:683-9.
50. Hofmann-Wellenhof R, Wolf P, Smolle J, Reimann-Weber A, Soy-
er HP, Kerl H. Influence of UVB therapy on dermoscopic features
of acquired melanocytic nevi. J Am Acad Dermatol 1997;37:
559-63.
51. Hofmann-Wellenhof R, Soyer HP, Wolf Ich, Smolle J, Reischle S,
Rieger E et al. Ultraviolet radiation of melanocytic nevi: a dermo-
scopic study. Arch Dermatol 1998;134:845-50.
52. Stanganelli I, Rafanelli S, Bucchi L. Seasonal prevalence of digital epi-
luminescence microscopy patterns in acquired melanocytic nevi. J
Am Acad Dermatol 1996;34:460-4.
53. Stanganelli I, Bauer P, Bucchi L, Serafini M, Cristofolini P, Rafanel-
li S et al. Critical effects of intense sun exposure on the expression of
epiluminescence microscopy features of acquired melanocytic nevi.
Arch Dermatol 1997;133:979-82.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHIMENTI
54. Schiffner R, Schiffner-Rohe J, Vogt T, Landthaler M, Wlotzke U,
Cognetta AB et al. Improvement of early recognition of lentigo
maligna using dermatoscopy. J Am Acad Dermatol 2000;42:
25-32.
55. Ronger S, Touzet S, Ligeron C, Balme B, Viallard AM, Barrut D et al.
Dermoscopic examination of nail pigmentation. Arch Dermatol
2002;138:1327-33.
56. Tosti A, Argenziano G. Dermoscopy allows better management of
nail pigmentation. Arch Dermatol 2002;138:1369-70.
57. Baran R, Kechijian P. Hutchinson’s sign: a reappraisal. J Am Acad Der-
matol 1996;34:87-90.
58. De Giorgi V, Massi D, Carli P. Dermoscopy in the management of pig-
mented lesions of the oral mucosa. Oral Oncol 2003;39:534-5.
59. Massi D, De Giorgi V, Carli P, Santucci M. Diagnostic significance of
the blue hue in dermoscopy of melanocytic lesions: a dermoscopic-
pathologic study. Am J Dermatopathol 2001;23:463-9.
60. Ferrari A, Peris K, Piccolo D, Chimenti S. Dermoscopic features of
cutaneous local recurrent melanoma. J Am Acad Dermatol
2000;43:722-4.
61. Braun RP, Calza AM, Krischer J, Saurat JH. The use of digital der-
moscopy for the follow-up of congenital nevi: a pilot study. Pediatr Der-
matol 2001;18:277-81.
62. Seidenari S, Martella A, Pellacani G. Polarized light-surface
microscopy for description and classification of small and medium-
sized congenital melanocytic naevi. Acta Derm Venereol 2003;83:
271-6.
63. Marghoob AA, Kopf AW. Persistent nevus: an exception to the ABCD
rule of dermoscopy. J Am Acad Dermatol 1997;36:474-5.
64. Kaddu S, Soyer HP, Wolf IH, Rieger E, Kerl H. Reticular lentigo.
Hautarzt 1997;48:181-5.
65. Menzies SW. Dermoscopy of pigmented basal cell carcinoma. Clin
Dermatol 2002;20:268-9.
66. Peris K, Altobelli E, Ferrari A, Fargnoli MC, Piccolo D, Esposito M
et al. Interobserver agreement on dermoscopic features of pigment-
ed basal cell carcinoma. Dermatol Surg 2002;28:643-5.
67. Argenziano G, Rossiello L, Scalvenzi M, Staibano S, Ruocco E,
Cicale L et al. Melanoma simulating seborrheic keratosis: a major
dermoscopy pitfall. Arch Dermatol 2003;139:389-91.
68. Braun RP, Rabinovitz HS, Krischer J, Kreusch J, Oliviero M, Naldi
L et al. Dermoscopy of pigmented seborrheic keratosis: a morpho-
logical study. Arch Dermatol 2002;138:1556-60.
69. De Giorgi V, Massi D, Stante M, Carli P. False “melanocytic” para-
meters shown by pigmented seborrheic keratoses: a finding which is
not uncommon in dermoscopy. Dermatol Surg 2002;28:776-9.
70. Saida T, Oguchi S, Miyazaki A. Dermoscopy for acral pigmented
skin lesions. Clin Dermatol 2002;20:279-85.
71. Gasparini S, Giovene GL, Ferranti G. Trattato di Dermoscopia. Milan:
Spriger-Verlag Italia; 2003.
72. Worret WI. Halo eczema and nevus cell nevi (Meyerson nevi). Hau-
tarzt 1990;41:262-4.
73. Carli P, de Giorgi V, Salvini C, Mannone F, Chiarugi A. The gold
standard for photographing pigmented skin lesions for diagnostic
purposes: contact versus distant imaging. Skin Res Technol
2002;8:255-9.
74. Kittler H, Seltenheim M, Pehamberger H, Wolff K, Binder M. Diag-
nostic informativeness of compressed digital epiluminescence
microscopy images of pigmented skin lesions compared with pho-
tographs. Melanoma Res 1998;8:255-60.
75. Bittorf A, Fartasch M, Schuler G, Diepeng TL. Resolution requirements
for digital images in dermatology. J Am Acad Dermatol 1997;37:
195-8.
76. Abramovits W, Stevenson LC. Changing paradigms in dermatology:
new ways to examine the skin using noninvasive imaging methods. Clin
Dermatol 2003;21:353-8.
77. Oliveria SA, Sachs D, Belasco KT, Halpern AC. Adoption of new
technologies for early detection of melanoma in dermatologic prac-
tice. J Am Acad Dermatol 2003;49:955-9.
339
CHIMENTI
Linee guida in dermatoscopia
L e linee guida proposte in questo articolo riflettono lo
stato dell’arte in dermoscopia al momento in cui sono sta-
te stilate. È, quindi, probabile che i risultati di studi in cor-
so possano determinare alcune modifiche delle definizioni e
delle indicazioni che verranno di seguito riportate. L’appli-
cazione clinica di linee guida non deve prescindere dalla
prudenza e dalla coscienziosità, che vanno sempre esercita-
te nell’interpretare i criteri dermoscopici. Gli obiettivi fina-
li, che sono stabilire la diagnosi preoperatoria più accurata
ed effettuare la scelta terapeutica appropriata, devono esse-
re raggiunti considerando tutte le condizioni individuali del
paziente in esame. Le linee guida non possono mai sostitui-
re le responsabilità mediche individuali.
Significato della dermoscopia
La dermoscopia è una metodica diffusamente utilizzata nel-
la pratica clinica per la diagnosi precoce del melanoma 1.
Studi di valutazione sull’accuratezza diagnostica del solo
esame clinico hanno mostrato che il dermatologo è in grado
di individuare il melanoma nel 65-80% dei casi 2. Una recen-
te revisione sistematica della letteratura ha dimostrato che la
dermoscopia è in grado di incrementare la sensibilità dia-
gnostica del melanoma del 10-35% rispetto alla sola osser-
vazione clinica 3. È stato, inoltre, riportato che tale miglio-
ramento diagnostico può essere ottenuto solo se l’osservatore
100
90
80
70
60
%
50
40
30
20
10
0
Specificità Sensibilità
Clinica Analisi di pattern 7-point
ABCD Approccio integrato
Figura 1. — Valori di sensibilità e specificità per la diagnosi di melanoma
relativi a differenti sistemi di valutazione a confronto.
340 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GUIDELINES IN DERMOSCOPY
ha un buon livello di esperienza nell’utilizzo della metodica,
mentre l’accuratezza della diagnosi dermoscopica può risul-
tare anche peggiore rispetto alla sola diagnosi clinica, per i
non esperti 4, 5. Pertanto, un’adeguata preparazione è fon-
damentale ai fini di un’applicazione diagnostica realmente
efficace 6-8.
Da qui si può dedurre l’importanza di corsi di insegna-
mento formali sulla metodica, tenuti da personale qualifi-
cato ed esperto. Sembrerebbe auspicabile anche l’inseri-
mento di un esame di dermoscopia all’interno delle scuole di
specializzazione in Dermatologia e Venereologia.
Integrazione fra clinica e dermoscopia
L’integrazione della dermoscopia nel contesto della valu-
tazione clinica globale del paziente si è dimostrata capace di
migliorare ulteriormente la diagnosi preoperatoria del mela-
noma (Figura 1). Nel 1996, Menzies et al. hanno riscontra-
to che 9 (8%) di 107 melanomi inclusi nel loro studio erano
privi dei caratteri diagnostici dermatoscopici specifici e per
tale motivo erano definiti «featureless». Per tali lesioni, l’a-
sportazione venne effettuata solo sulla base di un cambia-
mento dell’aspetto clinico riferito dal paziente 9. Proprio in
considerazione dell’importanza del criterio clinico evoluti-
vo, è stato proposto un nuovo sistema ABCD che prevede l’in-
serimento di un criterio ulteriore, denominato E, relativo
alla storia evolutiva della lesione in esame 10. È stato, inol-
tre, osservato che la percentuale di melanomi diagnosticati
correttamente mediante l’osservazione dermoscopica effet-
tuata dal vivo (faccia a faccia con il paziente) risulterebbe
maggiore rispetto a quella ottenuta esaminando le immagi-
ni dermoscopiche, in diapositiva, dei medesimi casi di mela-
noma 11. Questo significa che esistono dei parametri clinici,
quali l’età del paziente, il fototipo, il numero e la tipologia
degli altri nevi, la sede della lesione, la storia evolutiva, ecc.,
che sono in grado di aumentare l’accuratezza della diagno-
si finale. In presenza di dati clinici sospetti, il dermatologo
sarà più stimolato all’attenta valutazione anche di parametri
dermoscopici sfumati o appena percettibili, che altrimenti
potrebbero sfuggire 11.
Impatto della dermoscopia sul management clinico
In uno studio recente è stato valutato il ruolo della der-
moscopia come ausilio nella gestione clinica delle lesioni
pigmentate cutanee e, in particolare, quanto essa permetta di
ridurre il numero di escissioni chirurgiche delle lesioni beni-
gne 12. A questo scopo, una serie di lesioni, tutte asportate e
istologicamente confermate, sono state retrospettivamente
Agosto 2005
GUIDELINES IN DERMOSCOPY CHIMENTI

~ 65% Lesioni = Diagnosi definitiva Follow-up in

ESAME CLINICO di BENIGNITÀ selezionati

Paziente
(circa 1 ~ 35% lesioni
DERMOSCOPIA
melanoma ogni clinicamente
(Analisi pattern, ABCD, 7-Point)
100 pazienti EQUIVOCHE
osservati)

Sospetto melanoma -40% delle lesioni equivoche:

Approccio integrato: diagnosi definitiva di benignità
almeno 1 diagnosi (clinica o (escissione evitata)
dermoscopica) del melanoma

Follow-up in
ASPORTAZIONE
casi selezionati

Figura 2. — Sistema di gestione integrata delle lesioni pigmentate, elaborato sulla base dell’attività media di un Centro di Riferimento per lo scree-
ning delle lesioni pigmentate cutanee e la diagnosi dermoscopica precoce del melanoma.

valutate dal punto di vista sia clinico che dermoscopico. I
risultati hanno mostrato che, pur essendo la sensibilità dia-
gnostica del melanoma del 90%, tutte le lesioni maligne
(melanomi e carcinomi basocellulari) venivano comunque
giudicate tali da essere asportate. Inoltre, il 40% delle lesio-
ni clinicamente classificate come sospette (falsi positivi)
venivano invece diagnosticate come lesioni benigne all’esame
dermoscopico e, quindi, da non asportare. Ne consegue che,
quando la dermoscopia viene utilizzata per stabilire se la
lesione debba essere asportata o meno, questa metodica con-
sente di migliorare significativamente in termini di specifi-
cità la gestione clinica delle lesioni pigmentate 12. Un siste-
ma di gestione di lesioni pigmentate, elaborato sulla base
dell’attività media di un centro di riferimento per lo screening
delle lesioni pigmentate cutanee e la diagnosi precoce del
melanoma e basato sull’approccio diagnostico integrato, è
schematizzato nella Figura 2.
Procedura in due fasi per la diagnosi dermoscopica
Il metodo diagnostico per l’esame dermoscopico delle
lesioni pigmentate è stato recentemente standardizzato e
proposto come riferimento nel corso del Consensus Net
Meeting on Dermoscopy (CNMD). Il CNMD, tenutosi nel
2000 via internet tra 40 esperti di 14 diversi Paesi, aveva
come obiettivo stabilire alcune linee guida fondamentali in
dermoscopia: definizione, standardizzazione e semplifica-
zione della terminologia, e verifica della riproducibilità e
validità dei diversi criteri e degli algoritmi diagnostici 13. Il
metodo diagnostico dermoscopico proposto consta di una
procedura in 2 fasi. La prima fase consiste nel differenziare
la natura melanocitica o non melanocitica della lesione pig-
mentata in questione. A tale proposito, dovranno, pertanto,
essere identificati i criteri che consentono di definire una
lesione come melanocitica, quali: reticolo pigmentato, glo-
buli marroni, strie, pigmentazione blu omogenea, pattern
parallelo. In assenza di tali aspetti verrà esaminata la presenza
di criteri per le lesioni non melanocitiche e, in particolare, i
criteri per la diagnosi di cheratosi seborroica (pseudocisti
cornee, sbocchi simil-comedonici, strutture a impronta digi-
tale, aree con aspetto cerebriforme con giri e solchi), carci-
noma basocellulare (vasi arboriformi, aree a foglia d’ace-
ro, grandi aree ovoidali grigio-blu, multipli globuli grigio-blu,
aree a ruota di carro e ulcerazione) o di lesioni vascolari
(lacune rosso-blu, aree omogenee da rosso-bluastre a rosso
nerastre) 14. Nel caso in cui anche questi criteri siano assen-
ti, il pattern viene definito aspecifico e deve far, comunque,
sospettare una diagnosi di lesione melanocitica 13.
La seconda fase diagnostica prevede la differenziazione tra
le lesioni melanocitarie benigne e il melanoma mediante
l’applicazione di differenti algoritmi diagnostici: l’analisi
di pattern modificata 14, 15, l’ABCD di Stolz 16, il metodo di
Menzies 9 e il seven-point check-list 17. In base ai risultati del
CNMD è stato osservato che tutti questi metodi assicurano
una sensibilità elevata nella diagnosi di melanoma, ma che

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 341

CHIMENTI
TABELLA I. — Criteri dermoscopici melanoma-specifici, che hanno mostrato un valore predittivo significativamente elevato per la diagno-
si di melanoma, e loro correlati istopatologici 13-15, 19-28.
Criteri dermoscopici Descrizione dermoscopica
melanoma-specifici
Pattern polimorfo (o mul- Combinazione di 3 o più strutture dermosco-
ticomponente) piche distinte in una stessa lesione
Reticolo pigmentato atipico Caratterizzato da maglie irregolari e trama
ispessita, brusca interruzione alla periferia,
Strie irregolari (o pseudo- e di colore marrone-nerastro
podi) Strutture lineari di spessore variabile, irrego-
larmente distribuite, non chiaramente asso-
Strutture di regressione ciate alle maglie del reticolo
Associazione di aree bianche simil-cicatri-
ziali, e di aree grigio-blu tipo «peppe-
Punti deboli irregolari ring»
Strutture rotondeggianti di forma e dimen-
sione irregolari e disomogeneamente distri-
buiti nel contesto della lesione
Pigmetazione irregolare
Aree pigmentate nere, marroni o grigie di for- Intensa pigmentazione melaninica distribuita
ma e/o distribuzione irregolare
Velo blu-biancastro
Pigmentazione diffusa e confluente, di colo- Epidermide acantosica con ipergranulosi foca-
re variabile dal grigio-blu al blu-bianca-
stro, associata a reticolo pigmentato
Asimmetria strutturale Asimmetria nella forma della lesione (ma
anche di colori e strutture dermoscopiche),
calcolata sia con il metodo dell’ABCD che
con quello di Menzies
l’analisi di pattern mostra una più elevata specificità rispet-
to agli algoritmi alternativi semplificati, anche se richiede una
maggiore esperienza da parte dell’osservatore 13.
Recentemente, è stato proposto il three-point checklist 18
per consentire anche ai dermoscopisti meno esperti di dia-
gnosticare il maggior numero di melanomi (anche se con
una diminuzione della specificità). Si tratta di un metodo
semplificato basato sulla valutazione di soli 3 criteri der-
moscopici: l’asimmetria della lesione, la presenza di reti-
colo pigmentato atipico e di strutture bianco-blu (definite
come la presenza di qualsiasi struttura di colore blu e/o bian-
co). Il three-point checklist potrebbe rappresentare un effi-
cace sistema dermoscopico di screening delle lesioni pig-
mentate, anche per dermoscopisti non esperti 18.
Criteri dermoscopici melanoma-specifici
Negli ultimi anni numerosi studi hanno dimostrato la vali-
dità e la riproducibilità di criteri dermoscopici che sono più
frequentemente osservati nel melanoma e, per questo, sono
definiti melanoma-specifici 13-15, 19-28 (Tabella I.) Il pattern
342 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GUIDELINES IN DERMOSCOPY
Associazione con la
Correlati istopatologici diagnosi di melanoma
(Odds ratio) 13
4.3
Rate ridges irregolare e ispessita, un suo scom- 9,3
paginamento correla con la progressione
del melanoma
Teche di melanociti confluenti alla giunzione 5.8
dermo-epidermica
Associazione di fibrosi e melanofagi a livel- 5.4
lo del derma papillare ispessito
Accumuli di melanina nello strato corneo (o 4.8
anche segno di invasione pagetoide del-
l’epidermide) i primi, e nidi di melanoci-
ti situati alla giunzione o nel derma papil-
lare i secondi
4.1
a tutti i livelli dell’epidermide o nel derma
superficiale
2.9
le che sovrasta teche di melanociti forte-
mente pigmentati nel derma
13,7 (asimmetria in 2 assi
secondo l’ABCD di Stolz)
43,8 (asimmetria secondo
il metodo di Menzies)
dermoscopico globale che è risultato avere un valore pre-
dittivo maggiore riguardo al melanoma è stato quello poli-
morfo (o multicomponente), definito come combinazione
di 3 o più strutture dermoscopiche distinte in una stessa
lesione 13. Al contrario, i pattern globulare, ad acciottolato,
omogeneo e «a stella che esplode», sono risultati maggior-
mente predittivi per la diagnosi di lesioni melanocitiche
benigne 13. Per quanto riguarda i criteri dermoscopici loca-
li, il reticolo pigmentato atipico, le strie irregolari e le strut-
ture di regressione hanno mostrato il valore predittivo più
elevato nei confronti del melanoma, seguite da punti e globuli
irregolari, pigmentazione irregolare e velo blu-biancastro 13.
Invece, il reticolo pigmentato tipico, i punti e i globuli rego-
lari, le strie regolari e la pigmentazione regolare, sono risul-
tati maggiormente associati con lesioni melanocitiche beni-
gne 13. Anche l’asimmetria strutturale della lesione (calco-
lata sia con il metodo dell’ABCD che con quello di Menzies)
è risultata statisticamente predittiva di malignità 13. Questi cri-
teri melanoma specifici, espressione di alterazioni istopa-
tologiche ben definite (Tabella I), vanno sempre accurata-
mente ricercati nel contesto di una lesione pigmentata e la loro
osservazione giustifica il più delle volte l’escissione chirur-
gica e l’esame istologico.
Agosto 2005
GUIDELINES IN DERMOSCOPY
Il problema dei falsi negativi e delle lesioni equivoche
In dermoscopia, i falsi negativi comprendono i melanomi
che non vengono diagnosticati e, quindi, non asportati chi-
rurgicamente, e includono essenzialmente i melanomi che
mimano lesioni melanocitiche benigne quali il nevo di Clark
e il nevo di Spitz/Reed, oppure i melanomi «featureless»,
ossia lesioni che non mostrano criteri dermoscopici specifi-
ci o sono apigmentate 9, 14, 29. Dal punto di vista pratico, nel
caso di lesioni clinicamente sospette ma dermoscopicamente
apparentemente benigne, è importante esaminare accurata-
mente la lesione al fine di individuare possibili aspetti der-
moscopici atipici. La diagnosi di melanoma deve, comunque,
essere sempre sospettata quando la lesione mostra un pattern
aspecifico. In questi casi, è, comunque, fondamentale inte-
grare i dati dermoscopici con la storia clinica della lesione 10.
In particolare, per le lesioni poco pigmentate, la ricerca di un
pattern vascolare atipico (essenzialmente aree/globuli rosso-
lattescenti, vasi lineari-irregolari o la combinazione vasi
puntiformi e lineari-irregolari) 29, 30 può suggerire la dia-
gnosi di melanoma se associato ai criteri quali pigmenta-
zione irregolare, globuli/punti irregolari, strutture di regres-
sione e velo bianco-bluastro 30. Per le lesioni «rosa» o nul-
la pigmentate, i pattern vascolari possono da soli non esse-
re sufficienti a porre diagnosi di melanoma e vanno integra-
ti con le informazioni cliniche quali età, sesso, familiarità per
melanoma, numero delle lesioni, sede, epoca d’insorgenza
e modificazioni nel tempo della lesione. L’approccio integrato
(indagine dermoscopica-informazioni cliniche) può con-
sentire di diagnosticare un melanoma amelanotico in uno
stadio più precoce 30.
Vi è, infine, una percentuale di lesioni melanocitiche per
le quali è difficile o impossibile stabilire una diagnosi di
benignità o malignità, dal punto di vista sia clinico sia der-
moscopico, e, in alcuni casi, anche istopatologico 31. Appar-
tengono a questo gruppo le lesioni melanocitiche in cui la dia-
gnosi differenziale tra nevo di Clark di tipo giunzionale e
melanoma in situ, o tra nevo di Spitz/Reed e melanoma spit-
zoide è particolarmente difficile. In questi casi, le lesioni
dovrebbero essere asportate chirurgicamente o sottoposte a
uno stretto monitoraggio dermoscopico (1-3 mesi) che per-
mette di apprezzare un eventuale accrescimento asimmetri-
co della lesione stessa o modificazioni delle strutture der-
moscopiche 32, 33. Recentemente è stata proposta una nuova
classificazione dermoscopica dei nevi di Clark, utile nella
selezione delle lesioni da sottoporre a escissione chirurgica 1.
È stato osservato, infatti, che un’iperpigmentazione eccen-
trica (periferica) e la coesistenza di 3 strutture nell’ambito del-
la stessa lesione (reticolare, globulare e omogenea) sono
caratteristiche significativamente più frequenti nel melano-
ma; per tale motivo, queste lesioni dovrebbero essere aspor-
tate 35. Per le lesioni con aspetto spitzoide, è stato descritto
un modello di possibile evoluzione naturale nel tempo: da un
pattern globulare a un pattern a stella che esplode («star-
burst») 36, per poi andare incontro a una graduale scompar-
sa delle strie periferiche e a una pigmentazione centrale più
diffusa e omogenea 37, 38. In una percentuale di casi, il riscon-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHIMENTI
tro in una lesione spitzoide di un reticolo nero superficiale
(che, istopatologicamente, corrisponde a focali aree di para-
cheratosi pigmentata, che producono un aspetto reticolato e
nero sul piano orizzontale) può essere di ausilio nel porre dia-
gnosi di benignità 39. Per le lesioni con aspetto spitzoide,
l’età del paziente risulta, comunque, di fondamentale impor-
tanza: nei pazienti adulti esse vanno senz’altro asportate;
nei pazienti pediatrici, se presentano pattern dermoscopici
tipici, possono essere monitorate nel tempo 40.
Infine le lesioni melanocitiche che dermoscopicamente
possono essere di difficile interpretazione sono quelle che pre-
sentano strutture di regressione, quali aree bianche simil-
cicatriziali e/o aree blu tipo peppering. Un recente studio di
correlazione dermoscopico-patologica su lesioni melanoci-
tiche clinicamente equivoche con caratteristiche di regressione
ha evidenziato che la maggioranza dei nevi con regressione
mostra aree blu che coinvolgono <50% della lesione e han-
no una distribuzione prevalentemente centrale, mentre le
lesioni istologicamente equivoche mostrano una combina-
zione di aree bianche e aree blu, irregolarmente distribuite,
che coinvolgono >50% della lesione 41. In base a tali risul-
tati, è stato proposto un algoritmo per la gestione clinica
delle lesioni che dermoscopicamente presentano regressio-
ne: le lesioni che mostrano un basso grado di strutture di
regressione (<10%) potrebbero essere sottoposte a moni-
toraggio dermoscopico, al contrario, l’escissione andrebbe
sempre effettuata per le lesioni che mostrano un alto grado
di regressione (>50%) o che presentano un grado modera-
to di strutture di regressione (compreso tra 10% e 50%) ma
con la presenza contemporanea di aree bianche e di aree
blu 41.
Follow-up dermoscopico e modificazioni
nel tempo delle lesioni pigmentate
La necessità di sottoporre un paziente a esami clinici
periodici deriva dalla duplice esigenza di monitorare un sog-
getto che presenta significativi fattori di rischio per lo sviluppo
di un melanoma (e.g. storia personale o familiare di mela-
noma, elevato numero totale di nevi ecc.) e di osservare l’e-
voluzione nel tempo di singole lesioni melanocitiche mode-
ratamente atipiche, ma non tali da sospettare un melanoma.
Inoltre, il monitoraggio dermoscopico è particolarmente uti-
le nei soggetti che presentano un elevato numero di nevi,
molti dei quali clinicamente atipici, la cui asportazione con-
temporanea sarebbe praticamente impossibile 1, 42, 43.
In uno studio recente sono state descritte le caratteristiche
dermoscopiche dei nevi in accrescimento: su una casistica di
1 612 nevi melanocitici comuni, il 5% dei nevi ha mostrato
un aumento delle dimensioni in un periodo di 12 mesi. Nel
50% di queste lesioni era possibile evidenziare un anello di
globuli marroni simmetricamente distribuito alla periferia
della lesione, espressione dell’attività proliferativa delle cel-
lule neviche 44. Sebbene questo fenomeno fosse riscontrabile
più comunemente in soggetti di età inferiore ai 20 anni, l’ac-
crescimento simmetrico di una lesione melanocitaria (in
343
CHIMENTI
assenza di altri segni di atipia) non è di per sé indicativo di
malignità, come dimostrato dall’esame istologico condotto
su queste lesioni 44. Lesioni in accrescimento in soggetti
adulti, soprattutto se in presenza dell’anello globulare peri-
ferico, devono tuttavia essere attentamente seguite attraver-
so un monitoraggio digitale stretto (3 mesi) o asportate 45.
Altri studi hanno dimostrato l’efficacia del monitoraggio
digitale nell’individuare i pattern di modificazione delle
lesioni 46. I nevi atipici hanno mostrato essenzialmente un
accrescimento focale in assenza di importanti modifiche
strutturali, mentre nei melanomi è stato rilevato un accre-
scimento focale, associato a modifiche della forma e com-
parsa di caratteri dermoscopici quali punti neri irregolari,
rete pigmentaria irregolare, strutture di regressione, strie
irregolari, velo blu-biancastro 47. Quando a un esame der-
moscopico di follow-up si riscontrano modificazioni di cri-
teri dermoscopici, quali espansione o riduzione del reticolo
pigmentato, distribuzione o numero di punti neri, e/o aree di
ipopigmentazione o di regressione, è sempre consigliabile l’a-
sportazione chirurgica. Analogamente, la comparsa di ulte-
riori caratteri dermoscopici atipici, quali reticolo pigmen-
tato atipico, punti neri irregolari, strutture di regressione,
strie irregolari, velo blu-biancastro e pattern vascolare atipico,
sono indicativi di lesione sospetta che deve essere escissa
chirurgicamente.
Inoltre Menzies et al. hanno dimostrato che, effettuando
un follow-up a breve termine (in media 3 mesi) su 318 lesio-
ni solo moderatamente atipiche, è stato possibile identifica-
re ben 7 melanomi in stadio iniziale dermoscopicamente
«featureless», cioè che non mostravano criteri dermoscopi-
ci atipici ma che la sola modificazione in un intervallo stret-
to di tempo ha permesso di diagnosticare 32.
Quando si effettua un follow-up di lesioni atipiche, vi
sono, comunque, alcuni rischi da non sottovalutare. A tale pro-
posito è stato recentemente dimostrato che il ricorso indi-
scriminato al monitoraggio dermoscopico non è raccoman-
dabile, in quanto la sua efficacia clinica dipende dall’espe-
rienza dell’osservatore e dalla compliance del paziente e,
quindi, la sua adesione a un programma di monitoraggio nel
tempo 48. La scelta delle lesioni e dei pazienti da sottoporre
a follow-up digitale va, dunque, sempre valutata attenta-
mente per non rischiare di perdere un melanoma 33. Recen-
temente Carli et al. Hanno, infatti, dimostrato, in uno studio
randomizzato effettuato su 938 pazienti, che l’archiviazione
delle immagini dermoscopiche di lesioni equivoche si asso-
cia, da un lato, a una diminuzione di casi sottoposti a escis-
sione chirurgica, dall’altro, a un rischio non trascurabile di
melanomi iniziali non asportati 49.
Infine, nel contesto delle variazioni dermoscopiche osser-
vabili nelle lesioni melanocitiche, bisogna tener presenti
quelle dovute a esposizioni alle radiazioni ultraviolette, che
consistono in una maggiore pigmentazione e irregolarità
nella distribuzione del pigmento, un incremento delle dimen-
sioni dei globuli marroni, una diminuzione delle aree ipo-
pigmentate e una minore visibilità del reticolo pigmenta-
rio 50-53. Questi cambiamenti morfologici sono, tuttavia,
transitori e legati presumibilmente a un’attivazione reversi-
344 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GUIDELINES IN DERMOSCOPY
bile delle cellule neviche 50-53. Risulta necessario, quindi,
esaminare nuovamente queste lesioni 4-6 settimane dopo
l’esposizione solare, a causa della loro difficile differenzia-
zione dal melanoma nel periodo immediatamente successi-
vo alle esposizioni solari stesse 50-53.
In generale, le lesioni che possono essere sottoposte a un
monitoraggio digitale nel tempo sono quelle solo lievemen-
te atipiche, piane e non rilevate e non devono avere una sto-
ria di variazioni morfologiche né presentare criteri melano-
ma-specifici.
Il follow-up non dovrebbe mai essere eseguito nelle lesio-
ni nodulari che presentano caratteri di atipia, data l’impos-
sibilità di escludere con certezza una diagnosi di melanoma
nodulare. In questi casi, infatti, è sempre consigliata l’a-
sportazione chirurgica.
Aspetti morfologici salienti e indicazioni
per la gestione clinica delle lesioni pigmentate
di più difficile interpretazione
Precedentemente sono stati esaminati in dettaglio i crite-
ri dermoscopici melanoma specifici, che, quando osservati
in una lesione, devono far procedere a un’asportazione chi-
rurgica. In Tabella II 13, 14, 32, 34-36, 38, 40, 41, 54-70 sono ripor-
tati gli aspetti dermoscopici delle lesioni pigmentate che più
comunemente possono costituire dei falsi positivi o dei fal-
si negativi, nonché alcuni suggerimenti per un più accurato
inquadramento diagnostico e per una migliore gestione del-
le lesioni che generano problemi di diagnosi differenziale.
Per i nevi di Clark, come già accennato, è importante l’in-
dividuazione delle lesioni che presentano un’iperpigmenta-
zione eccentrica (periferica) e delle lesioni in cui vi è la coe-
sistenza di strutture reticolari, globulari e omogenee. Questi
nevi dovrebbero, infatti, essere asportati chirurgicamente,
in quanto le stesse caratteristiche possono riscontrarsi anche
nel melanoma 34, 35. Quando un paziente presenta lesioni
multiple con aspetti atipici si può ricorrere all’escissione di
quella/e maggiormente sospetta/e ed effettuare un follow-up
dermoscopico stretto (3 mesi) per le altre 32, 34. Un discorso
a parte meritano le lesioni con regressione di cui viene pro-
posto un modello di gestione in Tabella II 41.
I nevi di Spitz/Reed (nevi a cellule epitelioidi e/o fusate,
compresa la variante pigmentata, prima considerata distin-
ta, detta nevo di Reed), come precedentemente accennato,
possono presentarsi dermoscopicamente con una varietà di
pattern: a stella che esplode, globulare, reticolare, omogeneo,
ipopigmentato e atipico 37. Queste lesioni pongono spesso
problemi di diagnosi differenziale con un melanoma spit-
zoide, dal punto di vista sia clinico-dermoscopico sia isto-
logico. Per questo si consiglia l’escissione di qualsiasi lesio-
ne che nell’adulto mostri un aspetto spitzoide 40. Nei pazien-
ti pediatrici, un nevo di Spitz/Reed che presenta aspetti tipi-
ci può essere sottoposto a uno stretto follow up, in modo da
poter monitorare i diversi pattern di modificazione e evita-
re inutili asportazioni di lesioni benigne 36, 38.
Agosto 2005
GUIDELINES IN DERMOSCOPY
TABELLA II. — Aspetti dermoscopici salienti delle lesioni pigmentate che più comunemente possono costituire dei falsi positivi o dei falsi nega-
tivi, e suggerimenti per una migliore gestione clinica.
Lesioni pigmentate
Falsi positivi
1) Nevo di Clark con iperpigmentazio-
ne eccentrica
2) Nevo di Clark con regressione
3) Nevo di Spitz
Falsi negativi
1) Melanoma che simula una delle
seguenti lesioni pigmentate:
— Nevo di Clark
— Nevo di Spitz
— Nevo dermico
— Lesioni pigmentate del volto
— Lesioni in sede acrale
— Lesioni in sede ungueale
— Melanosi delle mucose
— Nevo blu
— Nevo congenito
— Nevo ricorrente
— Nevo irritato
— Lentigo reticolare
— Carcinoma basocellulare
— Cheratosi seborroica
— Lesioni vascolari
2) Melanoma con pattern aspecifico
3) Melanoma amelanotico
4) Metastasi di melanoma
Vol. 140 - N. 4
CHIMENTI
Aspetti morfologici salienti e indicazioni
Escissione se unico elemento; follow-up stretto (3 mesi) se più lesioni con aspetto simile nello stesso
paziente 32, 34, 35
Escissione se regressione >50% della superficie lesionale; follow-up se regressione <10% lesione; escis-
sione se regressione 10-50% ma presenza contemporanea di aree bianche e di aree blu 41
Escissione nell’adulto 40, nel bambino se aspetti tipici, stretto follow-up dermoscopico 36, 38
Follow-up a 3 mesi se in paziente con multiple lesioni atipiche 32, 34, 35; escissione se unico elemento
anche se lievemente atipico. Escissione se contemporanea presenza di strutture reticolari, globulari e omo-
genee 35
Escissione di qualsiasi lesione spitzoide nell’adulto 40 nel bambino stretto follow-up di lesioni con aspetti
tipici 36, 38
Criteri differenziali da valutare:
— pattern ad acciottolato, vasi a virgola e peli sono in favore di un nevo dermico 14
— asimmetria, velo blu, punti/globuli irregolari e pattern vascolare atipico sono in favore di un melanoma
anche in presenza di un pattern ad acciottolato 13. Mai follow-up in lesioni nodulari sospette
Criteri diagnostici da valutare 54:
— follicoli pigmentati asimmetrici
— strutture anulari-granulari
— strutture romboidali.
Escissione chirurgica o biopsia incisionale in aree sospette (grigio-blu) se lesioni estese
Escissione chirurgica delle lesioni che mostrano 70:
— pattern a creste parallele
— pattern atipici o multicomponenti
Biopsia incisionale se si rinviene:55-57
— pigmentazione marrone di fondo e linee longitudinali da marroni a nere, irregolari
— micro-segno di Hutchinson
Follow-up negli altri casi meno dubbi 56
Biopsia incisionale se lesioni con distribuzione irregolare del pigmento e colore variegato 58
Asportazione delle lesioni nodulari
Escissione se storia clinica e aspetti dermoscopici dubbi 59, mai follow-up in lesioni nodulari sospette
Escludere una metastasi da melanoma 60
Follow-up attraverso immagini dermoscopiche di:61, 62
— intera lesione se possibile
— zone rappresentative del pattern architetturale
— bordi della lesione
— zone di particolare interesse 61, 62
Escissione chirurgica delle lesioni sospette o biopsia incisionale
Aspetti dermoscopici spesso atipici, fondamentale la storia clinica nel decidere se asportare o monitorare
nel tempo 63
Follow-up stretto (anche 1-2 settimane). Escissione chirurgica nei casi che non si risolvono
Reticolo pigmentato, prominente e molto scuro (anche nerastro) a maglie irregolari 64. Escissione chirur-
gica nelle lesioni sospette
Ricercare la presenza di: vasi arboriformi, aree a foglia d’acero, grandi aree ovoidali grigio-blu, multipli glo-
buli grigio-blu, aree a ruota di carro ed ulcerazione 65, 66 Asportazione chirurgica in ogni caso
— valutare il numero di pseudocisti cornee e sbocchi simil-comedonici (numerosi nella cheratosi, pochi nel
melanoma) 67
— escissione delle lesioni sospette che mostrano caratteristiche quali falso reticolo pigmentario e strutture
pseudo-globulari 67-69
— Emangioma: differenziare le lacune blu-rosse (ben circoscritte nell’angioma) dalle milky red areas (non
ben circoscritte nel melanoma) 14
— Granuloma piogenico: escissione ed esame istologico nell’adulto
Escissione in assenza di criteri definiti per la diagnosi 14
Criteri diagnostici da valutare nelle lesioni rosa, che devono far propendere per un’escissione 29, 30
— lesione nodulare e/o ulcerata
— residui di pigmento, specialmente se di colore blu-grigio
— milky red areas (aree rosso lattescenti)
— pattern vascolare atipico (vasi punteggiati e/o lineari-irregolari)
— pigmentazione omogenea bluastra
— pigmentazione diffusa non omogenea marrone-bluastra
— pattern a globuli rosso-bluastri 60
GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 345
CHIMENTI
I nevi dermici (Unna e Miescher) dermoscopicamente
sono caratterizzati frequentemente da una pigmentazione
omogenea (che a livello del volto assume l’aspetto di pseu-
doreticolo pigmentato), da un pattern ad acciottolato e da
vasi disposti a virgola 14. Spesso presentano i caratteri der-
moscopici tipici delle lesioni esofitiche: strutture papillari eso-
fitiche e cripte irregolari 14. Tuttavia, il riscontro di asim-
metria, velo blu-biancastro, punti e globuli irregolari e pat-
tern vascolare atipico, deve far sospettare un melanoma 13.
Non è mai consigliabile ricorrere a un follow-up in lesioni
nodulari quando si ha anche solo un minimo dubbio dia-
gnostico.
Le lesioni pigmentate del volto si presentano all’esame der-
moscopico con il caratteristico pseudoreticolo pigmentato,
dovuto alla distribuzione del pigmento attorno agli osti fol-
licolari 14. I criteri che permettono di sospettare una lentigo
maligna sono la presenza di strutture anulari-granulari (di
colorito grigio-bluastro) e la pigmentazione asimmetrica
degli sbocchi follicolari 54. Le strutture romboidali e le aree
omogenee di invasione dei follicoli indicano, invece, una
fase più avanzata di progressione del melanoma (lentigo
maligna melanoma) 54. Quando si osservano questi caratte-
ri, è bene procedere all’asportazione chirurgica; in caso di
lesioni molto estese, si può procedere a una biopsia incisio-
nale, preferibilmente a livello delle zone che mostrano una
pigmentazione bluastra.
I nevi in sede acrale (a livello palmo-plantare) mostrano
un caratteristico pattern parallelo dovuto alla disposizione del
pigmento lungo i solchi. In particolare, nei nevi acrali, si
osservano comunemente il pattern a solchi paralleli, a rete di
metallo e fibrillare, mentre il pattern a creste parallele è alta-
mente suggestivo di un melanoma acrale-lentigginoso 70.
Le lesioni che mostrano dei pattern atipici o multicomponente
vanno comunque sottoposte a escissione chirurgica.
Tra le lesioni a livello ungueale la diagnosi differenziale
si pone essenzialmente con l’emorragia subungueale (storia
clinica di trauma e aspetto dermoscopico caratterizzato dal-
la presenza di aree tondeggianti ben circoscritte di colore
nero-rossastro e da «blood spots» o punti di colorito rosso-
nerastro), con i nevi melanocitici (linee longitudinali rego-
lari per spessore e parallelismo, su di un fondo marrone
omogeneo) e con la melanonichia indotta da farmaci (pig-
mentazione omogenea di colorito grigiastro con linee lon-
gitudinali regolari) 55. Appare utile, in questi casi, il follow-
up dermoscopico a conferma della diagnosi 56. Il melanoma
subungueale, invece, si presenta dermoscopicamente con
pigmentazione marrone di fondo e linee longitudinali da
marroni a nere, irregolari per spessore, parallelismo e colo-
razione. È importante anche ricercare il micro-segno di Hut-
chinson (pigmentazione a livello della cuticola e della cute
periungueale, visibile solo all’esame dermoscopico) che,
anche se raro, deve far sospettare un melanoma 55, 57. Nei casi
sospetti si deve ricorrere a una biopsia incisionale che coin-
volga anche la matrice ungueale.
Le melanosi labiali e genitali sono lesioni pigmentate beni-
gne che si presentano dermoscopicamente con una pigmen-
tazione diffusa di fondo con rinforzi del pigmento di tipo
346 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GUIDELINES IN DERMOSCOPY
granulare, globulare (con globuli spesso allineati) o lineare-
curvilineo, sovente parallelo, di colore marrone chiaro, bru-
no o grigiastro 71. Appare, comunque, utile un follow-up di
queste lesioni e di quelle che mostrano una pigmentazione lie-
vemente irregolare. Nel caso di lesioni sospette, caratterizzate
da colore variegato e distribuzione irregolare del pigmento,
la dermoscopia può permettere di identificare la zona più ati-
pica dove praticare una biopsia incisionale 58.
I nevi blu sono facilmente diagnosticabili quando mostra-
no la tipica pigmentazione omogenea bluastra, si riscontra-
no nelle sedi caratteristiche e senza storia di modificazioni.
Tuttavia qualche dubbio di diagnosi differenziale con il mela-
noma può insorgere nelle lesioni che presentano aree bian-
co-giallastre, indice di una fibrosi associata. Data l’importanza
delle strutture blu in dermoscopia, è sempre richiesta un’at-
tenta valutazione di tali aspetti, mentre l’asportazione è con-
sigliabile nei casi in cui la storia clinica di queste lesioni sia
dubbia o la dermoscopia anche solo lievemente sospetta 59.
Inoltre, bisogna tenere presente che le metastasi da mela-
noma possono simulare un nevo blu 60.
La dermoscopia risulta molto utile nello studio e nel fol-
low-up dei nevi congeniti. La presentazione dermoscopica di
tali lesioni è eterogenea: il pattern più comune è quello ad
acciottolato, ma, frequentemente, è possibile osservare anche
un pattern multicomponente, per la presenza di colori mul-
tipli, punti e globuli marroni, zone di reticolo pigmentato, aree
omogenee ipopigmentate e aree omogenee bluastre 61. Sono
stati recentemente descritti anche peculiari aspetti dermo-
scopici come il reticolo a bersaglio, i globuli a bersaglio e i
vasi a bersaglio 62. In ogni caso, un monitoraggio attento
delle immagini cliniche digitali dell’intera lesione, integra-
te dalle immagini dermoscopiche rappresentative del pat-
tern architetturale, dei bordi della lesione e di zone di parti-
colare interesse, può permettere di seguire nel tempo queste
lesioni congenite e di individuare l’eventuale comparsa di
caratteri atipici 61. In questi casi, se la lesione è di piccole
dimensioni è consigliabile l’asportazione chirurgica men-
tre in caso di nevi grandi o giganti è possibile effettuare una
biopsia incisionale nelle aree dubbie.
I nevi ricorrenti (o persistenti), spesso mostrano caratteri-
stiche dermoscopiche talmente bizzarre e atipiche da far sospet-
tare un melanoma, quali strie irregolari, punti e globuli irregolari
in prossimità o nel contesto dell’area cicatriziale 63. Se i nevi
ricorrenti compaiono dopo un’escissione incompleta di un
nevo istologicamente non atipico, possono essere monitorati
nel tempo. Quando non vi è certezza sulla precedente diagnosi
clinico-istopatologica, la lesione deve essere asportata chi-
rurgicamente.
Un altro utile campo di applicazione del monitoraggio
digitale riguarda i nevi irritati (da traumatismi, infezioni) o
nevi di Meyerson 72 in cui un follow-up molto stretto (anche
di 1-2 settimane) dopo trattamento locale può permettere di
dirimere il dubbio diagnostico.
Un’altra lesione pigmentata che può mimare un melano-
ma (in situ) è la lentigo reticolare (o «ink spot lentigo»), che
si riscontra frequentemente su cute intensamente foto-dan-
neggiata. Dermoscopicamente la lesione è caratterizzata da
Agosto 2005
GUIDELINES IN DERMOSCOPY
un reticolo pigmentato, prominente e di colore marrone scu-
ro-nerastro, a maglie irregolari e sfrangiate, che tuttavia è
uniformemente distribuito su tutta la lesione 64. Nelle lesio-
ni dubbie è bene ricorrere all’escissione chirurgica.
Il carcinoma basocellulare, specie se pigmentato, può por-
re problemi di diagnosi differenziale con il melanoma. Gli
aspetti dermoscopici tipici del BCC sono i vasi arborifor-
mi, le aree a foglia d’acero, le grandi aree ovoidali grigio-blu,
i globuli multipli grigio-blu, le aree a ruota di carro e l’ul-
cerazione, in assenza di reticolo pigmentato 65, 66.
La cheratosi seborroica, specie nella sua variante acantoti-
ca e pigmentata, può simulare un melanoma. Tipicamente, le
caratteristiche dermoscopiche includono le pseudocisti cor-
nee e gli sbocchi simil-comedonici, ma anche strutture a
impronta digitale, e aree con aspetto cerebriforme con giri e sol-
chi. Recentemente sono stati descritti come caratteristici del-
le cheratosi seborroiche i limiti nettamente demarcati e i bor-
di «a morsicatura concava» («moth-eaten borders») 68 osser-
vabili soprattutto nelle cheratosi seborroiche in fase iniziale e/o
lentigo solari. Talvolta si possono, però, osservare anche un fal-
so reticolo pigmentato (nelle cheratosi seborroiche di tipo reti-
colare) e strutture pseudo-globulari 68, 69. Nelle lesioni dubbie
è consigliabile l’asportazione, nell’evenienza, sebbene rara,
di un melanoma che simula una cheratosi seborroica 67.
L’esame dermoscopico delle lesioni di natura vascolare,
attraverso la ricerca di criteri specifici, consente di esclude-
re il melanoma con elevata accuratezza. Gli emangiomi sono
caratterizzati da un pattern lacunare, per la presenza di nume-
rose aree ovoidali ben circoscritte, di colore dal rosso al ros-
so bluastro, denominate lacune rosse. Queste strutture vanno
differenziate dalle aree rosso-lattescenti, meno ben definite,
che possono talvolta, ma in maniera specifica, essere osser-
vate nel melanoma 14. Nel granuloma piogenico il pattern
lacunare può non essere facilmente riconoscibile, quindi è
consigliabile l’escissione con esame istologico nell’adulto. Gli
angiocheratomi sono caratterizzati da lacune di colore dal
rosso-bluastro al nero, associate ad aree cheratosiche bianco-
giallastre 14. Infine gli ematomi subcornei mostrano dermo-
scopicamente un’area omogenea di colorito nerastro ma anche
un aspetto pseudo-parallelo o pseudo-globulare.
Standard tecnologici in dermoscopia
Il dermatoscopio e il videodermatoscopio rappresentano
gli strumenti più utilizzati per eseguire l’esame dermosco-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHIMENTI
pico 1-3, 14. In alcuni centri si impiega anche lo stereomi-
croscopio, implementato da sistemi digitali con telecame-
re ad alta risoluzione (3 CCD) 3, 52, 53. L’attuale standard
di riferimento, per quanto riguarda la fotografia dermosco-
pica, è costituito dal sistema di acquisizione Dermaphot
(Heine Optotechnik, Herrsching, Germany) che garantisce
un ottimo potere risolutivo e un’elevata qualità di immagi-
ne 73-75. Attualmente è disponibile in commercio una serie
di sistemi e strumenti di videodermoscopia digitale, che
consente di ottenere, in alcuni casi, un’immagine di qualità
sovrapponibile allo standard 76, 77. È auspicabile che, in un
futuro prossimo, le aziende si adeguino a tali standard, even-
tualmente attraverso l’istituzione di una commissione ad
hoc per la valutazione e la validazione dei sistemi di video-
dermoscopia.
Le problematiche della refertazione
dell’esame videodermoscopico
Anche se attualmente non esistono a riguardo né una nor-
mativa precisa né dati in letteratura, vogliamo farci portavoci
di una proposta di unificazione e standardizzazione del refer-
to che è opportuno rilasciare a seguito di un esame video-
dermoscopico.
Alla luce delle responsabilità di ordine medico-legale, di
deontologia professionale e, non ultimo, anche per un rico-
noscimento completo della demoscopia quale esame stru-
mentale di secondo livello, appare fondamentale il rilascio
di una refertazione idonea a seguito della prestazione spe-
cialistica effettuata. Pertanto, nel referto di un esame video-
dermoscopico, suggeriamo di includere sempre i seguenti
punti essenziali (criteri minimi):
— simmetria/asimmetria della lesione;
— aspetto globale;
— aspetti locali;
— conclusione diagnostica;
— indicazione sul trattamento.
Considerando i limiti attuali nella standardizzazione dei
diversi sistemi videodermoscopici (differente qualità del-
l’immagine e diversa risoluzione e qualità di stampa), rite-
niamo per il momento che l’immagine dermoscopica stam-
pata possa essere rilasciata al paziente, a discrezione del
dermatologo, specificando le attuali problematiche tecno-
logiche.
347
 ORIGINAL ARTICLES
The problem of clinically atypical nevi
submitted to verification biopsy:
can dermoscopy help excluding histologically common lesions?
Aim. Early diagnosis of cutaneous melanoma is associated
with costs in term of false positive diagnosis, i.e. benign pig-
mented lesions defined suspicious or equivocal by clinical
examination and submitted to verification biopsy. The aim of
the study is to investigate the possible role of dermoscopy in
selecting histologically common nevi from those with histo-
logic atypia.
Methods. Two hundred and sixthy-four clinically atypical
melanocytic nevi were classified by experienced pathologists as
nevi with histologic atypia and without histologic atypia. Two
observers analysed dermoscopic features with classic pattern
analysis and simplified algorithms. The study population was
divided in a training set and in a test set (observers aware of his-
tologic classification and blinded as to histologic classification
respectively).
Results. In the training set, nevi with histologic atypia signif-
icantly differed from common nevi since they showed with
higher frequency the following dermoscopic criteria: atypi-
cal pigment network, regression features, ABCD score ≥4 and
seven point check list ≥2. In the test set, however, only atypi-
cal pigment network and seven point score ≥2 were still more
frequent in atypical nevi than in common nevi. The negative
predictive value was less than 70% for any of selected der-
moscopic features. This means that when an observer predict
that a nevus lacking the above-mentioned criteria of atypia
is histologicaly common, it will be true in less than 70% of
cases.
Conclusion. Dermoscopy can play a role in detecting banal
melanocytic lesions, ie. nevi without histologic atypia, within the
pool of clinically equivocal lesions submitted to verification
biopsy. Further study with evidence-based design (prospec-
tive, randomised study) are needed to investigate the impact of
Address reprint requests to: P. Carli, Department of Dermatology, Uni-
versity of Florence, Via della Pergola 58/60, 50121 Firenze, Italy.
E-mail: carli@unifi.it
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:349-58
A. CHIARUGI, P. NARDINI, V. DE GIORGI, P. CARLI
Department of Dermatology
University of Florence, Florence, Italy
dermoscopy in a better selection of lesions to remove in real
practice.
KEY WORDS: Dermoscopy - Pigmented lesions - Atypical nevi -
False positive.
E arly diagnosis of cutaneous melanoma is associ-
ated with costs in term of false positive diagnosis,
i.e. benign pigmented lesions defined suspicious or
equivocal by clinical examination and submitted to
verification biopsy. This leads to inevitable scarring and
morbidity.
The frequency of verification biopsies is not negli-
gible: according to recent australian data, the ratio of
melanomas to benign lesions excised is 1:17 (1:26
including seborrheic keratosis) among family doc-
tors.1 When selection of lesions to be removed is made
by dermatologists at specialised pigmented lesion clin-
ics, less false positive excisions are made (about 1:6-
7 according to italian and British data).2, 3
It is likely that only small improvement of the malig-
nant: benign ratio is attainable without increasing the
risk of leaving a melanoma unexcised. Dermoscopy
(dermatoscopy, epiluminescence microscopy, surface
microscopy), a non invasive technique able to improve
diagnostic performance of melanoma diagnosis, may
349
CHIARUGI THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY

play a key role in a better preselection of lesions to TABLE I.—Training set. Clinical features of histologically atypical nevi
be removed. An improvement of specificity of and common nevi (N=114). The two subsets of nevi (common and
atypical on histology) are largely comparable as to clinical cha-
melanoma screening by use of dermoscopy has been racteristics.
recently shown.4
Atypical nevi Common nevi Pearson’s
We investigated the possible role of dermoscopy in Clinical feature (N=68) (N= 46) χ2 test
selecting, within the pool of lesions submitted to ver-
ification biopsy according to their clinical features, Palpability 27/68 (40%) 23/46 (50%) 0.277
Macular-papular aspect 22/68 (32%) 10/46 (22%) 0.216
histologically banal nevi from those with histologic
Asimmetry
atypia. Many studies have demonstrated that clinical N 3/68 (4%) 2/46 (4%)
atypia and histologic atypia in nevi have poor corre- 1 axis 30/68 (44%) 22/46 (48%)
lation.5 This means that a clinically atypical nevi can 2 axis 35/68 (51%) 22/46 (48%) NS
be histologic common and viceversa.5 Since the risk of N Colours
histopathologic misclassification between nevi with 1 colours 14/68 (20.5%) 14/46 (30%)
2 colours 46/68 (68%) 28/46 (61%)
histologic atypia and early melanoma is likely to occur,6 3 colours 8/68 (12%) 4/46 (9%) NS
the exclusion of histologically atypical nevi from ver- Type of predominant colour
ification biopsy, once judged equivocal from a clinical Black 3/68 (4%) 2/46 (4%)
point of view, could be hazardous. Conversely, an Dark brown 34/68 (50%) 25/46 (54%)
Light brown 27/68 (40%) 15/46 (33%)
histopathologic misclassification between early Blue/grey 1/68 (1.5%) 2/46 (4%)
melanoma and common nevi, even if equivocal by Pink/reddish 3/68 (4%) 2/46 (4%) NS
clinical examination, is unlikely. The latter lesions are Border irregularity 51/68 (75%) 31/46 (67%) NS
those who would mostly benefit of a non invasive clas-
sification by dermoscopy in order to be left in place
without risk for the patient. The training set (68 atypical nevi and 46 common
nevi) was aimed to identify dermoscopy features asso-
ciated with histologic atypia. In this study phase the
Materials and methods observers were aware of the histologic diagnosis.
Since the criteria for entry the study was the selection
This study included 264 clinically atypical of the lesion for diagnostic verification, clinical fea-
melanocytic nevi, consecutively excised for diagnos- tures of nevi, either histologically common or atypi-
tic verification in the period January 2001-December cal, were those of clinical atypical lesions, i.e. show-
2002 at the First Dermatology Unit, Florence. ing some or all the ABCD features of melanoma.
A staff of pathologists with experience in diagnosis Table I shows in details clinical features of common
of melanocytic lesions classified the lesions in accor- and atypical nevi included in the training set. No sta-
dance with the criteria of the NIH Consensus Conference tistically significant difference in the frequency of
on Diagnosis and Treatment of Early Melanoma 7 as selected clinical parameters was found. Therefore,
common melanocytic nevi (compound and junctional the two subsets of nevi (common and atypical on his-
nevi, n=133) and melanocytic nevi with histologic tology) were largely comparable as to clinical char-
atypia (architectural disorders and cytological atyp- acteristics.
ia, n=131). Therefore, in this set of lesions, nevi with The training set analysis should enable us in iden-
histologic atypia represented 131/264 (49.6%) of tifying dermoscopic features significantly associated
cases. with histologic atypia.
Before excision, clinical and dermoscopic images The possible role of these features in removing
were obtained using an F50 Nikon camera with objec- nevi without histologic atypia from the pool of lesions
tive AF micro Nikkor 60. For dermoscopy, images were selected for excision on the basis of their clinical
obtained after oil application on the surface of the lesions features will be investigated by means of test set
using a Dermaphot objective (magnification ×10). analysis, with observers blinded as to histologic diag-
We randomly divided the series of lesions in a train- nosis.
ing set and a test set. The same two experienced Test set included 63 nevi with histologic atypia and
observers (AC and PN) examined both training and 87 nevi without histological atypia.
test sets. Dermoscopy terminology adopted in the study was

350 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY
TABLE II.—Training set. Difference in the frequency of selected dermoscopic features between histologically common and atypical nevi (only
parameters with significant difference are shown).
Dermoscopic feature
Irregular/prominent pigment network
Regression structures
that of the Internet consensus meeting, held in 2000-
2001.8
We also tested the diagnostic algorithms: ABCD
rule of dermoscopy as published by Stolz et al.9 and 7-
Point Checklist according to Argenziano et al.10
The TDS derived from ABCD rule signifies a benign
lesion if the value is <4.75, malignant lesion if it is
>5.45 and doubtful lesion for intermediate values; the
TDS for histologically atypical nevi is frequently
between 4.5 and 5.8. We tested our series of melanocyt-
ic lesions using different cut-off values (4.75 and 4) of
TDS.
The 7- Point Checklist is a method based on sim-
plified ELM pattern analysis; a total score ≥3 indi-
cates a malignant lesion. Also for this algorithm we
used two different cut off values : 3 and 2 points.
Additionally, both training set and test set were
classified in accordance with the dermoscopic clas-
sification of Clark’s nevi suggested by Hofmann-
Wellenhof et al.11 including the following global pat-
terns: reticular, globular, homogeneous, or reticular-
globular, reticular-homogeneous, globular-homoge-
neous if two components were dominant. Concerning
the distribution of pigmentation the nevi were classi-
fied with central hyperpigmentation/hypopigmenta-
tion, eccentric peripheral hyperpigmentation/hypopig-
mentation, multifocal hyperpigmentation/hypopig-
mentation.
Statistical analysis
For statistical analysis non parametric test were used
(χ2 test, Fisher’s exact test when appropriate). In order
to evalute the power of selected dermoscopic meth-
ods in the identification of histologic common nevi
(i.e. without histologic atypia), the negative predic-
tive value (NPV) was calculated (true negative/true
negative + false negative). The NPV represents the
probability that a nevus defined “without histologic
atypia” by dermoscopy will be eventually confirmed
by histologic examination.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHIARUGI
Atypical nevi Common nevi Pearson’s
(N=68) (N=46) χ2 test
40/68 (59%) 18/46 (39%),3 P=0.039
29/68 (43%) 9/46 (19,5%) P=0.010
Results
Training set
As expected, nevi with and without histologic atyp-
ia were not dissimilar from a clinical point of view
(Table I). To the contrary, these two subsets of nevi
showed significant difference about dermoscopic fea-
tures.
Concerning the frequency of major features select-
ed on the basis of the Consensus Conference 2001,
an irregular/prominent pigment network, i.e. a net-
work with dark or thick lines and large holes, was
found in 59% of nevi with histologic atypia and in
39% of those without atypia (P=0.039). Dermoscop-
ic features of regression, i.e. blue and with areas, were
found in 43% of nevi with atypia and in 19.5% of
those without atypia (P=0.010) (Table II). No differ-
ence was found for other features. Figures 1 and 2
show major dermoscopic features of nevi with histo-
logical atypia.
Concerning simplified algorithms, nevi with histo-
logic atypia more frequently than common nevi
reached a value >4.75 (43% vs 19.5%) (P=0.02)
according to ABCD rule (threshold of suspicion) and
≥3 according to the 7- Point Checklist rule (threshold
for malignancy) (20.6 vs 6.5%) (P=0.039) (Table III).
A statistically significant difference between nevi
with and without histologic atypia was also found
concerning the classification of Clark’s nevi reported
by Hofmann-Wellenhof, both concerning global fea-
tures (Table IV) and distribution of pigmentation
(Table V).
Test set
Dealing with test set, the presence of an atypical
pigment network, but no that of regression features, was
significantly associated with histologic atypia (60.3%
vs 42.5%)(Table VI). According to the NPV associ-
ated with this parameter (66.6%), the use of this fea-
ture as a marker for excision, would have resulted in
351
CHIARUGI THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY

TABLE III.—Training set. Distribution of the SCORE according to 7

- Point Check List and TDS according to ABCD rule of dermoscopy
in histologically atypical and common nevi.
Atypical nevi Common nevi Pearson’s
(N=68) (N= 46) χ2 test

7-Point check list

<3 54/68 (79.4%) 43/46 (93.5%)
≥3 14/68 (20.6%) 3/46 (6.5%) P=0.039
<2 22/68 (32.4%) 32/46 (69.6%)
≥2 46/68 (67.6%) 14/46 (30.4%) P=0.000
ABCD score
<4.75 39/68 (57%) 37/46 (80.5%)
≥4.75 29/68 (43%) 9/46 (19.5%) P=0.020
<4 26/68 (38%) 28/46 (61%)
≥4 42/68 (62%) 18/46 (39%) P=0.018

TABLE IV.—Training set. Dermoscopic classification of histologi-

Figure 1.—Dermoscopic image (10×) of a nevus with histologic atypia. The cally atypical and common nevi according to “Hofmann-Wel-
arrow indicates atypical pigment network. lenhof”: global dermoscopic pattern.
Exact test
Global dermoscopic Atypical nevi Common nevi (Monte-
pattern (N=68) (N=46) carlo)

Reticular 32/68 (47%) 19/46 (41%)

Globular 1/68 (1.5%) 8/46 (17%)
Homogeneous 2/68 (3%) 1/46 (2%)
Reticular-globular 10/68 (15%) 7/46 (15%)
Reticular-homogeneous 18/68 (26%) 10/46 (21%)
Globular-homogeneous 4/68 (6%) 1/46 (2%)
Unclassified 1/68 (1.5%) 0/46 (0%)
P=0.011

TABLE V.—Training set. Dermoscopic classification of histological-

ly atypical and common nevi according to “Hofmann-Wellenhof”:
distribution of pigmentation.
Exact test
Distribution Atypical nevi Common nevi (Monte-
of pigmentation (N=68) (N=46) carlo)
Figure 2.—Dermoscopic image (10×) of a nevus with histologic atypia. The
arrows indicate regression structures. Blue area (black arrow), white area Uniform pigmentation 3/68 (4%) 12/46 (26%)
(empty arrow). Central hyperpigmentation 13/68 (19%) 6/46 (13%)
Central hypopigmentation 6/68 (9%) 7/46 (15%)
Eccentric peripheral hyper- 9/68 (13%) 7/46 (15%)
pigmentation
about 33% of false negative diagnosis, ie. nevi with his- Eccentric peripheral hypo- 13/68 (19%) 3/46 (6.5%)
tologic atypia left unexcised. pigmentation
Multifocal hyperpigmenta- 24/68 (35%) 11/46 (24%)
Concerning semiquantitative algorithms, both the tion and hypopigmenta-
ABCD rule and the seven point check-list confirmed tion
their possible role in selecting nevi with from those P=0.011
without atypia (Table VII). As occurred in the train-
ing set, even in the test set nevi with histologic atyp-
ia more frequently reached the threshold value of (instead of 3) as threshold value for the seven point
suspicion compared to common nevi. The NPV check list (Table VII).
ranged from 59.2% with a threshold point of 4 According to the dermoscopic classification of Clark
(instead of 4-75) for the ABCD rule to 66.6% for 2 nevi, no significant difference neither in the frequen-

352 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY CHIARUGI

TABLE VI.—Test set. Difference in frequency of selected dermoscopic features between histologically atypical and common nevi.
Atypical nevi Common nevi Pearson’s
Dermoscopic feature (N=63) (N=87) χ2 test NPV

Irregular/prominent pigment network 38/63 (60,3%) 37/87 (42,5%) P=0.032 66.6%

Regression structures 19/63 (30%),3 24/87 (27,5%) P=0.700 58.8%

TABLE VII.—Training set. Distribution of the score according to 7 - TABLE VIII.—Test set. Dermoscopic classification of histologically aty-
Point Check List and TDS according to ABCD rule dermoscopy in pical and common nevi according to “Hofmann-Wellenhof”: glo-
histologically atypical and common nevi. bal dermoscopic pattern.
Atypical nevi Common nevi Pearson’s Exact test
(N=63) (N=87) χ2 test NPV Global dermoscopic Atypical nevi Common nevi (Monte-
pattern (N=63) (N=87) carlo)
7-Point check list
<3 46/63 (73%) 76/87 (87.4%) Reticular 360/63 (57.1%) 39/87 (41%)
≥3 17/63 (27%) 11/87 (12.6%) P=0.026 62.2% Globular 1/63 (1.6%) 15/87 (17.2%)
Homogeneous 7/63 (11.1%) 8/87 (9.2%)
<2 27/63 (42.9%) 56/87 (64.4%) Reticular-globular 6/63 (9.5%) 5/87 (6%)
≥2 36/63 (57.1%) 31/87 (35.6%) P=0.009 66.6% Reticular-homogeneous 13/63 (20.6%) 18/87 (21.7%)
ABCD score Unclassified 0/63 (0%) 0/87 (0%)
<4.75 44/63 (69.8%) 71/87 (81.6%) P=0.093 NS
≥4.75 19/63 (30.2%) 16/87 (18.4%) 61.7%
<4 33/63 (52.4%) 58/87 (66.7%) P=0.077
≥4 30/63 (46.6%) 29/87 (33.3%) 59.2%
TABLE IX.—Training set. Dermoscopic classification of histologi-
cally atypical and common nevi according to “Hofmann-Wel-
lenhof”: distribution of pigmentation.
cy of global pattern nor in that of distribution of pig- Exact test
mentation between common and atypical nevi was Distribution Atypical nevi Common nevi (Monte-
of pigmentation (N=63) (N=87) carlo)
found dealing with test set (Table VIII, IX).
Uniform pigmentation 19/63 (30.2%) 28/87 (32.2%)
Central hyperpigmentation 9/63 (14.3%) 12/87 (13.8%)
Central hypopigmentation 7/63 (11.1%) 12/87 (13.8%)
Discussion and conclusions Eccentric peripheral hyper- 15/63 (23.8%) 19/87 (21.8%)
pigmentation
In recent years, dermoscopy proved to be able to Eccentric peripheral hypo- 7/63 (11.1%) 12/87 (13.8%)
pigmentation
increase the accuracy of melanoma diagnosis com- Multifocal hyperpigmenta- 6/63 (9.5%) 4/87 (4.5%) NS
pared to that achieved by visual examination alone.12, 13 tion and hypopigmenta-
Although further studies are still needed, it is con- tion
ceivable that dermoscopy reduce the false positive rate
in melanoma screening.4, 14
Among benign melanocytic lesions, verification diagnostic verification, excluding histologically com-
biopsy to exclude melanoma is currently undertaken mon nevi.
not only for Spitz nevi and nevi with histologic atyp- Since the risk of misclassification between nevi with
ia but also for histologically common nevi.4 As report- atypia and early melanoma by pathologist has been
ed by many studies, clinical features of a nevus do demonstrated by several authors,6, 15 a further reduc-
not correlate with histologic atypia.5 Therefore, even tion of false positives by means of avoid excision of
an histologically common nevus may show some or nevi with histologic atypia, when clinically doubtful,
all the ABCD signs of melanoma sometimes repre- seems to the contrary less desirable.
senting a cause of concern both for patient and physi- According to training set analysis, two dermoscop-
cian. ic features were more frequently found in nevi with his-
We sought to evaluate if dermoscopy may help to tologic atypia than in common nevi: atypical /promi-
achieve—compared to visual examination alone—a nent pigment network and regression features. This
better selection of melanocytic lesions to be excised for finding is in perfect agreement with the data provided

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 353

CHIARUGI THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY
by a previous study of our group conducted on a dif-
ferent series of lesions examined by another group of
observers.16
Also the semiquantitative score assigned in accor-
dance with two of the most popular simplified algo-
rithms for melanoma diagnosis, the ABCD rule of der-
moscopy 9 and the new seven-point check-list,10 sig-
nificantly differed between the two subset of nevi for
both the threshold values adopted for each method (4
and 4.75 for ABCD, 2 and 3 for seven point). This
finding disagree with a previous study where no sig-
nificant difference between nevi with atypia and those
without was found concerning ABCD score.16 Possi-
ble explainations include a different source of lesions
(clinically equivocal and unequivocal lesion in the
previous study), and a different procedure adopted in
the analysis of the data (comparison of median val-
ues in the previous study, comparison among fre-
quencies in pre-established categories in the present
study). However, only the seven point score confirmed
a statistically significant difference between the two
subsets of nevi according to test set analysis.
No difference in the frequency of dermoscopic sub-
types of Clark nevi 11 was found. In this context one
should take into account that our series was small and
the data have to be interpreted cautiously. Since the
large number of categories proposed for the classifi-
cation of nevi according to Hofman-Wellenhof et al.,11
our small series of lesions may not reach a sufficient
statistical power. Our data strongly cool down the
expectations of this new classification of Clark’s nevi:
In the opinion of the proposers “this classification
should be regarded not just as an academic morpho-
logic exercise but as a classification system that will
lead to a better understanding of the biological char-
acteristics of these melanocytic nevi”.11 The fact that
no statistical difference in the frequency of the Hof-
mann’s classification subtypes has been found between
nevi with atypia and common nevi greatly limits its
validity in the study of biological features of benign
melanocytic lesions.
In sum, according to training set, observers may be
able to identify—within a pool of nevi indistiguinsh-
able among them as to clinical features—nevi with
histologic atypia from those without atypia. The “pro-
totypic” nevus with atypia would therefore be a nevus
showing atypical network and/or regression structures,
that scores >2 in accordance with seven point check list
and >4 in accordance with ABCD rule of dermoscopy.
354 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
This scenario has been in part confirmed by test set
analysis: only atypical pigment network and seven
point score ≥2 or ≥3 confirmed their different distrib-
ution between the two subset of nevi with observers
blinded as to histologic diagnosis. Neither regression
features nor the ABCD score were to the contrary
longer associated with histologic atypia in the test set
analysis.
Among the points of strength of this study we men-
tion the design of the study, with a training set and a test
set; this allows to verify prospectively to what extent
what found in the training set by open observers is
confirmed in a new series of lesions by blinded exam-
iners. Moreover, this study includes only nevi excised
for diagnostic verification decided on the basis of clin-
ical features, thus approaching as far as possible the
diagnostic setting found in practice. Being in this study
the clinical characteristics of nevi with histologic atyp-
ia not dissimilar to those without atypia we minimized
the risk of lesion’s classification influenced by clini-
cal factors acting as confounders. Among the study’s
weaknesses, we mention that fact that this study is
based on lesion’s classification on photographic
images, in a posteriori diagnostic setting
In order to investigate what consequences this find-
ing may have in clinical practice, we calculated the
negative predictive value associated with each of the
above-mentioned feature, i.e. the probability that a
nevus lacking these features eventually be histologi-
cally common. Unfortunately, the NPV was not high-
er than 60% as average for any of the selected der-
moscopic features. This means that when an observ-
er predict that a nevus lacking the above mentioned cri-
teria of atypia is histologically common, it will be true
in 60% only of cases.
Further prospective, randomized studies should con-
firm if the above-mentioned dermoscopic criteria asso-
ciated with histologic atypia in nevi, i.e. atypical net-
work, regression features, ABCD score and Seven
point score can help dermatologist in change the
lesion’s management with fewer excision of clinical-
ly equivocal but histologically common nevi.
References
1. English DR, Burton RC, del Mar CB, Donovan RJ, Ireland PD, Emery
G. Evaluation of aid to diagnosis of pigmented skin lesions in gener-
al practice: controlled trial randomised by practice. BMJ 2003;327:375.
2. Carli P, De Giorgi V, Betti R, Vergani R, Catricala C, Mariani G et al.
Relationship between cause of referral and diagnostic outcome in
Agosto 2005
THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY
pigmented lesion clinics: a multicenter survey of the Italian Multi-
disciplinary Group on Melanoma (GIPMe). Melanoma Res
2003;13:207-11.
3. Gibbon KL. Pigmented lesion clinics – are they a waste of resources?
Clin Exper Dermatol 1998;23:3-8.
4. Carli P, De Giorgi V, Crocetti E, Mannone F, Massi D, Chiarugi A et
al. Improvement of malignant/benign ratio in excised melanocytic
lesions in the “dermoscopy era”: a retrospective study 1997-2001.
Br J Dermatol 2004;150:687-92.
5. Annessi G, Cattaruzza MS, Abeni D, Baliva G, Laurenza M, Macchini
V et al. Correlation between clinical atypia and histologic dysplasia
in acquired melanocytic nevi. J Am Acad Dermatol 2001;45:77-85.
6. Corona R, Mele A, Amini M, De Rosa G, Coppola G, Piccardi P et al.
Interobserver variability of the histopathologic diagnosis of cuta-
neous melanoma and other pigmented skin lesions. J Clin Oncol
1996;14:1218-23.
7. National Institute of Health Consensus Conference. Diagnosis and
treatment of early melanoma. JAMA 1992;268:1314-9.
8. Argenziano G, Soyer HP, Chimenti S, Talamini R, Corona R, Sera F
et al. Dermoscopy of pigmented skin lesions: results of a consensus
meeting via the Internet. J Am Acad Dermatol 2003;48:679-93.
9. Stolz W, Riemann A, Cognetta AB, Pillet l, Abmayr W, Holzel D et al.
ABCD rule of dermatoscopy: a new practical method for early recog-
nition of melanoma. Eur J Dermatol 1994;4:521-7.
Il problema dei nevi clinicamente atipici sottoposti a verifica diagnostica:
la dermoscopia può essere di aiuto per escludere nevi istologicamente comuni?
L a diagnosi precoce del melanoma cutaneo è associata a
dei costi in termini di diagnosi di falsi positivi, vale a dire
lesioni pigmentate benigne definite dubbie o sospette all’e-
same clinico e sottoposte a verifica bioptica. Questo porta ine-
vitabilmente ad una serie di conseguenze per il paziente (sta-
to di malattia, formazione di cicatrici).
La frequenza delle verifiche bioptiche non è trascurabile:
secondo recenti dati australiani il rapporto fra melanomi e
lesioni benigne asportate è 1:17 (1:26 se includiamo anche le
cheratosi seborroiche) quando la diagnosi viene effettuata dai
medici di medicina generale 1. Quando la selezione delle lesio-
ni da asportare avviene ad opera di dermatologi in cliniche
specializzate nella diagnosi delle lesioni pigmentate, risulta-
no effettuate un minor numero di escissioni di falsi positivi (cir-
ca 1:6-7 secondo dati italiani ed anglosassoni) 2, 3.
È probabile che si possa ottenere solo un piccolo miglio-
ramento del rapporto lesioni maligne/lesioni benigne senza
aumentare il rischio di lasciare in sede un melanoma. La der-
moscopia (o dermatoscopia, microscopia a epiluminescenza,
microscopia di superficie), una tecnica non invasiva capace di
migliorare la performance nella diagnosi del melanoma, può
giocare un ruolo chiave nel migliorare la selezione delle lesio-
ni pigmentate che devono essere escisse. Dati recenti hanno
mostrato che l’uso della dermoscopia può aumentare la spe-
cificità nello screening del melanoma 4.
Abbiamo indagato il possibile ruolo della dermoscopia nel
selezionare, all’interno di un pool di lesioni sottoposte a
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHIARUGI
10. Argenziano G, Fabbrocini G, Carli P, De Giorgi V, Sammarco E,
Delfino M. Epiluminescence microscopy for the diagnosis of doubt-
ful melanocytic skin lesions: comparison of the ABCD rule of der-
matoscopy and a new 7-point check list based on pattern analysis.
Arch Dermatol 1998;134:1563-70.
11. Hofmann-Wellenhof R, Blum A, Wolf IH, Piccolo D, Kerl H, Garbe
C et al. Dermoscopic classification of atypical melanocytic nevi (Clark
nevi). Arch Dermatol 2001;137:1575-80.
12. Pehamberger H, Steiner A, Wolff K. In vivo epiluminescence
microscopy of pigmented skin lesions. I. Pattern analysis of pig-
mented skin lesions. J Am Acad Dermatol 1987;17:571-83.
13. Carli P, De Giorgi V, Giannotti B. Dermoscopy and early diagnosis of
melanoma. The Light and the Dark. Arch Dermatol 2001;137:1641-4.
14. Carli P, De Giorgi V, Chiarugi A, Nardini P, Weinstock MA, Crocetti
E et al. Addition of dermoscopy to conventional naked-eye examination
in melanoma screening: a randomized study. J Am Acad Dermatol
2004;50:683-9.
15. Ferrara G, Argenziano G, Soyer HP, Corona R, Sera F, Brunetti B et
al. Dermoscopic and histopathologic diagnosis of equivocal melanocyt-
ic skin lesions: an interdisciplinary study on 107 cases. Cancer
2002;95:1094-100.
16. Carli P, De Giorgi V, Massi Giannotti B. The role of pattern analysis
and the ABCD rule of dermoscopy in the detection of histologic atyp-
ia in melanocytic naevi. Br J Dermatol 2000;143:290-7.
verifica bioptica per le loro caratteristiche cliniche, nevi
istologicamente banali da quelli con atipia istologica. Mol-
ti studi hanno dimostrato che nei nevi l’atipia clinica e l’a-
tipia istologica hanno una scarsa correlazione 5. Questo
significa che un nevo clinicamente atipico può essere isto-
logicamente comune e viceversa 5. Poiché il rischio di un
errore di classificazione istopatologica fra nevi con atipia isto-
logica e melanoma iniziale si può verificare con una certa fre-
quenza 6, escludere i nevi istologicamente atipici da una
verifica bioptica, una volta giudicati equivoci da un punto di
vista clinico, può essere rischioso. Al contrario, una misclas-
sificazione istopatologica fra melanoma iniziale e nevi comu-
ni, anche se valutati equivoci all’esame clinico, risulta poco
probabile. Sono queste ultime lesioni che potrebbero bene-
ficiare di una classificazione non invasiva da parte della der-
moscopia ed essere quindi lasciate in sede senza rischi per il
paziente.
Materiali e metodi
Questo studio ha incluso 264 nevi melanocitici clinica-
mente atipici, consecutivamente escissi per la verifica dia-
gnostica nel periodo gennaio 2001-dicembre 2002 presso la
I Clinica Dermatologica, Università di Firenze.
Uno staff di patologi esperti nella diagnosi delle lesioni
355
CHIARUGI THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY
melanocitiche ha classificato le lesioni in accordo con i cri-
teri della NIH Consensus Conference per la Diagnosi e il
Trattamento del Melanoma Precoce 7 come nevi melanoci-
tici comuni (nevi giunzionali e composti, n=133) e nevi
melanocitici con atipia istologica (nevi con disordine archi-
tetturale e atipia citologica, n=131). Quindi, in questo grup-
po di lesioni, i nevi con atipia istologica rappresentavano
131/264 (49,6%) dei casi.
Prima dell’escissione, sono state ottenute le immagini cli-
niche e dermoscopiche utilizzando una macchina fotografi-
ca Nikon F50 con obiettivo micro Nikkor 60. Per la dermo-
scopia, le immagini sono state ottenute, dopo applicazione
di olio sulla superficie delle lesioni, usando un obiettivo
Dermaphot (ingrandimento 10×).
Abbiamo suddiviso casualmente la serie delle lesioni in un
training set e in un test set. Gli stessi due osservatori esper-
ti (AC e PN) hanno esaminato sia il training che il test set.
Il training set (68 nevi atipici e 46 nevi comuni) aveva lo
scopo di identificare le caratteristiche dermoscopiche asso-
ciate con atipia istologica. In questa fase dello studio gli
osservatori erano a conoscenza della diagnosi istologica.
Poiché il criterio per l’entrata nello studio era la selezione del-
la lesione per verifica diagnostica, le caratteristiche clini-
che dei nevi, sia istologicamente atipici sia comuni, erano
quelle di lesioni clinicamente atipiche, quindi che mostravano
alcuni o tutti i criteri ABCD del melanoma. La Tabella I
mostra in dettaglio le caratteristiche cliniche dei nevi comu-
ni ed atipici inclusi nel training set: non sono state trovate dif-
ferenze statisticamente significative nella frequenza dei para-
metri clinici selezionati. Quindi i due sottogruppi di nevi
(istologicamente comuni ed atipici) erano ampiamente com-
parabili dal punto di vista delle caratteristiche cliniche.
L’analisi del training set dovrebbe permetterci di identi-
ficare le caratteristiche dermoscopiche significativamente
associate all’atipia istologica.
Il possibile ruolo di queste caratteristiche nell’escludere
i nevi senza atipia istologica dal pool di lesioni selezionate
per l’escissione chirurgica sulla base delle loro caratteristi-
che cliniche sarà valutato per mezzo dell’analisi del test set,
in cui gli osservatori non sono a conoscenza della diagnosi
istologica.
Il test set comprendeva 63 nevi con atipia istologica e 87
nevi senza atipia istologica.
La terminologia dermoscopica adottata nello studio è sta-
ta quella dell’Internet Consensus Meeting, svoltosi nel 2000-
2001 9.
Sono stati anche testati gli algoritmi diagnostici (regola
dell’ABCD della dermoscopia come pubblicata da Stolz et
al. 8) e la 7 – Point Checklist in accordo con Argenziano et
al. 10.
Il TDS (Total Dermoscopy Score) ricavato dalla regola
dell’ABCD indica che la lesione è benigna se il valore è
<4,75, che la lesione è maligna se il valore è >5,45 e che la
lesione è dubbia se il valore è intermedio; il TDS per i nevi isto-
logicamente atipici risulta frequentemente compreso fra 4,5
e 5,8. Abbiamo testato le nostre serie di lesioni melanocitiche
usando due differenti valori di cut-off del TDS (4,75 e 4).
356 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
La 7 – Point Checklist è un metodo basato sull’analisi di
pattern semplificata; un punteggio totale ≥3 indica una lesio-
ne maligna. Anche per questo algoritmo abbiamo usato due
differenti valori di cut-off: 3 e 2 punti.
Infine sia il training set che il test set sono stati classificati
in accordo con la classificazione dei nevi di Clark suggeri-
ta da Hofmann-Wellenhof et al. 11 che comprende i seguen-
ti pattern globali: reticolare, globulare, omogeneo, o retico-
lare-globulare, reticolare-omogeneo, globulare-omogeneo
se ci sono due componenti dominanti. Riguardo la distribu-
zione della pigmentazione i nevi sono stati classificati con
iperpigmentazione/ipopigmentazione centrale, iperpigmen-
tazione/ipopigmentazione eccentrica periferica, iperpig-
mentazione/ipopigmentazione multifocale 11.
Analisi statistica
Per l’analisi statistica sono stati utilizzati test non para-
metrici (test del χ2, test esatto di Fisher quando appropria-
ti). Al fine di valutare il potere dei metodi dermoscopici
selezionati nell’identificazione dei nevi istologicamente
comuni (senza atipia istologica), è stato calcolato il valo-
re predittivo negativo (NPV) (vero negativo/vero negati-
vo+falso negativo). Il NPV rappresenta la probabilità che
un nevo definito «senza atipia istologica» mediante la der-
moscopia, sarà eventualmente confermato tale dall’esame
istologico.
Risultati
Training set
Come atteso, i nevi con e senza atipia istologica non sono
risultati differenti dal punto di vista clinico (Tabella I). Al con-
trario, questi due sottogruppi hanno mostrato differenze
significative per quanto riguarda le caratteristiche dermo-
scopiche.
Riguardo alla frequenza dei maggiori criteri selezionati sul-
la base della Consensus Conference del 2001, è stato evi-
denziato un reticolo pigmentario irregolare/prominente (per
esempio un reticolo con linee scure o spesse e maglie larghe)
nel 59% dei nevi con atipia istologica e nel 39% di quelli sen-
za atipia (P=0,039). I caratteri dermoscopici di regressione
(per esempio aree blu ed aree bianche) sono stati trovati nel
43% dei nevi con atipia istologica e nel 19,5% di quelli sen-
za atipia (P=0,010) (Tabella II). Nessuna differenza è stata
trovata per gli altri caratteri. Le Figure 1 e 2 mostrano le
maggiori caratteristiche trovate in nevi con atipia istologica.
Riguardo agli algoritmi semplificati, i nevi con atipia isto-
logica hanno raggiunto più frequentemente dei nevi comu-
ni il valore >4,75 (soglia di sospetto) secondo la regola del-
l’ABCD (43% vs 19,5%) (P=0,02) e il valore ≥3 (soglia di
malignità) in accordo alla 7 – Point Checklist (20,6% vs
6,5%) (P=0,039) (Tabella III).
Anche per quanto riguarda la classificazione dei nevi di
Clark riportata da Hofmann-Wellenhof è stata trovata una dif-
Agosto 2005
THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY
ferenza statisticamente significativa fra nevi con e senza ati-
pia istologica relativa sia ai caratteri globali (Tabella IV) sia
alla distribuzione della pigmentazione (Tabella V).
Test set
Nel test set la presenza di un reticolo pigmentario atipico
è risultata significativamente associata con l’atipia istologi-
ca (60,3% vs 42,3%), mentre non lo è stata la presenza del-
le strutture di regressione (Tabella VI). Secondo il NPV asso-
ciato al reticolo pigmentario atipico (66,6%), l’uso di questo
parametro come marker per l’escissione, potrebbe compor-
tare una diagnosi di falsi negativi in circa il 33%, cioè nevi
con atipia istologica non asportati.
Riguardo agli algoritmi semiquantitativi, sia la regola del-
l’ABCD sia la 7 – Point Checklist, hanno confermato il loro
possibile ruolo nel selezionare nevi con atipia da quelli comu-
ni (Tabella VII). Come si è verificato nel training set, anche
nel test set i nevi con atipia istologica hanno raggiunto più fre-
quentemente il valore soglia di sospetto se comparati con i
nevi comuni. Il NPV oscillava fra il 59% per un valore soglia
di 4 (invece di 4,75) con la regola dell’ABCD ed il 66,6% per
un valore soglia di 2 (anziché 3) con la 7 – Point Checklist
(Tabella VII).
Nel test set, seguendo la classificazione dermoscopica dei
nevi di Clark secondo Hofmann-Wellenhof non sono emer-
se differenze significative fra i due sottogruppi di nevi né
nella frequenza del tipo di pattern globale né in quella della
distribuzione della pigmentazione (Tabella VIII, IX).
Discussione e conclusioni
In anni recenti, la dermoscopia ha dimostrato di poter
incrementare l’accuratezza della diagnosi del melanoma
rispetto a quella ottenuta con il solo esame visivo 12, 13. Seb-
bene ulteriori studi siano ancora necessari, è concepibile che
la dermoscopia riduca il tasso di falsi positivi nello screening
del melanoma 4, 14.
Nell’ambito delle lesioni melanocitiche, la verifica biop-
tica per escludere il melanoma, viene eseguita correntemente
non solo per i nevi di Spitz e per i nevi con atipia istologica,
ma anche per i nevi istologicamente comuni 4. Come ripor-
tato da numerosi studi, i caratteri clinici di un nevo non cor-
relano con i caratteri istologici 5; quindi anche un nevo isto-
logicamente comune può mostrare alcuni o tutti i segni del-
l’ABCD del melanoma, rappresentando talvolta una causa di
ansietà sia per il paziente sia per il medico.
Abbiamo cercato di valutare se la dermoscopia può aiutare
a raggiungere, rispetto al solo esame visivo, una miglior
selezione delle lesioni melanocitarie da sottoporre ad escis-
sione per verifica diagnostica, escludendo i nevi istologica-
mente comuni.
Poiché il rischio di misclassificazione da parte dei pato-
logi fra nevi con atipia e melanoma iniziale è stato dimo-
strato da parecchi Autori 6, 15, una riduzione ulteriore di fal-
si positivi evitando l’ escissione di nevi con atipia istologi-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHIARUGI
ca, quando siano dubbi clinicamente, sembra, al contrario,
meno auspicabile.
Analizzando il training set è stato evidenziato che due
caratteristiche dermoscopiche ricorrono più frequentemen-
te nei nevi con atipia istologica piuttosto che nei nevi comu-
ni: il reticolo pigmentario atipico/prominente e le strutture di
regressione. Questa evidenza è in perfetto accordo con i dati
forniti da un precedente studio del nostro gruppo condotto su
una serie diversa di lesioni esaminate da un altro gruppo di
osservatori 16.
Anche il punteggio semiquantitativo, assegnato in accor-
do con i due più popolari algoritmi semplificati, la regola
dell’ABCD della dermoscopia 9 e la nuova 7 – Point Check-
list 10, differiva significativamente fra i due sottogruppi di
nevi per entrambe i valori soglia utilizzati per ciascun meto-
do (4 e 4,75 per l’ABCD, 2 e 3 per la 7 – Point Checklist).
Questo risultato è in disaccordo con un precedente studio
dove non è stata rilevata nessuna differenza significativa, uti-
lizzando la regola dell’ABCD, fra nevi con atipia e nevi sen-
za atipia 16. Fra le possibili spiegazioni emerge la differente
provenienza delle lesioni (nel precedente studio erano inclu-
se sia lesioni clinicamente equivoche che lesioni non equi-
voche), e una diversa procedura adottata nell’analisi dei dati
(confronto della mediana nello studio precedente, confronto
fra frequenze in categorie prestabilite nel presente studio).
Tuttavia, analizzando il test set, solamente il punteggio del-
la 7 – Point Checklist ha confermato una differenza statisti-
camente significativa fra i due sottogruppi di nevi.
Analizzando i sottotipi dermoscopici dei nevi di Clark 11
nel test set non è emersa nessuna differenza significativa. In
questo contesto si può pensare che la nostra serie di lesioni fos-
se troppo piccola e che quindi i dati debbano essere inter-
pretati con cautela. È possibile che la nostra serie di lesioni sia
piccola per poter raggiungere un potere statistico sufficiente
dato il grande numero di categorie proposte per la classifi-
cazione dei nevi da Hofmann-Wellenhof et al. 11. I nostri dati
attenuano molto le aspettative di questa classificazione dei nevi
di Clark: nell’opinione dei proponenti «questa classificazio-
ne dovrebbe essere guardata non solo come un esercizio
morfologico accademico ma come un sistema classificativo
che condurrà ad una miglior conoscenza delle caratteristi-
che biologiche di questi nevi melanocitici» 11. Il fatto che
non sia stata trovata nessuna differenza statisticamente signi-
ficativa nella frequenza dei sottotipi della classificazione di
Hofmann-Wellenhoff fra nevi con atipia e nevi comuni limi-
ta fortemente la sua validità nello studio delle caratteristiche
biologiche delle lesioni melanocitiche benigne.
Riassumendo, secondo il training set, gli osservatori pos-
sono essere in grado di identificare, all’interno di un pool di
nevi non distinguibili fra loro per le caratteristiche cliniche,
nevi con atipia istologica e nevi senza atipia. Il «prototipo»
di nevo con atipia dovrebbe quindi essere un nevo che pre-
senta un reticolo pigmentario atipico e/o strutture di regres-
sione, che ha un punteggio >2 secondo la 7 – Point Check-
list e >4 secondo la regola dell’ABCD della dermoscopia.
Questo scenario è stato in parte confermato dall’analisi del
test set: con i due osservatori non a conoscenza della diagnosi
357
CHIARUGI THE PROBLEM OF CLINICALLY ATYPICAL NEVI SUBMITTED TO VERIFICATION BIOPSY
istologica, solamente il reticolo pigmentario atipico e il pun-
teggio secondo la 7 – Point Checklist ≥2 o ≥3 hanno con-
fermato la loro differente distribuzione fra i due sottogrup-
pi di nevi. Nell’analisi del test set né le strutture di regressione
né il punteggio secondo l’ABCD si sono mantenuti asso-
ciati all’atipia istologica.
Fra i punti di forza di questo studio vogliamo sottolinea-
re il disegno dello studio, strutturato con un training set ed
un test set; questo porta a verificare prospetticamente fino a
che punto ciò che è stato trovato nel training set in aperto dagli
osservatori viene confermato in una nuova serie di lesioni esa-
minate in cieco dai medesimi osservatori. Inoltre, questo
studio include solo nevi asportati per verifica diagnostica
sulla base delle caratteristiche cliniche, avvicinandosi quin-
di quanto più possibile alla seduta diagnostica che avviene
nella pratica. Poiché in questo studio le caratteristiche cliniche
dei nevi con atipia istologica sono simili a quelle dei nevi sen-
za atipia abbiamo ridotto al minimo il rischio di fare una
classificazione influenzata da fattori clinici che agiscano da
confondenti. Come punti deboli dello studio menzioniamo
il fatto che questo studio è basato su una classificazione del-
le lesioni fatta su immagini fotografiche, in una sessione
diagnostica a posteriori.
Al fine di capire quali conseguenze possano avere nella
pratica clinica queste evidenze, abbiamo calcolato il valore
predittivo negativo associato a ciascuno dei caratteri prece-
dentemente citati, vale a dire la probabilità che un nevo che
non abbia tali caratteri sia istologicamente comune. Sfortu-
natamente il NPV non è risultato mediamente più del 60%
per ogni carattere dermoscopico selezionato; ciò significa
che quando un osservatore predice che un nevo mancante
dei sopra citati criteri di atipia è istologicamente comune, ciò
sarà vero solamente nel 60% dei casi.
In futuro, studi randomizzati potrebbero confermare se i
criteri dermoscopici associati ad atipia istologica sopra men-
zionati, come il reticolo pigmentario atipico, le strutture di
regressione, il punteggio della regola dell’ABCD e della 7 –
Point Checklist possono aiutare i dermatologi a modificare
la gestione delle lesioni pigmentate, con un minor numero di
escissioni di nevi istologicamente comuni, ma clinicamen-
te equivoci.
358 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
Riassunto
Obiettivo. La diagnosi precoce del melanoma cutaneo è
associata a dei costi in termini di diagnosi di falsi positivi, cioè
lesioni pigmentate benigne definite sospette o equivoche
all’esame clinico e sottoposte a verifica bioptica. Scopo del-
lo studio è indagare il possibile ruolo della dermoscopia nel
selezionare i nevi istologicamente comuni da quelli con ati-
pia istologica.
Metodi. Duecentosessantaquattro nevi clinicamente ati-
pici sono stati classificati da patologi esperti come nevi con
atipia istologica e nevi senza atipia istologica. Due osser-
vatori hanno analizzato le caratteristiche dermoscopiche
utilizzando la classica analisi di pattern e gli algoritmi sem-
plificati. La popolazione oggetto dello studio è stata divisa
in un training set e in un test set (osservazione in aperto e in
cieco rispettivamente riguardo alla classificazione istolo-
gica).
Risultati. Nel training set i nevi con atipia istologica
differivano significativamente dai nevi comuni, avendo
mostrato con frequenza maggiore i seguenti criteri der-
moscopici: reticolo pigmentario atipico, strutture di regres-
sione, ABCD score ≥4 e seven point check list ≥2. Nel test
set, tuttavia, solamente il reticolo pigmentario atipico e il
seven point score ≥2 si sono mantenuti più frequenti nei nevi
atipici che nei nevi comuni. Il valore predittivo negativo è
risultato inferiore al 70% per ciascun carattere dermosco-
pico selezionato. Questo significa che quando un osser-
vatore predice che un nevo mancante dei suddetti criteri di
atipica è istologicamente comune, ciò sarà vero in meno del
70% dei casi.
Conclusioni. La dermoscopia può giocare un ruolo nel
rilevare lesioni melanocitiche banali, come i nevi senza ati-
pie istologiche, all’interno di un pool di lesioni clinicamen-
te equivoche sottoposte a verifica bioptica. Studi ulteriori
(studi prospettici randomizzati) sono necessari per indaga-
re l’impatto della dermoscopia nella pratica reale per una
miglior selezione delle lesioni da asportare.
Parole chiave: Demoscopia - Lesioni pigmentate - Nevi ati-
pici - Falsi positivi.
Agosto 2005
 G ITAL DERMATOL VENEREOL 2005;140:359-72

The Italian Registry of Hereditary Epidermolysis Bullosa

G. TADINI 1, L. GUALANDRI 2, M. COLOMBI 3, M. PARADISI 4, C. ANGELO 4, G. ZAMBRUNO 4, L. CASTIGLIA 4,
G. ANNICCHIARICO 5, M. EL HASHEEM 6, S. BARLATI 3, R. GARDELLA 3, L. NALDI 7, E. BONIFAZI 8,
L. GAROFALO 8, G. MORETTI 9, R. CAVALLI 1, S. CAMBIAGHI 1, S. PERCIVALLE 1, M. BELLINVIA 1,
A. DI BENEDETTO 1, A. LOCATELLI 1, L. LUNARDON 1, E. BRUNI 1, A. PATRIZI 10, G. LEMBO 11, T. CAINELLI 12

Aim. At present, in Italy no exhaustive epidemiological study 1Institute of Dermatological Sciences,

exists on inherited epidermolysis bullosa (EB). The necessity to Center for Inherited Diseases
have an exact evaluation of Italian cases encouraged the setting University of Milan, IRCCS, Ospedale Maggiore Policlinico
up of the national Registry, in order to collect all notifiable Mangiagalli e Regina Elene, Milan, Italy
cases of this disease, with important implications in clinical Department of Dermatology I, IRCCS, Milan, Italy
2Department of Dermatology IV
knowledge, the development of prenatal diagnosis instruments
and the start of epidemiological genetic studies. S. Paolo Hospital, Milan, Italy
3Unit of Biogenetics
Methods. A hospital registry has been prepared: initially, it
collected the cases already known to study centers in the peri- Department of Biomedical and Biotechnological Sciences
od 1985-1993; then, it collected the new cases, until December University of Brescia, Brescia, Italy
4Istituto Dermopatico dell’Immacolata, Rome, Italy
31, 2002. The registry envisaged the presence of a data coor- 5Associazione Pugliese Epidermolisi Bollose, Bari, Italy
dinating center (Dermatology Clinic in Bergamo), 3 regional 6Department of Pediatric Dermatology
centers (CMCE in Milan, IDI in Rome and Bari Hospital) that Bambin Gesù Pediatric Hospital, Rome, Italy
collected patients from North, Center and South Italy respec- 7Department of Pediatric Dermatology V
tively, and DEBRA Italy. University of Milan, City Hospital, Bergamo, Italy
Results. In total, 697 cases have been notified (9 not yet classi- 8Department of Pediatric Dermatology
fied), with 28% epidermolytic EB, 10% junctional EB and University of Bari, Bari, Italy
62% dermolytic EB. EB incidence at December 31, 2002 was 9DEBRA Italy, Catania, Italy
0.1 new cases per million live births; prevalence at December 10Department of Dermatology
31, 2002 was 10.1 affected patients per million Italians. S. Orsola Hospital, Bologna, Italy
11Department of Dermatology
Conclusion. This epidemiological evaluation is representative
of the Italian situation; from these data a geographic distrib- University of Naples, Naples, Italy
ution of the disease in our country can be traced, with signifi-
cant effects on prevention strategy.
KEY WORDS: Hereditary epidermolysis bullosa - Registry - Inci-
dence - Prevalence.

Paper presented at the GISED National Congress, October 3-6, 2001,

Genoa, Italy.
Fundings. Grant offered by Lions Club Multidistretto Italy - Leo Club
Telethon.
Address reprint requests to: Dr. G. Tadini, Institute of Dermatological
Sciences, Center for Inherited Cutaneous Diseases, University of Milan,
IRCCS, Ospedale Maggiore Mangiagalli e Regina Elena, Via Pace 9,
Milan, Italy.
H ereditary epidermolysis bullosa (EB) represents
a heterogeneous group of genetically determined
mechano-bullous dermatosis.
Though these diseases are rare in the general pop-
ulation, they have great clinical relevance because of
the possible reduction in life quality and life expectan-
cy they could cause.
The common clinical features of all EB is the marked
cutaneous and mucous fragility, which lead to blister

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 359

TADINI HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY

DERMO-EPIDERMICAL JUNCTION AND FORMS OF EB

Keratin
BASAL TONO-FILAMENTS K5 K14
KERATINOCYTE EBS
EMI-DESMOSOME
Plactin
Iam 5,
Lamina lucida ANCHORAGE α6β4, EBG
FILAMENTS BP180
Lamina densa

ANCHORAGE VII coll. EBD

Sub FIBERS
Lamina densa

Figure 1.—Pattern of dermo-epidermal junction and proteins implicated in the pathogenesis of inherited epidermolysis bullosa.

formation after minimal traumatism. The adnexial
structures could be involved, resulting sometimes in a
partial or complete absence.
Depending on the ultrastructural level within which
the cleavage plane of the blister occurs, we can clas-
sify epidermolytic, junctional and dermolytic EB. This
is possible with immunohistochemical assays and with
electronic microscopy.
In epidermolytic EB, the molecular defect involves
keratins K5 and K14 and plectin, a protein which is also
present in the neuromuscular plaque, inducing a cleav-
age at the basal level of the epidermis.
In junctional EB the altered molecules are laminin
5 (chain alfa 3, beta 3, gamma 2), collagen XVII and
the integrins alfa 6 and beta 4; the cleavage is at the lev-
el of lamina lucida.
In dermolytic EB the molecular defect involves col-
lagen VII and the cleavage occurs below lamina den-
sa 1 (Figure 1).
From a clinical point of view, the bullous manifes-
tations of epidermolytic EB tend to resolve without
any scars or milia, and generally don’t affect extracu-
taneous areas with the exception of the oral mucosa
(Figure 2-4). Epidermolytic EB is divided into 4 major
subgroups (Table I), each of them with characteristic
aspects (Table II).
In junctional EB blisters resolve with difficulty and
tend to result in atrophic lesions without retractions or
milia; in some subtypes the extracutaneous involvement
(mucosa of the gastroenteric, respiratory and geni-
tourinary tract) could be impressive and give rise to
severe complications which could also provoke death.
Dermolytic EB types are characterized by scarring,
which in time could cause retractions and contrac-
tures of limbs and sometimes also esophageal stric-
tures.2
Until now epidemiological studies about EB have
been mostly incomplete or regard only small samples

360 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY TADINI

Figure 2.—Image showing epidermolysis bullosa of the back. Figure 3.—Infant affected by epidermolysis bullosa.

or some subgroups of the disease, thus giving only

information which is not very indicative.3
The purpose of the Italian EB Registry is to fill this
gap, promoting the development of clinical knowl-
edge on EB by solving the problems related to the
extreme rarity of this disease.
Therefore, the purpose of the register is to:
— centralize important clinical information;
— develop a uniform clinical and pathological
approach to the disease (creation of reference insti-
tutes);
— promote systematic genetical studies (through a
network of laboratories collaborating with the Reg-
istry);
— assess periodically the cases included in the Reg- Figure 4.—Epidermolysis bullosa of the feet.
istry, using standardized methods;
— collaborate with the “rare diseases” study group
of the Superior Institute of Health in order to promote was essential for the realization of this project, which
knowledge on EB and its social and medical impact. made it possible to calculate the incidence and the
The creation of a register for EB was suggested prevalence of EB in the Italian population. In order to
about 10 years ago by the Center for hereditary cuta- get an epidemiologically valid result, in 1992 the Gen-
neous diseases of Milan, because of the need for an odermatoses Project was set up. This project involved
exact estimation of Italian EB patients. more than 500 dermatologists who were not members
The collaboration of the Italian Group for Epi- of the National Health service in order to reach as
demiologic Studies in Dermatology, of many univer- many EB patients as possible. Part of this project was
sity clinics and hospitals and of some dermatologists the distribution of educational material on genoder-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 361

TADINI HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY

TABLE I.—Classification of the main inherited EB.

Major type of EB Major subtype of EB Protein/gene system

EB epidermolytic (EB simplex, EBS) EBS, Weber-Cockayne (EBS-WC) K5, K14

EBS, Koebner (EBS-K) K5, K14
EBS, Dowling-Meara (EBS-DM) K5, K14
EBS with associated muscular dystrophy (EBS-MD) Plectin
EB type “Ogna”(without muscular dystrophy) Plectin
EB junctional (JEB) JEB, Herlitz (JEB-H) Laminin-5
JEB, non-Herlitz (JEB-nH) Laminin-5; collagen XVII
JEB with associated pyloric atresia (JEB -PA) α6β4 integrin+
EB dermolytic (EB dystrophica, DEB) DDEB Collagen VII
RDEB, Hallopeau-Siemens (RDEB-HS) Collagen VII
RDEB, non-Hallopeau -Siemens (RDEB-nHS) Collagen VII

TABLE II.—Classification of rare inherited EB.

Major EB type Major subtype EB Protein/gene system

EB epidermolytic (EB simplex, EBS) EBS “mottled pigmentation” (EBS-MP) K5

EBS autosomal recessive without muscular dystrophy (EBS-AR) K14
EBS superficialis (EBSS) Unknown
EB junctional (JEB) JEB, inversa (JEB-H) Laminin-5
JEB, delayed outbreak (JEB-nH) Unknown
EB dermolytic (EB dystrophica, DEB) DDEB, pretibial (DDEB-Pt) Collagen VII
DEB, transient dermolysis bullosa of newborn (DEB-TBDN) Collagen VII
DDEB or RDEB, “pruriginosa” (DDEB-Pr ; RDEB-Pr) Collagen VII
RDEB, inversa (RDEB-I) Collagen VII
RDEB, centripetalis (RDEB-Ce) Unknown
DEB, AD/AR heterozygote Collagen VII

matoses and the explanation of how to fill in the recruit-
ment cards.
Prospects
The register will give the opportunity to integrate dif-
ferent competencies in the study of EB and, on the
other hand, the possibility to build a representative
cohort of EB cases. The advantages will be as follows:
— estimation of genetic heterogeneity within homo-
geneous diagnostic groups, and study of the relation-
ship between specific mutations, phenotypes and clin-
ical evolution in a specific diagnostic group. The study
of genetic heterogeneity is important for the develop-
ment of prenatal diagnosis;
— possibility to organize studies and therapeutic
trials with a large group of patients, which make these
studies and trials statistically significant;
— possibility to estimate the incidence of some
clinical subgroup in certain geographic areas, studying
the possible differences between these areas and orga-
nizing studies of genetic epidemiology;
— estimate of the incidence of most severe com-
plications and study of their management;
— assessment of the causes of deaths.
The register will help to address the important chal-
lenges offered by EB in all medical fields; from a
methodological point of view it would be possible to
develop new research instruments or to evaluate old
ones (for example, it would be possible to evaluate
the validity of single patient trials in therapeutic ques-
tions).
The register could also promote, together with patient
associations, information campaigns about EB.
Materials and methods
Definition of register
In biomedical research “register” means: system-
atic and continuous collection of information regard-
ing all reported cases of a given disease. There are 2
kinds of registers: population registers and hospital

362 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY
registers. In population registers the main interests are
descriptive epidemiology (assessment of incidence)
and the organization of medical resources available
for a specific disease.
Hospital registers (like that of EB), on the other
hand, assess clinical problems such as the descrip-
tion of the medical evolution of a disease or the def-
inition of its prognosis. These kinds of registers also
allow the creation of a patient group, which is avail-
able for etiological, biological and therapeutic stud-
ies.
The validity of data obtained from a register depends
on the completeness of the information contained: so
the validity of a hospital register depends on the pos-
sibility of containing all the cases observed in the
structure.
If there are registered cases of disease which are of
good quality and with complete information and if
they regard all the cases seen in the past, it is possible
to use them for the register although the information
was collected before the setting up of the registry (ret-
rospective registry). This information is very useful
if someone wants to reconstruct the natural history of
a disease over a long period.
Structure of the registry
The Italian registry of EB, a hospital register, has
been proposed to all Italian centers interested. Dur-
ing the first stage of the project, all the cases known to
the participant centers until 1993 were collected. Dur-
ing the second stage, the register continued its activi-
ty by registering all new cases diagnosed.
The structure of the registry envisages a center for
the coordination of all data. In this center the Derma-
tological Clinic of Bergamo is responsible for the epi-
demiological aspects while the CMCE of Milan con-
trols the clinical aspects and manages the whole pro-
ject, collecting the data coming from all members tak-
ing part in the registry. Four supra-regional centers
refer to this coordination center: CMCE of Milan, Isti-
tuto Dermopatico dell’Immacoloata (IDI) of Rome,
Pediatric Dermatology Unit of the Hospital of Bari
and the Dystrophic Epidermolysis Bullosa Recessiva
Association (DEBRA) of Catania. Their first task has
been to collect all patients present in the territories of
pertinence, which are respectively the North of Italy,
the Center, the South and the Islands. These centers also
had to undertake laboratory assessments if necessary
to confirm the diagnosis of the different EB sub-types.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TADINI
Regarding this, every center had its own laboratory
or collaborated with a laboratory which could per-
form histological, immunohistochemical, ultrastruc-
tural and molecular assays.
Other hospital divisions and clinics joined these
first reference centers in time, and so the collection
web over the territory got bigger and wider.
In the centers a new figure was created: the mon-
itor (who already exists in the GISED). This figure is
a dermatologist with experience in EB, making it
possible to carry out a widespread targeted survey.
The 80 monitors, who took part in the registry gave
easily available and prompt information about the
disease to all ambulatorial dermatologists, called
observers, distributed all over Italy. These observers,
who numbered between 350 and 500 per year, rep-
resented a grassroots web of information over the
whole territory, which has identified and reported to
the nearest monitor all cases compatible with a diag-
nosis of EB.
Patient recruitment
Previous to patient recruitment an informative cam-
paign for dermatologists was carried out: training
courses in genodermatoses, with an average of 400
participants per year, were organized and informative
material about these rare diseases was distributed
(“Project Genodermatoses”).
Moreover, a promotional campaign for the general
population was organized: gadgets were distributed
in the streets of many cities and articles about this der-
matosis were published in different magazines. All
this was done with the help of the Italian Lions Club
organization. Essential both for financial help and for
marketing was the contribution of the TV program
“Telethon”, which has made it possible to collect a
conspicuous amount of money for the creation of many
scholarships.
Instruments for data collection
Report cards were developed where physicians have
to enter the personal data of the patient, the familiar-
ity of the disease, the clinical history (symptoms,
diagnostic procedures and therapies carried out) and,
if possible, the supposed type of EB affecting the
patient.
These report cards were distributed to all members
of the project during the meetings or by mail.
363
TADINI HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY

TABLE III.—Distribution of Registry cases by type and subtype of

EB. Epidermol. EB (192, 28%)
Type/Subtype Frequency %

Epidermolytic EB 192 28
— Koebner 75 10.8
— Weber-Cockaine 67 8.5 Junctional EB
— Dowling-Meara 50 8.2 (67, 10%)
— With muscular dystrophy 1 0.2
Junctional EB 67 10
— Non-Herlitz 46 5.8
— Herlitz 12 2.3
— With pyloric atresia 9 1.6
Dermolytic EB 438 62
— Cockaine-Touraine 159 24.5
— Hallopeau-Siemens 182 29
— Non Hallopeau-Siemens 91 8.8
EB unclassifiable 9 —
N.B.: Unclassified EB are not calculated for incidence and prevalence.
Dermol. EB (438, 62%)

Figure 5.—Total case of inherited epidermolysis bullosa.

TABLE IV.—List of main Clinical Centers involved in the study.
Clinical Centers
— Center for Inherited Cutaneous Diseases Milan
— Istitute Dermopatico dell’Immacolata (IDI) Rome
— DEBRA Italy, Catania
— Pediatric Dermatology Hospital, Bari
— Department of Genetics, University of Brescia
— Observers GISED
— Gaslini Hospital, Genova
— Dermatologic Department, Bologna
— Federico II Hospital, Napoli
Data used for the analysis
To estimate EB prevalence and incidence in our
country, all data contained in the registry regarding
the period 1991-2002 were used.
Diagnosis was based on the concordance with clin-
ical and laboratory criteria described in the literature
and validated in practice all over the world.
Approach to analysis
All EB patients of the registry having adequate data,
entered the survey. The diagnosis of every type of EB
was confirmed with the help of immunofluorescence
assays or electronic microscopy, which shows the
cleavage plane of the blister, with the study of genet-
ic transmission of the disease and, if possible, with
bio-molecular assays.
The patients were divided into the principal types and
subtypes of EB which are described in the new clas-
sification of EB presented by the Consensus Confer-
ence of Chicago in 1999:
— epidermolytic EB: Koebner, Weber-Cockaine,
Dowling Meara, muscular dystrophic-EB;
— junctional EB: Herlitz, not-Herlitz, junctional
with pyloric atresia
— dermolytic EB: Cockaine-Touraine, Hallopeau-
Siemens, not Hallopeau-Siemens.
The incidence is calculated considering the num-
ber of live-born EB patients per 1 million births in the
two-year period 1997-98.
The prevalence is based on all living patients affect-
ed by EB in the year, the last year of the survey.
Results
Total number of cases and contribution of every supra-
regional center
Table III and Table IV show all the cases observed
in 10 years of registry activity and the contribution of
the main supra-regional centers in data collection.
At the end of 2002 the patients affected by EB were
about 697 clearly defined cases and 9 non-classifiable
cases (Figure 5). This last number depends on the fact
that in the study some case reports were considered
which were taken before the project had been found and
which were therefore incomplete.

364 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY TADINI

Muscl. dystr. (1%)

Herlitz (22%)
Pyloric atrophy
(16%)
Dow-Meara
(30%)

Koebner
(38%)

Web-Cock (31%) Non Herlitz (62%)

Figure 6.—Table of data concerning epidermolytic EB. These percentages refer Figure 7.—Table of data concerning the junctional EB. These percentages refer
to total number of epidermolytic EB and not to the total of cases picked up. to the total number of junctional EB and not to total cases collected.

The Italian registry has operated constantly up to

now.4 Non Hallopeau (15%)
The distribution of cases between the different types
and subtypes of EB, according to the up to date clas-
sification,5 is the following:
— 192 cases of epidermolytic EB (28% of the total);
within this group the ratio of different subgroups fluc-
tuates from 8.2% for Dowling Meara to 10.8% for
Koebner (Figure 6); Hallopeau Cockayne
(46%) Touraine
— 67 cases of junctional EB (10%) of the total (Fig- (39%)
ure 7); the non-Herlitz cases represent 5.8% of cases.
It is important to point out that in all 9 cases of junc-
tional EB with pyloric atresia (1.6%) the genetic muta-
tion was identified.
The dermolytic forms represent most of the cases in
the Italian Registry: 438 patients (62%).
Among these the recessive form named Hallopeau- Figure 8.—Table of the data concerning dermolytic EB. These percenta-
Siemens, which represents 29% of all cases, together ges refer to total number of dermolytic EB and not to total cases collected.
with the dominant form of Cockaine-Touraines, which
represents 24.5% of all cases, account for 53.5% of all
is calculated as 20.1 per million live newborns. If we
patients with dermolytic EB (Figure 8).
consider the main groups of EB, we find that the inci-
dence of dermolytic EB is of 12.4 per million. The
Incidence and prevalence incidence of junctional and epidermolytic EB is sim-
In the period 1997-98, taken as a sample recruit- ilar: 3.8 per million.
ment period, 21 newborns affected by EB were In the period 1997-1998, taken as a sample period,
observed. At the same time 1 044 340 healthy new- the prevalence was of 10.1 EB patients per million (in
borns were registered in Italy: so the incidence of EB a population of 57 679 855 people).

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 365

TADINI
Table III shows epidemiologic data using the new
classification.
Discussion
Knowledge of the epidemiology of hereditary EB is
approximate for several reasons:
1) Only few countries have worked on this kind of
studies and among these are Japan,3 Finland,6 North-
ern Ireland,3 South Africa,7 Sweden and Norway,8
Scotland and Great Britain 9 and recently the USA.3
2) The epidemiologic studies considered parts of
the population or only some regions of a country. Some
studies also considered only one subset of EB
3) In general, data are old and/or not recently revised.
All these factors don’t allow us to know the situation
of all the patients with EB in a country accurately.
For example, in Croatia only 56 patients of 50 fam-
ilies were considered suggesting an incidence for EBDr
H-S of 19.2/1milion newborns.10
In Norway, in 1995, the incidence of Epidermolyt-
ic EB was estimated to be 42 cases/1million without
considering the different clinical subtypes of Epider-
molytic EB.
In the USA the prevalence was obtained consider-
ing the data from the American DEBRA: 327 new-
borns with EB from 1986 to 1991 with an incidence of
16.62/1milion newborns,3 these results are not com-
plete because they consider only the severe forms of
EB, excluding milder subtypes.
Later, in the USA, methods for recruitment were
reinforced and an incidence of 19.60/1milion new-
borns was obtained: 10.75 with EBE, 2.86 with EBDd,
EBDr and EBJ 2.04 each.
We note that in Croatia the incidence of EBDr H-S
was 47 times more than in the USA (19.23 vs 0.43)
because of different methods for epidemiologic stud-
ies in different countries.
In the USA in 1990 the prevalence was 2 044 patients
(8.22/1milion). It is very important to consider that our
prevalence data (10.1) are very similar to the data report-
ed in the USA Registry and underline the relevance of
the 2 Registries, even in the presence of a different
socio-sanitary and geographical environment. Never-
theless, the recruitment of patients in the foreign Reg-
istries does not follow an adequate sampling of the pop-
ulation given that the recruitment itself does not refer to
peripheral Centers distributed in the whole territory.
366 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY
In Italy, before the present collection, incidence and
prevalence data were not available and were estimat-
ed on the basis of foreign data.
The Italian Registry enables a widespread sampling
of patients with EB to be carried out, recruiting also the
highest possible number of milder and paucisympto-
matic cases (i.e. familiar mild forms of palmo-plantar
Epidermolytic EB recruited by peripheral- dermatol-
ogists or by the Centers of Military Medicine) that
often in this kind of investigation are underestimated
because of the relative paucity of their symptoms.
Our information reflects the real situation of patients
with EB in Italy and this is very important for social,
clinical and scientific purposes.
Knowing the clinical features of EB and how they are
spread all over Italy, we are able to organize medical
centers for the management of all these patients with
a chronic disease; that’s why the General Institute of
Health promoted this study and now the Italian Reg-
istry of EB is part of the Register of Rare Diseases of
the Ministry of Health.
From a scientific point of view all the data are a rich
store of knowledge about all the features of this pathol-
ogy:
— clinical features and all their variants in every
group or subgroup of EB;
— traditional and molecular diagnostic trials;
— study of recurrent or ancestral mutations in the
Italian population;
— examination of recurrent pathologies, especial-
ly cancer, and management of their prevention, ther-
apy and follow-up.
Finally, in perspective, we can hypothesize a sig-
nificant impact in the strategy of genetic counseling
which should classify EB patients following the mol-
ecular classification in order to find all carriers in the
families of EB affected patients and to perform mol-
ecular prenatal diagnosis when requested in lethal or
very severe forms of EB.
At this point we remember that 70 mutations were
discovered in Italian patients with EBJ and EBD in the
2 main laboratories in Rome (Istituto Dermopatico del-
l’Immacolata) and Brescia (Institute of Genetics).
Mortality rates
Until today it was possible to make a rough calcula-
tion of the mortality rate only in patients of the Center
of Inherited Diseases of Milan representing 44.7% of
all the patients considered in Italy: 18 patients died in
Agosto 2005
HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY
TABELLA V—Mortality of patients at Inherited Cutaneous Disease
Center (Milan).
Number of deaths
(period 1992-2002)
on the basis
of epidermolysis
bullosa subtypes
Epidermolytic — with medical and social organizations.
Junctional 8 EBs are genetically determined diseases with genet-
Dermolytic 10
Total 18
the 13 years of the survey in a group of 263 patients
(mortality rate = 6.8%). Of these, 8 had an EBJ and
10 an EBD (Table V). Of the patients with EBJ, 2 had
EBJ with pyloric atresia and 6 had the variant of Her-
litz, of these 2 died when they were 6 and11 year-old,
and 6 did not pass the first year of life. The most fre-
quent causes of death were infections, especially in
EBJ, and metastasis from squamous cell carcinomas of
the skin ( 8-10 patients with EBD died in adult age). The
last 2 EBD patients died soon after birth due to sepsis.
Complications
A lot of complications have been described in
patients with EB; our data refer only to 2 of the most
frequent and severe of them, squamous cell carcinomas
and esophageal stenosis. Of the 263 patients of the
Center of Milan 7 patients (2.7%) presented squamous
cell carcinomas and 2 of them died due to metastases.
Of these, 6 patients had a squamous cell carcinoma
and only 1 had a basal cell carcinoma.
Twenty-five patients (9.5%) presented an esophageal
stenosis.
Conclusions
The incidence and prevalence data are representative
of the Italian situation given that all the severe cases
have been collected and we think that almost every
case with milder forms of EB have also been identified
using the widespread network of dermatologists
throughout Italy. In fact, no other registry or collection
of patients shows such an accurate or complete sam-
pling as ours. The methods of recruitment of the Ital-
ian Registry allow us to state that just a very small
percentage of cases, of little consequence for the preva-
lence and incidence data, escaped recruitment.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TADINI
The rarity of these genodermatoses and their sever-
ity, considered from the human, medical or social point
of view, forced us to create a structure devoted to the
collection of all data regarding EB. From the epi-
demiological data, we can derive the geographical dis-
tribution of the patients nationally and its relationship
ic variability much greater than clinical and the respec-
tive data are available for anyone interested in these
fields at any level.
We can hypothesize a significant impact in the strat-
egy of prevention, especially for the identification of
the mutations in all EB families in order to find carri-
ers and avoid consanguineous marriages. Compari-
son with the epidemiological data of other Registries
may be of great utility. Finally, the medical staff of
the Italian Registry of EB promotes a European Reg-
istry of Epidermolyss Bullosa.
Acknowledgments.—The compilation of the Italian Registry of Hered-
itary Epidermolysis has been made possible by the contribution of about
600 Italian dermatologists that worked together giving their time and
experience to this project. We thank all contributors, wishing to have them
again as companions in the future projects of other epidemiological stud-
ies on genodermatosis.
References
1. Lin AN, Carter DM. Epidermolysis bullosa: basic and clinical aspects.
New York: Springer; 1992.
2. Tadini G, Brusasco A, Cambiaghi S, Camozzi S, Cavalli R, Restano
L. Epidermolisi bollose ereditarie - Ittiosi. Milano: EdiSES; 1995.
3. Fine J-D, Bauer EA, Mc Guire J, Moshell A. Epidermolysis bullosa:
clinical, epidemiologic and laboratory advances, and the findings of
the National Epidermolysis Bullosa Registry. Baltimore: John Hop-
kins University Press; 1999:101-13.
4. Tadini G, Naldi L, Locati L, Cainelli T. Epidemiological survey on epi-
dermolysis bullosa in Italy. J Invest Dermatol 1994;103:853.
5. Fine J-D, Eady RAY, Bauer EA, Briggman RA, Bruckner-Tuder-
man L, Christiano A et al. Revised classification system for inher-
ited epidermolysis bullosa Report of the Second International Con-
sensus Meeting on Diagnosis and Classification of Epidermoly-
sis Bullosa. Special report. J Am Acad Dermatol June 2000;42:
1051-66.
6. Kero M. Occurrence of epidermolysis bullosa in Finland. Acta Derm
Venereol (Stockh) 1984;64:57-62.
7. Winship IM. Epidermolysis bullosa in South Africa: formation of a
National Registry. Epidermolysis Bullosa. A comprehensive review
of classification management and laboratory studies.Crowthorne,
Berkshire, UK: DEBRA; 1990. p.134-6.
8. Gedde-Dahl Jr T. Epidermolysis bullosa. A clinical, genetic and epi-
demiological study. Universitets Forlaget (Oslo). Baltimore: The John
Hopkin Press; 1971.
9. Davison BCC. Epidermolysis bullosa. J Med Genet 1965;2:233-42.
10. Pavicic Z, Kmet-Vizintin P, Kansky A, Dobric I. Occurrence of hered-
itary bullous epidermolysis in Croatia. Pediatric Dermatol 1990;
7:108-10.
367
TADINI
Il registro italiano delle epidermolisi bollose ereditarie
L e epidermolisi bollose (EB) ereditarie costituiscono un
gruppo eterogeneo di patologie meccanobollose gene-
ticamente determinate.
Sono dermatosi rare nella popolazione generale, ma di
grande rilevanza clinica, in quanto possono determinare una
notevole riduzione della qualità di vita e dell’aspettativa di
vita dei pazienti affetti.
Le peculiarità cliniche comuni alle EB sono rappresenta-
te dalla marcata fragilità cutanea e mucosa, per cui anche il
minimo trauma determina l’insorgenza di lesioni bollose.
Gli annessi sono coinvolti in modo più o meno marcato,
risultando in alcuni casi mancanti.
Si distinguono EB epidermolitiche, giunzionali e dermo-
litiche in base al sito di clivaggio nella regione della giunzione
dermo-epidermica, evidenziabile con tecniche immunoi-
stochimiche (immunofluorescenza diretta) e con la micro-
scopia elettronica. Nelle EB epidermolitiche il difetto mole-
colare interessa le citocheratine K5 e K14 e la pectina, una
proteina non collagenica presente anche nelle giunzioni neu-
romuscolari, determinando il clivaggio a livello della mem-
brana basale dei cheratinociti; le EB giunzionali conseguo-
no all’alterazione genetica della laminina 5 (catene α3, β3 e
γ2), del gene del collagene XVII e delle integrine α6 e β4, e
le dermolitiche del collagene VII, con conseguente separa-
zione a livello della lamina lucida e al di sotto della lamina
densa, rispettivamente 1 (Figura 1). Dal punto di vista clini-
co le lesioni bollose di tipo epidermolitico tendono a guari-
re rapidamente senza esiti cicatriziali o grani di «milium»,
generalmente senza coinvolgere sedi extracutanee, se non
la mucosa orale (Figure 2-4): all’interno di questo gruppo esi-
stono comunque 4 varianti principali (Tabella I), ciascuna con
aspetti peculiari (Tabella II). Nelle EB giunzionali le bolle rie-
pitelizzano con difficoltà, ed esitano in lesioni atrofiche, ma
senza cicatrici retraenti o grani di milium; in alcune varian-
ti il coinvolgimento extracutaneo (mucose degli apparati
gastroenterico, respiratorio, genitourinario) può essere impo-
nente, con complicanze sistemiche anche molto gravi o leta-
li. Tipici delle forme dermolitiche sono gli esiti cicatriziali,
la cui continua formazione può provocare nel tempo retrazioni
e contratture degli arti interessati o restringimenti esofagei 2.
Le stime epidemiologiche sulle EB condotte fino a que-
sto momento sono scarse e incomplete, o comunque relati-
ve solo a piccoli campioni o a poche varianti della malattia,
fornendo perciò dati parziali e poco significativi 3.
Obiettivo generale della costituzione di un Registro italiano
delle EB è quello di favorire lo sviluppo delle conoscenze cli-
niche sulle EB superando i problemi connessi con la rarità di
tali patologie. A tal fine il Registro prevede:
— la centralizzazione delle informazioni cliniche rile-
vanti;
— lo sviluppo di un inquadramento clinico-patologico
uniforme e riproducibile (revisione esperta di casi, istitu-
zione di Centri di riferimento clinico all’interno del Registro);
368 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY
— la conduzione di indagini genetiche sistematiche (attra-
verso una rete di laboratori di biologia molecolare che col-
laborano al registro);
— la valutazione clinica periodica dei casi inseriti nel
Registro, secondo modalità il più possibile standardizzate;
— l’adesione al gruppo di studio dell’Istituto Superiore di
Sanità sulle malattie rare per portare a conoscenza delle Isti-
tuzioni preposte la realtà delle epidermolisi Bollose e il loro
impatto socio-sanitario.
La proposta di creazione di un Registro delle Epidermo-
lisi Bollose Ereditarie è stata fatta circa 10 anni fa dal Cen-
tro Malattie Cutanee Ereditarie di Milano, in seguito alla
necessità di avere una stima quanto più vicina al vero della
casistica italiana dei pazienti affetti da EB.
La collaborazione con il Gruppo Italiano Studi Epide-
miologici in Dermatologia (GISED), con numerosi centri
universitari e ospedalieri, e con dermatologi ambulatoriali ha
consentito la realizzazione di tale progetto, rendendo dispo-
nibili i valori di incidenza e prevalenza nella popolazione
italiana. Per ottenere un risultato epidemiologicamente vali-
do nel 1992 è stato organizzato il «Progetto Genodermato-
si» che ha consentito di programmare il coinvolgimento di più
di 500 dermatologi ambulatoriali esterni del Sistema Sani-
tario Nazionale, nell’intento di allargare la base di recluta-
mento dei pazienti con EB. Il Progetto veniva completato
con una parte didattica sulle Genodermatosi in generale,
compresa la fornitura di materiale didattico e l’informazio-
ne sulla compilazione delle schede di arruolamento dei
pazienti affetti da EB.
Prospettive
Il Registro offrirà la possibilità di integrare differenti com-
petenze nello studio delle EB e permetterà di costruire una
coorte rappresentativa dei casi di EB. I vantaggi sono evidenti;
sarà infatti possibile:
— valutare la presenza di un’eterogeneità genetica all’in-
terno di categorie diagnostiche omogenee, studiando la rela-
zione tra specifiche mutazioni, fenotipo e storia evolutiva
delle singole entità diagnostiche (prognosi). La documenta-
zione di un’eterogeneità genetica ha particolare importanza
per lo sviluppo di strumenti di diagnosi prenatale;
— avviare studi e trials terapeutici su una larga base di
pazienti, con disegni formali che garantiscano validità e
riproducibilità dei risultati e potere statistico sufficiente;
— stimare in aree campione l’incidenza delle varietà cli-
nico-patologiche più frequenti, valutando eventuali diffe-
renze tra aree geografiche ed avviando studi di epidemiolo-
gia genetica;
— analizzare l’incidenza delle complicazioni più severe
e il loro conseguente trattamento;
— studiare le cause dei decessi.
Agosto 2005
HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY
Il Registro potrà fungere da catalizzatore per affrontare pro-
blemi complessi all’interfaccia tra differenti discipline; da un
punto di vista metodologico potranno essere sviluppati stru-
menti di ricerca originali o potranno essere valutati stru-
menti esistenti (ad esempio, valutazione della validità di
disegni del tipo «single patient trial» in ambito terapeuti-
co). Il Registro potrà inoltre promuovere, in collaborazione
con le Associazioni dei pazienti, un’attività di informazione
e documentazione sulle EB.
Materiali e metodi
Definizione di registro
Si definisce «registro», nella ricerca biomedica, la rac-
colta sistematica e continua di alcune informazioni su tutti i
casi notificabili di una malattia. Si possono distinguere regi-
stri di popolazione e registri ospedalieri. Nei registri di popo-
lazione prevalgono gli interessi di epidemiologia descrittiva
(stime di incidenza nel tempo) e di pianificazione delle risor-
se sanitarie disponibili per la patologia in esame. I registri
ospedalieri, come è quello delle EB, sono invece maggior-
mente orientati alla valutazione di problemi clinici come la
descrizione della storia evolutiva della malattia e la definizione
della prognosi, permettendo inoltre la costituzione di una
base di pazienti per studi eziologici, biologici (banche bio-
logiche) o di terapia. L’interesse e la riproducibilità delle
informazioni ottenute in un registro dipendono dalla com-
pletezza della registrazione, per cui la validità di un registro
ospedaliero dipende dalla possibilità di segnalare tutti i casi
osservati presso l’istituzione.
Tuttavia, quando sia garantita una buona qualità e com-
pletezza delle informazioni retrospettive e sia possibile recu-
perare informazioni su tutti i casi di malattia, è possibile uti-
lizzare anche i casi diagnosticati in un definito periodo pre-
cedente l’avvio del Registro (registro retrospettivo). L’utilizzo
di tali informazioni da tali casi è particolarmente utile quan-
do si voglia ricostruire la storia naturale della patologia in esa-
me su di un lungo periodo di tempo (studi di coorte in par-
te retrospettivi).
Struttura del Registro
Il Registro Italiano delle EB è stato proposto come regi-
stro su base ospedaliera a tutti i centri italiani interessati. In
una prima fase sono stati raccolti tutti i casi noti ai Centri par-
tecipanti fino al 1993. Successivamente il Registro ha pro-
seguito la sua attività attraverso la segnalazione di tutti i
nuovi casi diagnosticati (casi incidenti).
La struttura del Registro prevede la presenza di un centro
di coordinamento dei dati, gestito dalla Clinica Dermatolo-
gica di Bergamo per la parte epidemiologica e dal CMCE di
Milano per la parte clinica, col compito di gestire e control-
lare l’intero progetto e raccogliere tutti i dati forniti dai diver-
si componenti del progetto stesso. A questo fanno riferi-
mento 4 centri sovraregionali: il Centro Malattie Cutanee
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TADINI
Ereditarie di Milano, l’Istituto Dermopatico dell’Immacolata
(IDI) di Roma, la Cattedra di Dermatologia Pediatrica del-
l’Ospedale di Bari e la Dystrophic Epidermolysis Bullosa
Recessiva Association (DEBRA) Italia con sede a Catania.
La loro funzione primaria è stata quella di individuare e rac-
cogliere i pazienti provenienti dalle aree del territorio ita-
liano di loro competenza, rispettivamente il Nord, il Cen-
tro, e il Sud peninsulare e insulare. Questi centri dovevano
inoltre ricorrere, quando indicato, ad appropriate indagini
di laboratorio per la conferma diagnostica del tipo di EB
incontrata. A questo proposito, si sottolinea che ogni centro
di riferimento era in contatto o disponeva di un laboratorio
di analisi per l’esecuzione di indagini istologiche, immu-
noistochimiche, ultrastrutturali e molecolari.
A queste strutture di riferimento si sono poi affiancati
divisioni ospedaliere e cliniche universitarie, ampliando così
la rete di arruolamento sul territorio.
Nell’ambito di questi centri è stata istituita la figura del
«monitor» (già operanti nella rete GISED), cioè quella di
uno specialista dermatologo, che, dimostrando una cono-
scenza specifica di queste patologie, ha permesso di racco-
gliere una casistica ampia e selezionata. Gli 80 monitor che
hanno preso parte al registro hanno inoltre fornito una con-
sulenza pronta e facilmente raggiungibile ai colleghi der-
matologi ambulatoriali, denominati «observers» e distribui-
ti sul territorio italiano. Questi, in numero variabile da 350
a 500 per anno, hanno costituito una rete informativa capil-
lare su tutto il territorio, identificando e segnalando al più vici-
no «monitor» o centro clinico regionale tutti i casi certi o
compatibili con diagnosi di EB.
Reclutamento dei pazienti
Il reclutamento dei pazienti è avvenuto mediante una cam-
pagna informativa rivolta ai dermatologi, nell’ambito della
quale sono state organizzate riunioni di aggiornamento cli-
nico, che hanno raccolto in media 400 partecipanti per anno,
ed è stato stampato materiale didattico in cui erano descrit-
te le più importanti caratteristiche di queste rare malattie
(Progetto Genodermatosi).
Inoltre è stata avviata una campagna promozionale anche
nei confronti della popolazione generale, mediante la distri-
buzione di gadgets nelle piazze e la pubblicazione di artico-
li informativi sui diversi quotidiani, organizzata in coopera-
zione con il Multidistretto Lions Clubs Italia. Fondamenta-
le, sia in termini economici che di risonanza pubblica, è sta-
to anche il contributo derivato dalla trasmissione televisiva
«Telethon», che ha permesso di raccogliere una cospicua
quota di fondi utilizzati per il finanziamento di borse di ricer-
ca.
Strumenti per la raccolta dei dati
Sono state sviluppate delle schede informative in cui veni-
vano richiesti i dati anagrafici del paziente, la familiarità per
tali malattie, la storia clinica (sintomi riferiti, indagini dia-
gnostiche eseguite e terapie intraprese) e, ove possibile, una
diagnosi orientativa per il tipo di EB espressa dal probando.
369
TADINI
Le schede sono state distribuite a tutti i collaboratori del
progetto, o durante gli incontri di aggiornamento o median-
te mezzo postale a chi ne facesse richiesta.
Dati utilizzati per l’analisi
Per stimare la prevalenza e l’incidenza delle EB nel nostro
paese, le analisi sono state eseguite su tutti i dati disponibi-
li del Registro italiano dal 1991 al 2002.
Le diagnosi si sono basate sulla concordanza con specifici
criteri clinici e di laboratori descritte in letterature e ampia-
mente validate dall’utilizzo in tutto il mondo.
Approccio all’analisi
Lo studio di popolazione si è basato su tutti i pazienti
affetti da EB facenti parte del Registro, sui quali fossero
disponibili dati informativi adeguati.
La diagnosi di ciascuna categoria principale di EB è sta-
ta confermata sulla base della sede di formazione del distac-
co dermo-epidermico identificabile a livello ultrastrutturale
con l’immunofluorescenza e la microscopia elettronica, e
sulla trasmissione genetica della malattia e, dove effettuata,
anche sulla diagnosi di biologia molecolare (mutazione).
I pazienti sono stati progressivamente suddivisi nei prin-
cipali tipi e sottotipi di EB di seguito riportati seguendo la
nuova classificazione per le EB redatta dalla Consensus Con-
ference di Chicago del 1999:
— EB epidermolitiche: Koebner, Weber-Cockaine, Dow-
ling-Meara, epidermolitiche con distrofia muscolare associata.
— EB giunzionali: non-Herlitz, Herlitz, giunzionale con
atresia pilorica.
— EB dermolitiche: Cockaine-Touraine, Hallopeau-Sie-
mens, Non-Hallopeau-Siemens.
L’incidenza viene calcolata in base al numero dei nati vivi
affetti da EB per 1 000 000 nati in Italia nel biennio 1997-98,
presi come biennio campione.
Il calcolo della prevalenza si è basato sulla determinazio-
ne di tutti pazienti vivi affetti da EB nel corso dell’ ultimo
anno dello studio, preso come anno campione.
Risultati
Casistica totale raccolta e contributo di ciascun centro sovra-
regionale
Nelle Tabelle III-IV viene segnalato il totale dei casi accer-
tati nei 10 anni di attività del Registro e il contributo dei
principali Centri Clinici Sovraregionali nella raccolta della
casistica.
Alla fine del 2002 la stima dei pazienti colpiti da EB
ammontava a 697 casi accertati e 9 casi di EB non ulterior-
mente classificabili (Figura 5). Quest’ultimo valore è legato
al fatto che sono state esaminate anche segnalazioni antece-
denti al progetto, le quali presentavano solo informazioni
parziali, tali da non poter meglio inquadrare questi pazienti.
370 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY
Il registro italiano rimane comunque, costantemente ope-
rativo 4.
La distribuzione per tipo e sottotipo di EB, basata sulla più
recente classificazione di tali genodermatosi 5, evidenzia
per le forme epidermolitiche 192 casi, circa il 28% del tota-
le; all’interno di questo raggruppamento i valori percentua-
li dei diversi sottotipi di EB oscillano tra 8,2% per la Dow-
ling-Meara e 10,8% per la forma generalizzata di Koebner
(Figura 6). Le EB giunzionali hanno formato una casistica di
67 pazienti, pari al 10% del totale (Figura 7); la variante
non-Herlitz costituisce il 5,8% dei casi. È importante segna-
lare che nei 9 casi (1,6%) di EB giunzionale con atresia pilo-
rica, è stata individuata la mutazione corrispondente.
Le forme dermolitiche costituiscono la parte preponderante
della casistica del registro italiano con 438 casi, pari al 62%
del totale. In quest’ambito la variante recessiva grave di Hal-
lopeau-Siemens con il 29% dei casi sommata a quella domi-
nante di Cockaine-Touraine con il 24,5%, costituiscono ben
il 53,5% di tutti i casi (Figura 8).
Incidenza e prevalenza
Nel corso del biennio 1997-98, preso come biennio cam-
pione, sono stati segnalati 21 nuovi casi di bambini nati vivi
e affetti da EB. Nello stesso periodo i nati vivi e sani ammon-
tavano nel nostro paese a 1 044 340. Il valore di incidenza che
ne è derivato è di 20,1 nati affetti per milione di nati vivi. Se
si considerano i principali gruppi di EB, troviamo che le for-
me dermolitiche in questi due anni mostravano un’inciden-
za pari a 12,4 nati affetti per 1 000 000 nuovi nati. Per le
giunzionali e le epidermolitiche il valore di incidenza è lo stes-
so ed è pari a 3,8.
La prevalenza, calcolata al 31 dicembre 1998, era pari a
10,1 pazienti affetti per milione di abitanti (calcolata su una
popolazione di 57 679 855).
Applicando la nuova classificazione, l’indagine epide-
miologica definitiva fornisce i dati raccolti nella Tabella III.
Discussione
Le conoscenze sugli aspetti epidemiologici delle epider-
molisi bollose ereditarie sono piuttosto approssimative per una
serie di differenti ragioni. La prima risiede nel fatto che solo
pochi paesi si sono impegnati in questi tipi di ricerca: tra que-
sti, Giappone 3, Finlandia 6, Nord Irlanda 3, Sud Africa 7,
Svezia e Norvegia 8, Scozia e Gran Bretagna 9 e recente-
mente anche gli Stati Uniti 3. La seconda ragione è legata al
fatto che tali indagini epidemiologiche sono state fatte su
campioni parziali di popolazione o prendendo in considera-
zione solo alcune regioni di uno stato. In alcuni inoltre, è
stato considerato solo un determinato sottogruppo di epi-
dermolisi bollose ereditarie, escludendo tutte le altre varian-
ti. La terza e ultima ragione è che si tratta di stime non recen-
ti o comunque non più aggiornate.
Questi fattori, combinati fra loro, non consentono di sta-
bilire una stima rappresentativa della reale situazione dei
Agosto 2005
HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY
pazienti affetti da EB in quel determinato paese. Per esem-
pio, in Croazia sono stati esaminati solo 58 pazienti di 50
famiglie, suggerendo un’incidenza per la forma dermolitica
recessiva di Hallopeau-Siemens di 19,2 affetti per milione di
nati vivi 10. In Norvegia, nel 1995, è stata calcolata un’inci-
denza delle EB epidermolitiche pari a circa 42 casi per milio-
ne, non considerando però i diversi sottotipi della forma
stessa.
Negli Stati Uniti la prevalenza è stata calcolata basando-
si su dati forniti dalla DEBRA americana: è stato infatti
riportato un totale di 327 nati affetti da EB nel periodo 1986-
1990, con un’incidenza pari a 16,62 nati per milione di nasci-
te 3. In realtà questi dati sono relativi a una maggioranza di
casi di pazienti affetti dalle forme più severe, mentre gran par-
te dei casi meno severi non erano noti al progetto. I redatto-
ri del Registro USA, nel tentativo di fornire una stima più
accurata, hanno perciò migliorato negli anni successivi le
metodiche di reclutamento e di sviluppo di centri regionali
di riferimento. Usando questo approccio, l’incidenza com-
plessiva di EB è risultata pari a 19,60 nuovi casi per milio-
ne di nascite, con una quota maggiore di EB epidermolitiche
(10,75), seguite dalla EB dermolitica dominante (2,86), dal-
la recessiva e dalla EB giunzionale (2,04 ciascuna). È inte-
ressante notare che l’incidenza stimata per la EB dermoliti-
ca tipo Hallopeau-Siemens in Croazia sia all’incirca 47 vol-
te (19,23 vs 0,43) più alta di quella osservata negli Stati Uni-
ti e ciò è spiegato proprio dai differenti sistemi di arruola-
mento dei pazienti e della precisione di tale raccolta.
La prevalenza complessiva di EB ereditaria negli Stati
Uniti nel 1990 è stata stimata essere 2 044 pazienti, corri-
spondenti ad un tasso di prevalenza di 8,22 casi di EB per
milione di abitanti. È importante sottolineare che la nostra
casistica e il nostro dato assoluto di prevalenza (10,1) sono
molto simili a quelli riferiti dal Registro statunitense e riba-
disce la validità di entrambi gli approcci fatte salve le diffe-
renze sociali socio-sanitarie e geografiche. Si fa notare che
comunque la raccolta della casistica finora utilizzata all’estero
non segue un campionamento adeguato della popolazione, in
quanto la raccolta dei casi non è stata affidata a centri di
riferimento periferici ben inseriti in ogni regione del territorio.
Nel nostro Paese, finora, i dati relativi all’incidenza o pre-
valenza di queste genodermatosi non erano disponibili, ma
erano desunti in modo proporzionale da quelli degli altri
Paesi.
Il Registro Italiano con la sua organizzazione ha permes-
so un campionamento sul territorio quasi in modo capillare
dei pazienti affetti da EB, identificando anche un numero
significativo di casi molto modesti di EB epidermolitiche
(forme familiari fruste di EB epidermolitiche palmo-plantari
individuate dai dermatologi del SSN a cui i pazienti si rife-
rivano per altre patologie oppure diagnosticate nei Centri di
Medicina Militare) che spesso, in questo tipo di indagini,
non riescono a essere individuati proprio per la esiguità del-
le manifestazioni cliniche.
La disponibilitá di informazioni sufficientemente rappre-
sentative della reale situazione dei pazienti affetti da EB nel
nostro paese è estremamente importante sia da un punto di
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TADINI
vista medico-sanitario sia per gli aspetti prettamente scien-
tifici.
Nel primo caso, conoscendo le caratteristiche cliniche
delle EB, e la loro distribuzione nel territorio si possono
organizzare centri sanitari con strutture congrue alla gestio-
ne di questi malati che sono portatori di una patologia cro-
nica. Proprio per questa necessitá l’Istituto Superiore di
Sanitá, ha stimolato lo sviluppo di tale indagine epidemio-
logica e a oggi i dati del Registro Italiano delle EB fanno
parte del Registro Italiano delle Malattie Rare del Ministe-
ro della Salute.
Da un punto di vista scientifico la casistica raccolta rap-
presenta un grosso bagaglio informazionale su tutti gli aspet-
ti della patologia:
— Aspetti clinici e loro varianti all’interno dello stesso
gruppo o sottotipo.
— Aspetti diagnostici tradizionali e molecolari.
— Studio delle mutazioni ricorrenti ed ancestrali della
popolazione italiana.
— Valutazione delle patologie ricorrenti nei pazienti con
particolare riguardo ai tumori ed alla loro prevenzione e tera-
pia.
Si prospetta infine un impatto significativo nella strategia
della consulenza genetica la quale prevede di disporre del-
la classificazione molecolare dei soggetti affetti per indivi-
duare i portatori fra loro consanguinei e gli incroci a rischio,
e infine per rendere possibile la diagnosi prenatale moleco-
lare per ciascuna delle famiglie a rischio per EB letali o gra-
vemente invalidanti A questo proposito ricordiamo che sono
state individuate circa 70 mutazioni nei pazienti italiani affet-
ti da epidermolisi bollosa giunzionale e dermolitica, nei 2
laboratori di riferimento italiani di Roma (Istituto Dermo-
patico dell’Immacolata e di Brescia, Dipartimento di Gene-
tica Medica dell’Università di Brescia).
Dati di mortalità
Fino ad oggi è stato possibile effettuare una stima dei dati
di mortalità esclusivamente sui pazienti afferenti al Centro
Malattie Cutanee Ereditarie di Milano, che costituiscono il
44,7% della casistica complessiva: su 263 pazienti, sono
stati registrati 18 decessi (tasso di mortalità = 6,8%), di cui
8 pazienti erano affetti da EB giunzionale e 10 da EB der-
molitica (Tabella V); non si sono verificati decessi di pazien-
ti con EB epidermolitica. Dei pazienti con EB giunzionale,
2 erano affetti dalla variante associata ad atresia del piloro e
6 dalla variante di Herlitz: tra questi ultimi, 2 pazienti sono
deceduti tra i 6 e gli 11 anni, mentre 6 non hanno superato
l’anno di vita. Tra le più frequenti cause di morte si segna-
lano le complicanze infettive (soprattutto nelle varianti giun-
zionali) e le metastasi a partenza da carcinomi spinocellulari
(presenti in 8 dei 10 soggetti con EB dermolitica, variante
Hallopeau-Siemens, deceduti entrambi in età adulta; gli altri
2 pazienti affetti da EB dermolitica sono deceduti poco dopo
la nascita per sepsi). I dati che si riferiscono ai decessi rife-
riti dagli altri Centri di reclutamento saranno elaborati ed
371
TADINI
esposti in successive comunicazioni di aggiornamento del
Registro.
Complicanze
Sono state descritte numerosissime complicanze nei
pazienti affetti da EB ereditarie: i dati disponibili sono rela-
tivi a due delle più frequenti e severe (sviluppo di carcinomi,
stenosi esofagea) e rappresentano quindi solo una stima par-
ziale rispetto alla loro reale incidenza.
Nella casistica dei pazienti del Centro Malattie Cutanee
Ereditarie di Milano, 7 pazienti su 263 (pari al 2,7%) hanno
sviluppato carcinomi cutanei: tra questi, 2 sono deceduti in
seguito allo sviluppo di metastasi. In 6 casi si trattava di car-
cinomi spinocellulari; in un solo paziente era presente un
carcinoma basocellulare.
Inoltre, 25 pazienti tra quelli afferenti al CMCE di Mila-
no (N=263) erano affetti da stenosi esofagea, ovvero una
percentuale pari al 9,5%.
Anche per le complicanze, i dati di riferimento degli altri
Centri sono in via di rielaborazione.
Conclusioni
Le stime di in incidenza e prevalenza sopra riportate sono
rappresentative della reale situazione italiana per una serie di
motivi. In primo luogo, si ritiene che tutti i casi più gravi
siano stati raccolti e conteggiati e, in secondo luogo, anche
i casi relativi alle forme di EB meno gravi sono stati quasi
completamente individuati grazie alla rete informativa capil-
lare creata sul territorio. Infatti, nessun altro registro ha potu-
to disporre di un campionamento così completo; in questo
modo solo una percentuale minima, non in grado di altera-
re significativamente il valore di incidenza, sarebbe sfuggi-
ta alla raccolta.
Comunque le cause di questo sia pure modesto «bias»
risiedono nel livello socio-culturale dei soggetti affetti e dei
familiari, dalla esiguità del quadro clinico, dalla inconsape-
volezza di essere affetto da EB e da ultimo dalla mancata dia-
gnosi.
La rarità di tali Genodermatosi e la loro gravità, intesa
sia da un punto di vista umano sia sanitario ed economico,
rendeva obbligatoria l’istituzione di una struttura atta a rac-
cogliere tutte le informazioni possibili su queste materie.
Dalla disponibilità dei dati epidemiologici sopra esposti
si può evidentemente derivare la distribuzione geografica
della malattia sul territorio nazionale e la sua interferenza nel-
la organizzazione degli interventi clinici assistenziali.
Le EB sono malattie geneticamente determinate e con
372 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
HEREDITARY EPIDERMOLYSIS BULLOSA ITALIAN REGISTRY
una variabilità genetica ben più ampia di quella clinica; dati
di qualunque natura così ottenuti, sono disponibili a chiun-
que ne sia a qualsiasi titolo interessato. Si prospetta inoltre
un impatto significativo nella strategia della prevenzione, la
quale prevede di disporre della classificazione molecolare dei
soggetti affetti per individuare i portatori fra loro consan-
guinei, per individuare i matrimoni a rischio e, infine, per ren-
dere possibile la diagnosi molecolare di tutte le famiglie ita-
liane affette.
In virtù di questa prerogativa, i dati ottenuti potranno esse-
re confrontati e scambiati con quelli di altri Paesi. Lo staff
medico del Registro italiano ha proposto infatti in numero-
si sedi congressuali e istituzionali la creazione di un Registro
Europeo delle Epidermolisi Bollose Ereditarie
Riassunto
Obiettivo. Attualmente in Italia non esiste uno studio epi-
demilogico completo sulle Epidermolisi Bollose (EB) ere-
ditarie: la necessità di avere una stima accurata della casisti-
ca italiana ha favorito la costituzione di un Registro nazionale,
che ha permesso di raccogliere tutti i casi notificabili di malat-
tia, con importanti implicazioni nell’ambito delle conoscen-
ze cliniche dell’EB, per lo sviluppo di strumenti di diagnosi
prenatale e per l’avvio di studi di epidemiologia genetica.
Metodi. Abbiamo costituito un registro su base ospedaliera
che, in una prima fase, ha raccolto i casi già noti ai centri par-
tecipanti dal 1985 al 1993, e, successivamente, ha prose-
guito con la segnalazione dei nuovi casi incidenti, fino al 31
Dicembre 2002. Il Registro prevedeva la presenza di un cen-
tro di coordinamento dati (Clinica Dermatologica di Berga-
mo), di 3 centri sovraregionali (CMCE di Milano, IDI di
Roma, Ospedale di Bari) che hanno raccolto i pazienti pro-
venienti rispettivamente dal Nord, Centro e Sud del Paese e
della DEBRA Italia.
Risultati. Sono stati notificati in tutto 697 casi e 9 casi
non ulteriormente classificabili, costituiti da un 28% di EB
epidermolitiche, 10% di EB giunzionali e 62% di EB der-
molitiche. L’incidenza delle EB al 31 Dicembre 2002 era di
0,1 pazienti affetti per milione di abitanti; la prevalenza, cal-
colata al 31 dicembre 2002, era pari a 10,1 pazienti affetti per
milione di abitanti.
Conclusioni. Tali stime epidemiologiche sono rappre-
sentative della situazione italiana e tracciano una distribuzione
geografica della malattia sul nostro territorio, con un impat-
to significativo nella strategia della prevenzione.
PAROLE CHIAVE: Epidermolisi Bollosa - Registro - Inciden-
za - Relevanza.
Agosto 2005
 G ITAL DERMATOL VENEREOL 2005;140:373-9

Methodological procedure for evaluation

of risk factors for cutaneous malignant melanoma
in a representative sample of the Tuscan population
P. RUBEGNI 1, P. SBANO 1, G. CEVENINI 2, M. RISULO 1, E. STANGHELLINI 1, P. BARBINI 2
M. R. MASSAI 2, L. ANDREASSI 1, M. FIMIANI 1

Aim. The incidence of melanoma is rising steadily in countries
with white populations and, despite all attempts at treatment,
a considerable proportion of patients with malignant melanoma
die of the disease. Primary prevention is, therefore, important
to reduce mortality at present. Here we report the results of a
case-control study of melanoma risk factors conducted in Tus-
cany (Italy) in 2002-2003.
Methods. One-hundred-forty Italian subjects who underwent
surgical exeresis of non familial cutaneous malignant melanoma
and 280 age- and gender-matched controls filled in a stan-
dardized questionnaire about occupational and recreational
sun exposure, underwent complete skin examination by a der-
matologist to assess the number of nevi and presence of clini-
cally atypical nevi, eye color, hair color and Fitzpatrick pho-
totype. Moreover skin color was measured with a Minolta CR-
300 colorimeter. Univariate and stepwise logistic regression
statistical analysis were performed to analyze differences in
variables between melanoma patients and control subjects.
Results. We demonstrated a highly significant difference
between controls and melanoma patients in our study: in nevi
number and presence of atypical nevi, constitutional skin col-
or (Fitzpatrick phototype was completely explained by colori-
metric variables of skin color) and eye color.
Conclusion. We agree with what was recently proposed by oth-
ers that objective skin color measurements must be combined
with phenotype parameters and sun exposure history for exact
assessment of individual risk
KEY WORDS: Cutaneous melanoma - Risk factors - Colorimeter -
Skin color.
1Unit of Dermatology, Department of Clinical Medicine and
Immunological Science
University of Siena, Siena, Italy
2Department of Surgery and Bioengineering
University of Siena, Siena, Italy
T he incidence of melanoma is rising steadily in
countries with white populations and, despite all
attempts at treatment, a considerable proportion of
patients with malignant melanoma (MM) die of the
disease.1-4 Early diagnosis and primary prevention are,
therefore, the only way to reduce mortality at present.
Many advances in early diagnosis have been made
through innovative non invasive techniques, such as
digital dermoscopy analysis (DDA), which have shown
the importance of objective numerical parameter mea-
surements.5 Less progress seems to have been made
with primary prevention, despite the fact that instru-
ments for objective numerical evaluation of impor-
tant phenotypic characteristics are now available.
Features known to be correlated with melanoma
risk include eye color, hair color, tanning ability, freck-
ling, nevus number, skin phototype and skin colour.6,
7 Skin color and phototype characterise skin biotype.8

Received: May 10, 2005 .

Accepted for publication: July 1, 2005.
Address reprint requests to: P. Rubegni, MD, Istituto di Scienze Der-
matologiche, Università degli Studi di Siena, Policlinico Le Scotte, 53100
Siena, Italy. E-mail: rubegni@unisi.it.
In an extensive review of case-control studies, Evans
et al. noted that 6 out of 9 studies demonstrated that the
relative risk of melanoma was 2-18 fold higher for
fair or pale complexions than for non fair complex-
ions.9 However, in almost all these studies, skin color
was evaluated by a visual score and was therefore sub-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 373

RUBEGNI METHODOLOGICAL PROCEDURE FOR EVALUATION OPF RISK FACTORS FOR CUTANEOUS MALIGNANT MELANOMA

jective, non reproducible and non quantifiable. The TABLE I.—Frequency counts of categorical risk factors together with
increasing availability of colorimeters now makes it statistical significances (HS=highly significant, P<0.01; S=signifi-
cant, P<0.05; NS=not significant, P>0.05) of χ2 test, Fisher exact test
possible to objectively evaluate skin color in an easy and Spearman correlation analysis applied to contingency tables.
and largely reproducible manner.10, 11
Risk factor Categories Controls Melanomas Totals
Here we report the results of a case-control study of
melanoma risk factors conducted in Tuscany (Italy) Occupational exposure Minor 20 6 26
in 2002-2003. The aim of the study was to assess the χ2 (NS) A few years 16 8 24
Spearman (NS) Part-time 20 8 28
significance of risk factors commonly associated with Most of time 8 10 18
MM in a representative sample of the Tuscan popula- Full-time 216 108 324
Recreational exposure None 40 16 56
tion, combining objective measurement of skin color χ2 (NS) Minor 152 54 206
with other known data acquisition methods (ques- Spearman (NS) Medium 76 70 146
tionnaire and dermatological examination). Strong 12 0 12
Number of nevi and High-risk 76 32 108
presence of atypical nevi Medium-risk 164 44 208
χ2 (NS) Low-risk 40 64 104
Materials and methods Spearman (NS)c
Fitzpatrick phototype I 16 4 20
χ2 (NS) II 88 60 148
Selection of patients and controls Spearman (NS) III 120 74 194
IV 56 2 58
From October 2002 to May 2003 we studied all sub- Eye color Fair (green, blue) 40 46 94
jects (49 men, 91 women, total 140) of native Italian Fisher exact (S) Dark (brown, black) 240 94 326
Spearman (S)
origin from Tuscany who underwent surgical exeresis Hair color Fair (red, blond) 48 26 74
of non familial cutaneous MM and non acral lentigi- Fisher exact (NS) Dark (brown, black) 232 114 346
Spearman (NS)
nous melanoma. Twenty four patients had nodular Freckles No 192 80 272
melanoma and 84 had superficial spreading melanoma. Fisher exact (NS) Yes 88 60 148
Their age and gender distributions were similar to Spearman (NS)
Sunburn No 180 74 254
those of the Tuscan Cancer Registry from 1985 to Fisher exact (NS) Yes 100 66 166
1987. A group of 280 age- and gender-matched con- Spearman (NS)
trols of native Italian origin living in Tuscany was
evaluated in the same period. The control group con-
sisted of subjects chosen by random sampling from strong). Patients were also asked to recall episodes of
the computerized demographic files of Siena, Arezzo sunburn in infancy or adolescence (yes/no). Complete
and Grosseto hospitals. These files contained all the skin examination was performed by a dermatologist to
data required for the present study of subjects attend- assess the number of nevi, distinguishing 3 categories:
ing the dermatology clinics. Subjects with a history less than 10, between 10 and 30, and more than 30. Eye
of phototherapy or skin tumors were excluded from the color was recorded as fair (green-blue) or dark (brown-
control group. The control group may, of course, have black), hair color as fair (red-blond) or dark (brown-
contained subjects who will develop melanoma, how- black), Fitzpatrick phototype as I, II, III or IV and
ever, since the recent mean prevalence of melanoma in freckling (as yes/no). The exact description of cate-
Tuscany is very low (68 cases per million), we regard gories is shown in Table I.
them as normal, though low-risk would be a more
appropriate term. Semi-quantitative risk factors

Qualitative risk factors Complete skin examination was performed by the

All subjects gave their informed consent and filled
in a standardized questionnaire about occupational
and recreational sun exposure. Occupational exposure
was scored in 5 categories (minor, a few years, part-
time, most of the time and full-time) and recreational
exposure in 4 categories (none, minor, medium and
same dermatologist to assess the number and type of
nevi, as recently suggested by Carli et al.,12 distin-
guishing subjects into 3 categories: high-risk when
there were 30 or more common acquired nevi and 3 or
more atypical nevi; medium-risk when there were less
than 30 common acquired nevi and 3 atypical nevi but
more than 15 acquired nevi; low-risk when there were

374 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

METHODOLOGICAL PROCEDURE FOR EVALUATION OPF RISK FACTORS FOR CUTANEOUS MALIGNANT MELANOMA
less than 15 acquired nevi. Clinically atypical nevi
were defined as acquired macular or slightly palpa-
ble nevi with a minimum of 3 of the following 5 cri-
teria: diameter 5 mm or more, asymmetrical shape,
ill-defined border, irregular brown pigmentation, and
erythema. Nevi showing only 1 or 2 criteria were
counted as normal nevi, although histological features
of dysplasia can sometimes be found in these nevi.
Quantitative risk factors
Six quantitative variables representing objective
skin color were also considered. Skin color was mea-
sured with a Minolta CR-300 colorimeter consisting of
a detector and a microcomputer. The detector contains
a pulsed xenon arc lamp in a mixing chamber and pro-
vides diffuse illumination over the sample to be ana-
lyzed. Six high-sensitivity silicon photocells, filtered
to match the Commission International d’Eclairage
standard observer response in a double-beam feed-
back system, measure incident and reflected light. The
CR-300 detects any slight deviation in the spectral
power distribution of the pulsed radiation and corrects
it automatically. The opening of the detector is fitted
with an applicator so that dermal vessels are not com-
pressed. The CR-300 expresses the results in 5 differ-
ent color systems. We chose the Yxy system because
it gives parametric color measurements and is widely
used in dermatology. After 15 min acclimatization in
a room with air conditioning at 18°C, skin color was
measured on the upper medial quarter of the buttock
(constitutional color) and on the cheek (facultative
color), taking care not to press the detector heavily
onto the surface, which could cause ischemia. The
chromameter was calibrated between individuals. The
color measurement was read 3 times and averaged.
Two terns of variables Yxy were, therefore, measured
for each subject, specifically 3 colorimetric values,
Yc, xc and yc, for constitutional skin color and 3 more
colorimetric values, Yf, xf and yf, for facultative sun-
exposed skin color.
Statistical analysis
Fourteen variables were analyzed for involvement
with melanoma risk. Frequency count and contingency
tables were calculated for each categorical variable and
the χ2 test was used to analyze differences in variables
between melanoma patients and control subjects. For
2×2 tables, the more powerful Fisher exact test was
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
RUBEGNI
used instead of the χ2 test. The Spearman correlation
coefficient was also computed to check statistically sig-
nificant associations between melanoma and the cate-
gories of ordinal variables used. Colorimetric variables
were described statistically as mean, standard devia-
tion (SD) and range for all cases and separately for
melanoma patients and controls. Univariate differences
between groups were tested by F statistics. Stepwise
logistic regression analysis was then carried out with all
14 variables (covariates) to identify a statistically sig-
nificant minimum subset of variables with the highest
possible power in discriminating melanoma patients
from controls and quantifying risk. Logistic discrimi-
nation is generally preferable to linear discrimination in
small samples, especially when distributions are sus-
pected to be non-Gaussian. In logistic regression the
dependent variable is binary, i.e. with only 2 possible val-
ues, in our study melanoma/control. The method
assumes that the posterior probabilities P1 and P2 of
group membership follow the logistic model:
eV
P1 = 1 + eV
1 + eV
1
P2 = 1 - P1 = 1 + eVV
1+e
where V is a linear function of one or more indepen-
dent variables, that is:
V = b0 + b1x1 + b2x2 + ... + +bnxn
where x1, x2, ..., xn are the n independent variables and
b0, b1, b2, ..., bn the corresponding model parameters
to be estimated from experimental data.
The ratio of probabilities
P1 = eV
P1 = eV
P2 = eV
expresses the risk of group 1 (melanoma patients) with
respect to group 2 (healthy subjects). The exponen-
tial quantities ebi (i=1,2, ..., n) are known as odds ratios.
Supposing the variables (risk factors) to be linearly
uncorrelated, the odds ratio represents the contribution
of unit variable to relative risk. Therefore, for dichoto-
mous variables binary coded 0 or 1, the odds ratio is
simply the relative risk of the category coded 1 with
respect to the category coded 0. For qualitative vari-
ables with n categories, we have n-1 odds ratios relat-
ed to the remaining category taken as reference. To
interpret odds ratios of quantitative variables it is con-
venient to standardize them, that is to transform them
375
RUBEGNI METHODOLOGICAL PROCEDURE FOR EVALUATION OPF RISK FACTORS FOR CUTANEOUS MALIGNANT MELANOMA

into z-scores with 0 mean and unit SD. Odds ratios TABLE II.—Descriptive statistics for colorimetric variables and Fisher
can then be interpreted as representing an increase of statistics significance of differences between control and melano-
ma cases: HS=highly significant, P<0.01; S=significant, P<0.05;
one SD from the mean value of the variable. NS=not significant, P>0.05.
In the presence of confusing variables which may
Colorimetric Standard
affect the relative risk due to causes not related to the variables
Group Mean
deviation
Min Max
study (for example sample distortions in age or gender),
estimates of odds ratios related to real risk factors can Yc Total 38.27 4.3 22.62 46.93
Fisher (S) Melanoma 39.09 3.3 30.86 46.93
be corrected by including these variables in the analy- Control 37.46 5 22.62 46.55
sis. Their inclusion should be decided on the basis of Xc Total 0.3591 0.0096 0.3354 0.3885
Fisher (HS) Melanoma 0.3557 0.0075 0.3354 0.3822
a preliminary univariate analysis for testing their effec- Control 0.3625 0.010 0.3456 0.3885
tive statistical influence on relative risk. Yc Total 0.3424 0.0066 0.3300 0.3648
In the stepwise process, an independent variable is Fisher (HS) Melanoma 0.3398 0.0059 0.3300 0.3576
Control 0.3449 0.0063 0.3334 0.3648
added to or removed from the model at each step on the YF Total 27.53 4.5 19.08 38.51
basis of a statistical criterion. Many statistical criteria Fisher (S) Melanoma 26.65 4.3 19.42 38.27
for inclusion or exclusion of variables are available Control 28.40 4.5 19.08 38.51
XF Total 0.3826 0.011 0.3539 0.4076
in the following 2 main configurations: Fisher (NS) Melanoma 0.3830 0.011 0.3539 0.4045
— forward, starting (step 0) with all the variables out Control 0.3823 0.0095 0.3619 0.4076
YF Total 0.3440 0.0074 0.3230 0.3643
of the model and adding them stepwise; the removal Fisher (NS) Melanoma 0.3432 0.0061 0.3277 0.3616
process can only begin when at least 2 variables have Control 0.3449 0.0071 0.3230 0.3643
been entered in the model;
— backward, starting with all the variables in the
model and removing them stepwise. Results
The process stops when addition or removal of vari-
ables no longer improves statistical significance. The Frequency counts of categorical variables are shown
subset of variables entered in the model is then consid- in Table I. Univariate statistically significant differ-
ered for interpretation of the multivariate model. We ences between control and melanoma subjects are also
used the criterion of maximum likelihood ratio in the for- reported for each categorical risk factor (P-value of
ward configuration. The Hosmer-Lemeshow test was χ2 or Fisher-exact test). Spearman correlation analy-
used to evaluate the fit of the logistic multivariate mod- sis assessed statistically significant associations
el step by step. An associated value of P=1 indicated per- between risk factors and melanoma. Highly signifi-
fect fit and P < 0.05 indicated 95% statistical significance cant differences between controls and cases were found
of disagreement between model and experimental data. for the 3 classes of number of nevi and presence of
Values of P between 0.05 and 1, therefore, demonstrate atypical nevi and for Fitzpatrick phototype (χ2 test,
a statistically significant fit. A classification matrix was P<0.01). Recreational sun exposure and eye color gave
formed by assigning each case to the group having a less significant differences (χ2 or Fisher exact test,
probability greater than 0.5. The percentage of correct P<0.05). Hair color type, freckles, sunburn and occu-
classification was calculated. Univariate logistic regres- pational sun exposure did not show significant differ-
sion was also obtained from stepwise results at step 0. ences (χ2 or Fisher exact test, P>0.05). A statistically
Any undesired confusing effect of age and gender on significant ordinal association with melanoma was
melanoma risk was also examined by univariate logis- only found for a number of nevi and the presence of
tic regression. Odds ratios and 95% confidence intervals atypical nevi (Spearman, P<0.01) and eye color (Spear-
(CI) were evaluated to establish the relative risk of man, P<0.01).
melanoma by uni- and multivariate models, for statis- Table II shows descriptive statistics for colorimetric
tically significant variables. For this purpose, quantita- variables and Fisher test for statistical comparison of
tive colorimetric variables were standardized, so that melanoma and control cases. Significant differences
odds ratios represented the relative risk associated with were found in constitutional colorimetric variables xc
an increase of one SD from their mean value. and yc (P<0.01) and in constitutional, Yc, and facul-
Statistical analysis was performed using SPSS sta- tative, Yf, reflectances (P<0.05). No statistically sig-
tistical software. nificant differences were found in facultative colori-

376 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

METHODOLOGICAL PROCEDURE FOR EVALUATION OPF RISK FACTORS FOR CUTANEOUS MALIGNANT MELANOMA RUBEGNI

metric variables xf and yf (P>0.05). Univariate analy- TABLE III.—Univariate logistic regression. Odds ratios and 95%
sis showed that the major colorimetric factor related to confidence intervals (CI) are only reported for statistically signi-
ficant (P<0.05) variables.
probability of developing melanoma was chromatic-
ity of unexposed skin, followed by reflectance of unex- Reported Odds 95% CI
Risk factors Catogories
posed and chronically exposed skin. Univariate logis- category ratios Lower Upper
tic regression of age and gender failed to reveal sta-
tistical significance, indicating that it is not necessary Eye color Fair Dark 2.4 1.1 5.3
Number of nevi and Medium risk Low risk 0.64 0.27 1.5
to correct for confusing effects of these variables on presence of atypical nevi High risk 3.8 1.4 10
melanoma risk. Fitzpatrick phototype I IV 7 0.50 98
The results of univariate logistic regression applied II 19.1 2.3 156
III 17.3 2.1 139
to the 14 variables investigated are shown in Table III. Yc * 1.5 1 2.1
Only risk factors giving a statistically significant xc * 0.44 0.29 0.66
odds ratio (P>0.05) are reported. It was found that fair yc * 0.39 0.25 0.60
Yf * 0.67 0.47 0.94
eyes are associated with a significantly greater risk of
melanoma than dark eyes (odds ratio 2.4, 95% CI *Odds ratios of quantitative colorimetric variables represent an increase of one stan-
dard deviation with respect to the mean
1.1-5.3). Referring to the class number of nevi and the
presence of atypical nevi, an increased risk differ-
ence was demonstrated for high-risk class in com-
parison with a low-risk class (odds ratio 3.8, 95% CI TABELLA IV.—Multivariate stepwise logistic regression. Odds ratios
and 95% confidence intervals (CI) are only reported for statistically
1.4-10), whereas medium-risk class does not supply significant (P<0.05) variables.
appreciable risk differences. Although odds ratios of
95% CI
Fitzpatrick phototypes had large CI indicating poor Step N. Risk factors Categories
Reported Odds
accuracy in phototype estimation, the risk associat- category ratios Lower Upper
ed with types I-III was much greater than that asso-
1 yc * 0.20 0.088 0.44
ciated with type IV. The fairest constitutional skin 2 Number of nevi Medium risk Low risk 0.83 0.29 2.40
color is also associated with increased risk of and presence of High risk 6.74 1.85 24.60
melanoma: an increase of one SD in reflectance Yc atypical nevi
3 Yf * 0.33 0.18 0.60
leads to a significant increase in risk (odds ratio 1.5, 4 Fitzpatrick I IV 11.3 1.36 92.90
95% CI 1-2.1) and the same increase (one SD) in phototype II 12.1 1.38 105.00
chromaticity xc and yc is associated with a significant III 0.65 0.027 15.70
5 Recreational Minor Null 1.35 0.35 5.20
decrease in risk (odds ratios: xc=0.44, 95% CI 0.29- exposure Medium 7.62 1.65 35.20
0.66; yc=0.39, 95% CI 0.25-0.60). With regard to Strong 0.073 0 1.4×1015
6 yf * 2.05 1.11 3.81
facultative skin color, an increase in reflectance Yf led
to a decrease in risk (odds ratio 0.67, 95% CI 0.47- *Odds ratios of quantitative colorimetric variables represent an increase of one
standard deviation with respect to the mean
0.94). This is in line with the fact that subjects who
expose their skin less to sunlight (and therefore have
paler skin color on the cheek) are at lower risk for
MM. No significant differences were found between or. Once quantitative constitutional and facultative
cases and controls with regard to type of exposure skin color was entered at step 3, the model already
(occupational or facultative), hair color, freckling or has a satisfactory fit and power of discrimination
early history of sunburn. between melanoma and healthy subjects. The Hos-
The multivariate logistic model obtained by step- mer-Lemeshow test gave a P-value of 0.25 indicating
wise regression is shown in Table IV. Some of the a significantly good model fit. Model sensitivity and
results are interesting. When the constitutional skin specificity were 70% and 75.7%, respectively. The
color component yc was entered, at the first step, the Fitzpatrick phototype was entered at step 4. Estimat-
other constitutional skin color components Yc and xc ed odds ratios indicated a risk of melanoma about 11
lose significance, indicating substantial correlation and 12 times greater for phototypes I and II, respec-
among the 3 colorimetric components. One colori- tively, than for phototype IV, though phototypes had
metric component is, therefore, sufficient to account for poor accuracy (large CIs). Recreational exposure
relative melanoma risk due to constitutional skin col- entered at step 5. This furnishes less accurate esti-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 377

RUBEGNI METHODOLOGICAL PROCEDURE FOR EVALUATION OPF RISK FACTORS FOR CUTANEOUS MALIGNANT MELANOMA
mates of odds ratio but towards an increase of risk for
greater exposures. At step 6 yf entered, with a well-esti-
mated odds ratio of about 2 which confirms that high-
er facultative pigmentation significantly increases the
risk of melanoma.
Finally, the multivariate logistic model gives a good
fit (Hosmer-Lemeshow test, P=0.42) with apprecia-
ble sensitivity, specificity and accuracy: 78.6%, 75.7%
and 77.1% respectively.
Discussion and conclusions
Many environmental and constitutional risk fac-
tors have been associated with MM.1 Among consti-
tutional factors, numerous melanocytic nevi or the
presence of atypical nevi is the major known risk
factor.1, 13, 14 The former has also emerged as a risk
factor in studies on Caucasian and Mediterranean
peoples,12, 15 and in Celtic people with fair photo-
type (northern Europe and Australia). 16 We also
encountered a highly significant difference in nevus
number and type between controls and melanoma
patients in our study. This finding emerged both with
univariate and multivariate statistical analysis. We
found that subjects we defined as high risk on the
basis of these 2 factors had a much higher risk of
melanoma than those defined as low risk. An inter-
esting new finding, different from those of other stud-
ies, was that subjects defined by us as medium risk
actually had a risk of melanoma similar to low risk
subjects. We could, therefore, have divided the pop-
ulation into 2 (high and medium-low risk) instead of
3 groups on the basis of nevus number.
A variable we found highly significant despite an
odds ratio indicative of high CIs was the Fitzpatrick
phototype, and specifically, the risk associated with
types I-III was much greater than that associated with
type IV. However, this variable turned out to be cor-
related and completely explained by colorimetric
variables of skin color. This means that we could
have omitted phototype and studied only skin color.
The major skin color parameters correlated with risk
of MM were chromaticity of unexposed skin (but-
tock y) and reflectance (fairness) of unexposed and
chronically exposed skin (buttock and cheek Y). It is
well known that fair subjects are more susceptible
to skin cancer because they are more vulnerable to UV
light, the major environmental risk factor for skin
cancer.17
378 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
With regard to facultative skin color, an increase in
reflectance (Yf), i.e. fair skin on the cheek, is associ-
ated with a decrease in melanoma risk. This finding is
in line with the fact that subjects with lower exposure
to sunlight (and hence fairer exposed skin on the cheek)
have a lower risk of MM.17 As indicated by all the
international literature,1, 12, 15, 16 univariate and multi-
variate analysis identified fair eyes as a significantly
greater risk factor than dark eyes, in our study. On the
other hand, no significant differences in type of expo-
sure (occupational or facultative), hair color, freck-
ling or early sunburn history were found between cas-
es and controls by univariate analysis. Some of these
findings disagree with the results of epidemiological
studies from other geographical areas,1, 16-18 suggest-
ing that risk factors on which to base prevention cam-
paigns have to come directly from population sub-
groups and cannot be based on generalizations from
studies on climatically, geographically and racially
different peoples. The absence of correlations between
risk of melanoma, hair color and freckling, for exam-
ple, may be due to the fact that few Tuscans have red
hair and develop freckles. With regard to early histo-
ry of sunburn, the lack of significance of this factor had
already been reported in a population geographically
and culturally similar to ours 12 and this could be due
to the large percentage of subjects with phototype III
or IV, who are unlikely to burn, and to nearness to the
sea and temperate climate which permit continuous
exposure, so that skin is protected against acute sun-
burn. Indeed, in our study population, only 6% of
melanoma patients and 3.8% of controls had a histo-
ry of sunburn.
As far as we know, 7 case-control studies have been
conducted on melanoma risk factors in the Italian pop-
ulation.12, 15, 19-24 Because these studies differed in aim,
method (qualitative and/or quantitative) and study
population, their results cannot readily be compared.
In particular, northern and southern Italy are charac-
terized by prevalences of fair and Mediterranean pho-
totypes, respectively, and by quite different climates.
Moreover, most of these case-control studies were
based on retrospective assessment of sun exposure
and self-reported information on individual sensitiv-
ity to UV radiation. In conclusion, we agree with what
was recently proposed by Brenner et al.23 that objec-
tive skin color measurements must be combined with
phenotype parameters and sun exposure history for
exact assessment of individual risk.
Agosto 2005
METHODOLOGICAL PROCEDURE FOR EVALUATION OPF RISK FACTORS FOR CUTANEOUS MALIGNANT MELANOMA
Riassunto
Procedura metodologica per la valutazione dei fattori di
rischio per melanoma cutaneo maligno in un campione rap-
presentativo della popolazione toscana
Obiettivo. Numerosi fattori di rischio costituzionali e
ambientali sono stati messi in relazione con l’insorgenza del
melanoma cutaneo. I risultati degli studi epidemiologici
effettuati sino ad oggi sono, tuttavia, spesso non confronta-
bili tra loro a causa della carenza di parametri oggettivi
(numerici) quantificabili e della diversità razziale esistente
tra le varie popolazioni esaminate. Lo scopo di questo lavo-
ro è stato valutare la significatività dei fattori di rischio comu-
nemente associati al melanoma, in un campione rappresen-
tativo della popolazione della Toscana, associando ai già
utilizzati strumenti di raccolta dati (questionario e visita der-
matologica), la misurazione oggettiva del colore della pelle.
Metodi. Centoquaranta soggetti di origine italiana, nati e
residenti in Toscana, sottoposti ad asportazione chirurgica di
melanoma cutaneo (non familiare e non acrale-lentiginoso) e 280
soggetti controllo, simili ai primi per età, sesso e origine hanno
compilato un questionario e sono stati sottoposti a visita der-
matologica e valutazione strumentale del colore della pelle.
Risultati. L’analisi statistica univariata e multivariata ha
dimostrato differenze statisticamente significative tra il grup-
po dei pazienti e quello dei controlli per le 3 classi “nume-
ro di nevi e presenza di nevi atipici” e per il fototipo secon-
do Fitzpatrick. Inoltre differenze significative sono state
apprezzate per il colore degli occhi e il colore costituziona-
le della cute. L’analisi univariata ha mostrato che la variabi-
le maggiormente correlata al rischio di ammalarsi di mela-
noma è la cromaticità della cute non esposta al sole, segui-
ta dalla reflettanza della cute non esposta e di quella esposta.
In questo studio è stata riscontrata una differenza statistica-
mente significativa di rischio di sviluppare melanoma cuta-
neo tra i pazienti e i controlli per quanto riguarda il numero
e la tipologia dei nevi. Variabili risultate altamente signifi-
cative sono state, inoltre, il colore cutaneo costituzionale
(misurato mediante colorimetria) e il colore chiaro degli
occhi. Tutti gli altri fattori di rischio esaminati non hanno rag-
giunto un elevato grado di significatività statistica.
Conclusioni. Come già suggerito anche da altri Autori, si con-
corda circa la necessità di associare la misurazione oggettiva del
colore cutaneo alle valutazioni delle caratteristiche fenotipiche
per una corretta valutazione del rischio di melanoma.
PAROLE CHIAVE: Melanoma cutaneo - Fattori di rischio -
Misura del colore cutaneo.
References
1. Desmond RA, Soong SJ. Epidemiology of malignant melanoma. Surg
Clin North Am. 2003;83:1-29.
2. Koh HK. Cutaneous melanoma. N Engl J Med1991;325:171-82.
3. Swerlick RA, Suephy C. The melanoma epidemic. Arch Dermatol
1996;132:881-4.
4. Holman CD, James IR, Gattey PH. An analysis of trends in mortali-
ty from malignant melanoma of the skin in Australia. Int J
Canc1980;26:703-9.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
RUBEGNI
5. Rubegni P, Burroni M, Cevenini G, Perotti R, Dell’Eva G, Barbini P
et al. Digital dermoscopy analysis and artificial neural network for the
differentiation of clinically atypical pigmented skin lesions: a retro-
spective study. J Invest Dermatol 2002;119:471-4.
6. MacKie RM, Freudenberger T, Aitchison TC. Personal risk-factor
chart for cutaneous melanoma. Lancet 1989;1:487-90.
7. Cockburn M, Black W, McKelvey W, Mack T. Determinants of
melanoma in a case-control study of twins (United States). Cancer
Causes Control 2001;12:615-25.
8. Andreassi L, Casini L, Simoni S, Bartalini P, Fimiani M. Measurement
of cutaneous colour and assessment of skin type. Photodermatol Pho-
toimmunol Photomed1990;7:20-4.
9. Evans RD, Kopf AW, Lew RA. Risk factors for the development of
malignant melanoma. I. Review of case control studies. J Dermatol
Surg Oncol 1988;14:393-408.
10. Rubegni P, Cevenini G, Barbini P, Flori ML, Fimiani M, Andreassi L.
Quantitative characterization and study of the relationship between con-
stitutive-facultative skin color and phototype in Caucasians. Pho-
tochem Photobiol 1999;70:303-7.
11. Weatherall IL, Coombs BD. Skin color in terms of CIELAB color
space values. J Invest Dermatol 1992;99:468-73.
12. Carli P, Balzi D, de Giorgi V, Massi D, Palli D, Chiarugi A et al.
Results of surveillance programme aimed at early diagnosis of cuta-
neous melanoma in high risk Mediterranean subjects. Eur J Derma-
tol 2003;13:482-6.
13. Bressac-de-Paillerets B, Avril MF, Chompret A, Demenais F. Genet-
ic and environmental factors in cutaneous malignant melanoma.
Biochimie 2002;84:67-74.
14. Rokuhara S, Saida T, Oguchi M, Matsumoto K, Murase S, Oguchi S.
Number of acquired melanocytic nevi in patients with melanoma and
control subjects in Japan: nevus count is a significant risk factor for
nonacral melanoma but not for acral melanoma. J Am Acad Derma-
tol 2004;50:695-700.
15. Naldi L, Lorenzo Imberti G, Parazzini F, Gallus S, La Vecchia C.
Pigmentary traits, modalities of sun reaction, history of sunburns,
and melanocytic nevi as risk factors for cutaneous malignant melanoma
in the Italian population: results of a collaborative case-control study.
Cancer 2000;88:2703-10.
16. Youl P, Aitken J, Hayward N, Hogg D, Liu L, Lassam N et al.
Melanoma in adolescents: a case-control study of risk factors in
Queensland, Australia. Int J Cancer 2002;98:92-8.
17. Kennedy C, Bajdik CD, Willemze R, De Gruijl FR, Bouwes Bavinck
JN; Leiden Skin Cancer Study. The influence of painful sunburns
and lifetime sun exposure on the risk of actinic keratoses, seborrhe-
ic warts, melanocytic nevi, atypical nevi, and skin cancer. J Invest
Dermatol 2003;120:1087-93.
18. Whiteman DC, Whiteman CA, Green AC. Childhood sun exposure as
a risk factor for melanoma: a systematic review of epidemiologic
studies. Cancer Causes Control 2001;12:69-82.
19. Cristofolini M, Franceschi S, Tasin L, Zumiani G, Piscioli F, Talami-
ni R et al. Risk factors for cutaneous malignant melanoma in a north-
ern Italian population. Int J Cancer 1987;39:150-4.
20. Zanetti R, Franceschi S, Rosso S, Colonna S, Bidoli E. Cutaneous
melanoma and sunburns in childhood in a southern European popu-
lation. Eur J Cancer 1992;28A:1172-6.
21. Carli P, Biggeri A, Giannotti B. Malignant melanoma in Italy: risks
associated with common and clinically atypical melanocytic nevi. J
Am Acad Dermatol 1995;32(5 Pt 1):734-9.
22. Landi MT, Baccarelli A, Calista D, Pesatori A, Fears T, Tucker MA
et al. Combined risk factors for melanoma in a Mediterranean popu-
lation. Br J Cancer 2001;85:1304-10.
23. Brenner AV, Lubin JH, Calista D, Landi MT. Instrumental measure-
ments of skin color and skin ultraviolet light sensitivity and risk of cuta-
neous malignant melanoma: a case-control study in an Italian popu-
lation. Am J Epidemiol 2002;156:353-62.
24. Fargnoli MC, Piccolo D, Altobelli E, Formicone F, Chimenti S, Peris
K. Constitutional and environmental risk factors for cutaneous
melanoma in an Italian population. A case-control study. Melanoma
Res 2004;14:151-7.
379

Photodynamic therapy of actinic keratoses
with methyl-aminolevulinate (METVIX®)
Aim. Actinic keratoses (AKs) are considered one of the most
common cutaneous intraepithelial neoplastic disorders. Some
authors have recently proposed to define AK as a keratinocyte
intraepithelial neoplasia with 3 possible steps of evolution
towards squamous cell carcinoma. According to others, AK is
a true neoplasia from the very beginning. It has been calcu-
lated that 60% of subjects with skin from I to III phototype over
40s present at least one AK. This diffuse and progressing con-
dition requires a prompt diagnosis and an adequate treatment.
The prevalence of AKs increases with age, with sun exposure
and is regulated by different factors, including immunosur-
veillance and low phototype. The more risk factors there are,
the more diffuse the lesions which required multiple therapies.
Up to now both medical (5-FU cream, imiquimod cream,
retinoids etc.) and surgical (cryosurgery, laser surgery,
diclofenac gel, etc.) treatments have been used to treat AK with
variable results. Photodynamic therapy (PDT) seems to rep-
resent a new, effective and well tolerated therapy for the treat-
ment of AKs with excellent cosmetic results and long term fol-
low up in terms of efficacy. In this paper the efficacy of PDT
treatment has been evaluated with special attention to the
employment of methyl-ester of aminolevulinic acid (MAL),
which is a prodrug recently introduced in our country.
Methods. We have treated 100 Caucasian patients (70 males, 30
females) with a skin phototype ranging from 1 to 3 according
to the Fitzpatrick classification, for a total of 170 AKs of the face
and scalp.
Results. 15 days after the treatment showed complete healing
in 114 lesions of the face (82.4%) and in 44 lesions of the scalp
(78%). 84% of the more superficial and less squamous ker-
atosis (grade I) presented a complete response against 80% of
Address reprint requests to: Dott. R. Rossi, U.O. Struttura Dermatolo-
gica Complessa Fisioterapia Dermatologica, Dipartimento di Scienze Der-
matologiche, Via Della Pergola 58, 50121 Firenze.
E-mail: r.rossi@med.unifi.it
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:381-7
R. ROSSI, L. MAVILIA, I. GHERSETICH, T. LOTTI
Unit of Dermatological Physiotherapy
Department of Dermatological Sciences
University of Florence, Florence, Italy
grade II lesions. The general response to the first treatment
was 77%. We showed 90% of complete healing with excellent
compliance and cosmetic results.
Conclusion. This study has shown that metylaminolevulinate
PDT is an effective, safe and well tolerated treatment for AKs,
which could probably be considered the treatment of choice
for this very common and emerging cutaneous disorder. PDT
is also a promising treatment modality with a good potential for
future development in different fields, such as T cell lymphoma,
acne, localized eczema, and human papillomavirus infection.
KEY WORDS: Skin neoplasms - Actinic keratoses - Photodynamic
therapy - Methyl-aminolevulinate.
A ctinic keratoses (AKs) are considered one of the
most common cutaneous intraepithelial neoplastic
disorders. Yantos et al.1 have recently proposed defin-
ing AK as a keratinocyte intraepithelial neoplasia with
3 possible steps in their evolution towards squamous
cell carcinoma. According to others, AK is a true neo-
plasia from the very beginning.2, 3 It has been calculated
that 60% of subjects over-40s showing a skin with I to
III phototype present at least one AK and that this per-
centage increases to 80% in the over-60s.4
In patients with diffuse signs of photocarcinogene-
sis AKs may undergo difficult therapeutic manage-
ment, especially if those subjects for professional rea-
381
ROSSI PHOTODYNAMIC THERAPY OF ACTINIC KERATOSES WITH METHYL-AMINOLEVULINATE (METVIX)

sons or life style are chronically photoexposed.5, 6 In
these cases, lesions are usually widespread and tend to
recur. Available topical treatments usually have bad
compliance, surgical treatments are usually consid-
ered too invasive for patients. Thus a new and less
invasive approach has become a hoped-for therapeu-
tic option.
In cases of patients with diffuse signs of photocar-
cinogenesis, photodynamic therapy (PDT) is consid-
ered the treatment of choice.
PDT, a treatment modality involving the use of a
photosensitizing agent, oxygen, and light of a specif-
ic wavelength to produce controlled cell death, has
gained increasing popularity in the treatment of pre-
malignant and malignant skin lesions.7 This treatment
offers the potential advantage of reduced scarring and
improved cosmetic outcome compared with conven-
tional treatments. PDT using topical 5-aminolevulin-
ic acid (ALA), a precursor of the active endogenous
photosensitizer protoporphyrin IX, is effective in the
treatment of Aks.8-13 In contrast with systemic photo-
sensitizers, persistent skin photosensitization seems not
to be a problem for topical ones. Numerous studies have
led to the approval of this therapy for dermatological pur-
poses, which was obtained in 1999 from the USA FDA
for a topical formulation (Levulan® Kerastick, DUSA
Pharmaceuticals, Tarrytown, NY) 14, 15 and more recent-
ly in Europe and Italy (March 2004) for methy-
laminolevulinate (MAL) (Metvix®, PhotoCure, Oslo,
Norway).
The aim of the present study was to evaluate the effi-
cacy and tolerability of topical MAL- PDT (recently
approved in our country) in the treatment of AKs.16-18
Materials and methods
We evaluated 100 Caucasian patients (70 males and
30 females) of average age 68 (range 32-93) and Fitz-
patrick skin phototype ranging from 1 to 3 (Table I).
These subjects presented one or more AKs, for a
total of 170 lesions of the face (114/170, 67%) and/or
the scalp (56/170, 33%). The lesions were not pig-
mented and of grade 1 (119/170, 70%) or moderate
(grade 2) (51/170, 30%).
Lesions was classified as follows:19
Grade 1: easily visible, slightly palpable
Grade 2: easily visible, palpable
Grade3: frankly visible and very hyperkeratosic.
TABLE I.—Baseline characteristics of patients.
Sex
Male 70 (70%)
Female 30 (30%)
Age (years)
Mean 68
Range (32-93)
Skin type
(Fitzpatrick skin type)
I 10%
II 30%
III 60%
Total no. of lesions 170
Lesion location
Face 114/170 (67%)
Scalp 56/170 (33%)
No. of lesions per grade
Grade I (AK thin) 119/170 (70%)
Grade II (AK moderate) 51/170 (30%)
Treatment
Each lesion was prepared before treatment to facil-
itate access of the cream and to ensure that illumina-
tion was not blocked. Scales and crusts were removed
by a small dermal curette and the surface of the lesion
was scraped gently. The intention of this very gentle
curettage was to remove scales and crusts without
bleeding. A thick layer of 160 mg/g MAL cream
(Metvix®, Photocure ASA, Oslo, Norway) (approxi-
mately 1 mm) was applied to each lesion and 5 mm of
surrounding tissue and covered with an occlusive
dressing (Tegaderm, 3M Health care, St Paul, MN,
USA) and covered with gauze to avoid photo exposure
and to prevent the accidental activation of the cream
(photobleaching). After 180 min lesions were exam-
ined by fluorescence with a light emitting diode (LED)
lamp (DICAM-UV® - Alpha Strumenti, Milan, Italy)
with UV at 405 nm irradiation to better appreciate
the AKs lesions and the penetration of the cream (Fig-
ure 1).
After 3 h (from the beginning) dressings were
removed and lesions treated with non-coherent red
light. We employed a device which uses so called LED
light. This lamp (Aktilite PDT - Model CL128- Pho-
tocure ASA, Oslo, Norway) has an average wave-
length of 630 nm, light dose 37J/cm2, light intensity 70-
100 mW/cm2. It illuminates areas from 80 to 180 mm
at a distance from 50 to 80 mm (Figure 2).
After PDT treatment, all lesions were treated with
topical antibiotic ointments until complete healing
(approximately 1 week).
Some patients requested burning and stinging dur-

382 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

PHOTODYNAMIC THERAPY OF ACTINIC KERATOSES WITH METHYL-AMINOLEVULINATE (METVIX) ROSSI

Figure 1.—DICAM UV - Alpha Strumenti, Milan, Italy.

ing the treatment; in these cases we pretreated the area

with EMLA for 1 h before treatment in the next ther-
apeutic session.
The first control was performed 15 days after the
first treatment. At this point clinically cleared lesions
were included in the follow up program (control every
3 months), while not completely cleared lesions were
treated again following the same algorithm.

Results

All selected patients (100 patients presenting with

170 lesions, spread for 67% over the face, 33% on the Figure 2.—Aktilite “ PDT- Model CL128 - Photocure ASA, Oslo, Norway,
scalp) were treated after informed consent, with a ses- distributed by Galderma.
sion of MAL-PDT.
15 days after the treatment we showed complete
healing in 114 lesions of the face (82.4%) and in 44 bility. All results were highly satisfactory from an
lesions of the scalp (78%). 84% of the more superfi- esthetic point of view.
cial and less squamous keratosis (grade I) (Figure 3A, None of the patients underwent local intralesional
3B) presented a complete response against 80% of anesthesia.
grade II lesions (Figure 4A, 4B).
The general response to the first treatment was 77%.
Over the lesions which did not completely heal a Discussion and conclusions
second treatment was performed. Three months after
the second treatment 10 more lesions of the face and The ideal treatment for AKs should be effective,
5 more lesions of the scalp were healed. well tolerated and have an excellent cosmetic out-
In conclusions we showed a total healing of 153 of come, particularly in exposed areas such as the face.
170 treated lesions (90%) (Table II) with a higher The fashion tendency to photoexposure even in the
response for grade I lesions of the face. After 6 months winter months (sun beds, tropical trips...) have deter-
of follow up there were no recurrencies. mined a significant reduction in the average time need-
All patients enjoyed good compliance and tolera- ed for the development of AKs.

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 383

ROSSI PHOTODYNAMIC THERAPY OF ACTINIC KERATOSES WITH METHYL-AMINOLEVULINATE (METVIX)

Figure 3.—A) Grade I Ak before treatment with MAL-PDT. B) Complete response 6 months after the first MAL-PDT session.

Figure 4.—A) Actinic keratosis of the nose before MAL-PDT treatment. B) Response 15 days after MAL-PDT session.

The pathogenetic role of ultraviolet rays in the induc-
tion and progression of AKs is proved both on exper-
imental and epidemiologic models. AKs are also
known as solar keratoses. The term solar is more spe-
cific because it refers to a variety of rays. Even among
sun rays, action spectrum evaluations indicate that
ultraviolet B rays (290-320 nm) are the most damag-
ing, while UVA rays (320-400 nm) can augment the
damaging effects of UVB rays.
AKs are very often associated with other aspects of
photodamage such as actinic elastosis, teleangiec-
tasias, wrinkles and solar lentigos.20 Five percent to
20% of these lesions will progress in 10-25 years into
squamous cell carcinomas.2
Immunosurveillance can modulate the progression
of AKs towards neoplastic lesions in single subjects.
Grafted or immunocompromised patients usually have
widespread AKs with a more rapid progression towards
malignancy.
AKs are usually seen as multiple lesions in sun-
exposed areas (face, dorsa of the hands, bald portions
of the scalp in men). Usually the lesions measure less

384 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

PHOTODYNAMIC THERAPY OF ACTINIC KERATOSES WITH METHYL-AMINOLEVULINATE (METVIX)
TABLE II.—Lesion response rate by lesion grade and location 6 months after the first and the second PDT session.
MAL-PDT
Lesion location
Face 114/170 (67%)
Scalp 56/170 (33%)
Lesion grade
Grade I (AK thin) 119/170 (70%)
Grade II (AK moderate) 51/170 (30%)
Complete response
than 1 cm in diameter. They are erythematous, are
often covered by adherent scales, and except in their
hypertrophic form, show little or no infiltration. Some
solar keratoses are pigmented and show peripheral
spreading, making clinical differentiation from lenti-
go maligna difficult. Occasionally, lesions show
marked hyperkeratosis and have the clinical appearance
of cutaneous horns.
Histologically AKs are squamous cell carcinoma
in situ. However, biologically, the lesions are still
benign; invasion into the dermis, if present at all, is lim-
ited to the most superficial portion of the papillary
dermis. In the typical histological pattern the epider-
mis is thickened and shows irregular downward pro-
liferation, which is limited to the uppermost dermis
and does not represent frank invasion. Most ker-
atinocytes show a loss of polarity and a disorderly
arrangement. Some of these cells show pleomorphism
and atypia of their nuclei, which appear large, irregu-
lar and hyperchromatic.21
PDT is a treatment modality involving the admin-
istration of a photosensitizing compound and the accu-
mulation of the sensitizer molecules in the target cells,
followed by selective irradiation of the lesion with
visible light with wavelength preferentially between
600 and 700 nm, in order to achieve a deep penetration
in the tissue. Basically, photodynamic action requires
the presence and interaction of 3 components: photo-
sensitizer, light and oxygen.
The initiating step of the photosensitizing mecha-
nism is the absorption of a light photon by the sensi-
tizer, causing a promotion of the drug molecule from
its ground state to the extremely unstable excited sin-
glet state. The singlet excited photosensitizer either
decays back to the ground state, resulting in the emis-
sion of light in the form of fluorescence, or undergoes
intersystem crossover to the longer lived triplet excit-
ed state by electron spin conversion. The in situ gen-
eration of singlet oxygen via the type II pathway
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
ROSSI
After 1 Sessiom After 2 sessions
94/114 (82.4%) 10/20 (104/114) 91%
44/56 (78%) 5/12 (49/56) 87.5%
100/119 (84%) 9/19 (109/119) 91.5%
38/51 (80%) 6/13 (44/51) 86%
138/170 (77%) 153/170 (90%)
appears to play a central role in photodynamic cyto-
toxicity because of the highly efficient interaction of
the O2 species with different biomolecules.22, 23 The tis-
sue damaging effect is realized via several pathways:
i. cell necrosis and apoptosis of dysplastic cells;
ii. microcirculation arrest: damage of endothelial
cells promotes thrombus formation and consequent
vascular neological stasis which also contributes to
tumor ablation.
iii. inflammation in the exposed tissue
iv. induction of host immune response.
MAL is a new topical photosensitizer that may offer
advantages over ALA in terms of improved skin pen-
etration as a result of enhanced lipophilicity and greater
selectivity for neoplastic cells than other ALA-induced
porphyrins.24-27
In addition, as the cellular uptake mechanisms for
these agents differ, the intensity of pain may be low-
er during PDT using MAL than ALA.
The good cosmetic results which are obtained due
to selective tissue destruction and mobilization of
the organism’s proper immune response is one of the
main PDT advantages. From the pharmacological
point of view, sensitizers show low toxicity and almost
no interaction with other medications, making PDT
a safe treatment modality. Many other treatments for
AKs are available, but most of them present side
effects, and they do not assure the same esthetic
results.28-30
Cryosurgery is generally limited to patients with
only a very limited number of lesions where hypopig-
mentation can resist after treatment. Curettage has
potential complications of scarring and infection, and
local anesthesia is required before the procedure. Top-
ical application of 5-fluorouracil cream has the dis-
advantages of prolonged erythema and exudation as
part of the treatment and recovery, lack of patient com-
pliance, morbidity, only partial effectiveness in remov-
ing deep or hyperkeratotic lesions, and the potential for
385
ROSSI PHOTODYNAMIC THERAPY OF ACTINIC KERATOSES WITH METHYL-AMINOLEVULINATE (METVIX)
exacerbation of other cutaneous conditions such as
acne rosacea.6
PDT has gained increasing popularity in the treat-
ment of premalignant or malignant skin lesions. The
treatment offers the advantages of reduced scarring and
improved cosmetic outcome compared with conven-
tional treatments. The use of gentle curettage before
MAL-PDT and the occlusion of the drug before illu-
mination probably contribute to the efficacy of the treat-
ment.31 Moreover with this treatment local intralesion-
al anesthesia is not required and the stinging sensation
is usually limited to the treated area and can be easily
treated with refrigerating devices which are usually pre-
sent in the most modern and sophisticated PDT lamps.
The intensity of itching or pain seem to be lower during
PDT using MAL instead of ALA, because of the dif-
ferent cellular uptake mechanisms for these agents.
From a practical perspective, MAL PDT can be use-
fully and safely integrated into clinical practice.
In conclusion, in this study we have shown that
MAL PDT is an effective, safe and well tolerated treat-
ment for AKs, which could be probably considered
the treatment of choice for this very common and
emerging cutaneous disorder.
PDT is also a promising treatment modality with a
good potential for future development in different
fields, such as T cell lymphoma, acne, localized
eczema, and human papillomavirus infections.32-43
Acknowledgement.—Mrs. S. Lombardi and Mrs. C. Izzap are
gratefully acknowledged for their technical assistance.
Riassunto
Terapia fotodinamica con Metil-aminolevulinato (METVIX®)
nel trattamento delle cheratosi attiniche
Obiettivo. La cheratosi attinica (CA) o cheratosi solare
rappresenta la più comune neoplasia cutanea circoscritta.
Alcuni Autori hanno recentemente proposto di definire la
CA una “neoplasia intraepiteliale cheratinocitaria” con 3 gra-
di di evoluzione verso il carcinoma squamocellulare. Secon-
do altri, la CA è una vera e propria neoplasia fin dall’inizio.
È stato calcolato, infatti, che il 60% dei soggetti a basso foto-
tipo (I-III) oltre i 40 anni di età presenta almeno una cheratosi
solare. Questa patologia rappresenta, inoltre, la principale
condizione per lo sviluppo dei carcinomi a cellule squamo-
se e richiede, perciò, una rapida diagnosi e un efficace trat-
tamento. La prevalenza della CA aumenta con l’età e con
l’esposizione a fattori di rischio (elioesposizione) e varia in
presenza di fattori predisponenti (immunosoppressione). In
questi casi le lesioni sono spesso diffuse e recidivanti e neces-
386 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
sitano di trattamenti multipli. Le CA devono sempre essere
trattate. Oggi sono disponibili terapie sia mediche (crema al
5-FU, peeling medio-profondi, imiquimod crema, retinoidi
orali, interferone a2b) sia chirurgiche (elettrochirurgia, escis-
sione chirurgica, laser-chirurgia, criochirurgia, dermoabra-
sione). La terapia fotodinamica rappresenta una recente effi-
cace acquisizione per il trattamento delle CA, ben tollerata e
con eccellenti risultati cosmetici. Può, inoltre, risultare par-
ticolarmente utile per la scarsa invasività e la possibile ripe-
tibilità che la contraddistinguono. In questo lavoro è stata
valutata l’efficacia di tale terapia con particolare riguardo
all’utilizzo del metil-estere dell’ALA, profarmaco fotosen-
sibilizzante di recentissima introduzione nel nostro Paese.
Metodi. Sono stati trattati 100 pazienti (70 di sesso maschi-
le, 30 di sesso femminile) di razza caucasica e fototipo 1 e 2
o 3 secondo la scala di Fitzpatrick, per un totale di 170 CA,
localizzate al volto e al cuoio capelluto.
Risultati. Dopo 15 giorni dal primo trattamento è stata
evidenziata una risposta completa (completa regressione cli-
nica della lesione) in 114 lesioni del volto (82,4%) e in 44
lesioni del cuoio capelluto (78%). In base al grado di evo-
luzione delle cheratosi, l’84% delle lesioni più superficiali e
meno squamose (grado I ) ha presentato risposta completa
contro l’80% di quelle di grado medio (grado II). La rispo-
sta complessiva al primo trattamento è risultata pari al 77%.
Sono state ottenute risposte di guarigione completa nel 90%
dei casi, risultato sovrapponibile o superiore se comparato ai
trattamenti convenzionali.
Conclusioni. In questo studio, la terapia fotodinamica con
metil-aminolevulinato si è dimostrata una terapia efficace e
ben tollerata per il trattamento delle CA; i risultati ottenuti,
pertanto, sembrano rimarcare il potenziale ruolo della tera-
pia fotodinamica come trattamento di prima scelta per que-
sta patologia molto comune e in costante aumento. Lo sce-
nario è, inoltre, destinato ad arricchirsi rapidamente in quan-
to studi pilota sono in corso per il trattamento fototerapeu-
tico di patologie anche molto diverse tra loro, come il linfo-
ma primitivo cutaneo a cellule T e a cellule B, alcune forme
di acne e di eczema, l’ipertricosi, il lichen sclerosus et atrophi-
cans e le infezioni da papillomavirus.
PAROLE CHIAVE: Cute, carcinoma - Cheratosi attiniche - Tera-
pia fotodinamica - Metil-aminolevulinato.
References
1. Yantos VA, Conrad N, Azbawski E, Cockerell CJ. Incipient intraepi-
dermal cutaneous squamous cell carcinoma: a proposal for reclassi-
fying and grading solar (actinic) keratosis. Semin Cutan Med Surg
1999;1:3-14.
2. Salasche SJ. Epidemiology of actinic keratoses and squamous cell
carcinoma. J Am Acad Dermatol 2000;42:8-10.
3. Cockerell CJ. Histopathology of incipient intraepidermal squamous
cell carcinoma (“actinic Keratosis”). J Am Acad Dermatol
2000;115:649-55.
4. Drake LA, Ceilley RI, Cornelison RL, Dobes WL, Corner W, Goltz
RW et al. Guidelines of care for actinic keratoses. J Am Acad Dermatol
1995;32:95-8.
Agosto 2005
PHOTODYNAMIC THERAPY OF ACTINIC KERATOSES WITH METHYL-AMINOLEVULINATE (METVIX)
5. Ortonne JP. From actinic keratosis to squamous cell carcinoma. Br J
Dermatol 2002;146:20-3.
6. Schmook T, Stockfleth E. Current treatment patterns in non melanoma
skin cancer across Europe. J Derm Treat 2003;14:3-10.
7. Kennedy JC, Pottier RH, Pross DC. Photodynamic therapy with
endogenous protoprophyrin IX: basic principles and present clinical
experience. J Photochem Photohiol B 1990;14:275-92.
8. Jeffes EW, McCullough JL, Weinstein GD, Fergin PE, Nelson JS,
Shull TF et al. Photodynamic therapy of actinic keratosis with topi-
cal 5-aminolevulinic acid. Arch Dermatol 1997;133:727-32.
9. Kurwa HA, Yong-Gee SA, Seed PT, Markey AC, Barlow RJ. A ran-
domized paired comparison of photodynamic therapy and topical 5-
fluorouracil in the treatment of actinic keratoses. J Am Acad Derma-
tol 1999;41:414-8.
10. Pinzi C, Campolmi P, Moretti S, Guasti A, Rossi R, Cappugi P. Ter-
apia fotodinamica di tumori cutanei (non melanoma) primitivi e sec-
ondari con applicazione topica di acido 5-aminolevulinico. G Ital
Dermatol Venereol 2000;135:427-31.
11. Wolf P, Rieger E, Kerl H. Topical photodynamic therapy with endoge-
nous porphyris after application of 5-aminolevulinic acid: an alternative
treatment modality for solar keratoses, superficial squamous cell carci-
nomas, and basal cell carcinomas? J Am Acad Dermatol 1993;28:17-21.
12. Morton CA, Whitehurst C, Moseley H, Moore JV, MacKie RM.
Development of an alternative light source to lasers for photodynamic
therapy. 1. Clinical evaluation in the treatment of premalignant and non-
melanoma skin cancer. Lasers Med Sci 1995;10:165-71.
13. Calzavara-Pinton PG. Repetitive photodynamic therapy with topical
delta-aminolevulinic acid as an appropriate approach to the routine
treatment of superficial nonmelanoma skin tumors. J Photochem Pho-
tobiol B 1995;29:53-7.
14. Jeffes EW. Levulan: the first approved topical photosensitizer for the
treatment of actinic keratosis. J Dermatolog Treat 2002;13 Suppl 1:
S19-23.
15. Lang K, Achulte KW, Ruzicka T, Tritsch C. Aminolevulinic acid
(Levulan) in photodynamic therapy of actinic keratoses. Skin Thera-
py Lett 2001;6:1-2, 5.
16. Uehlinger P, Zellweger M, Wagnieres G Juillerat L, Van Der Bergh H,
Lange N. 5-Aminolevulinic acid and its derivates: physical chemical
properties and protoporphyrin IX formation in cultured cells. Pho-
tochem Photobiol 2000;54:72-80.
17. Fritsch C, Homey B, Stahl W, Lehmann P, Ruzicka T, Sies H. Pref-
erential relative porphyrin enrichment in solar keratoses upon topical
application of 5-aminolevulinic acid methylester. Photochem Photo-
biol 2000;71:640-7.
18. Pariser DM, Lowe NJ, Steward DM, Jarratt MT, Lucky AW, Pariser
RJ et al. Photodynamic therapy with topical methyl aminolevulinate
for actinic keratosis: results of a prospective randomised multicenter
trial. J Am Acad Dermatol 2003;98:227-32.
19. Olsen EA, Abernethy ML, Kulp-Shorten C, Callen JP, Blazer SD,
Huntley A et al. A double-blind, vehicle-controlled study evaluating
masoprocol cream in the treatment of actinic keratoses on the head and
neck. J Am Acad Dermatol 1991;5:738-43.
20. Touma D, Yaar M, Whitehead S, Konnikov N, Gilchrest B. A trial of
short incubation, broad-area photodynamic therapy of facial keratoses
and diffuse photodamage. Arch Dermatol 2004;140:33-40.
21. Petrini N, Massi D, Rossi R, Cappugi P. Patologia precancerosa e
terapia fotodinamica. In: Cappugi P, Rossi R, Mavilia L, Campolmi
P editors. La terapia fotodinamica nella pratica clinica. Manuale prati-
co e testo-atlante dermatologico. 1° ed. Firenze: SEE editrice; 2005.
22. Prignano F, Bianchi B, Domenici L, Rossi R, Romagnoli P, Pimpinel-
li N et al. Early apoptosis plays an important role in the healing mech-
anism of cutaneous basal cell carcinomas after photodynamic thera-
py. Br J Dermatol 2003;149:205-6.
23. Nakaseko H, Kobayashi M, Akita Y, Tamada Y, Matsumoto Y. Histo-
logical changes and involvement of apoptosis after photodynamic
therapy for actinic keratoses. Br J Dermatol 2003;148:122-7.
24. Stefanidou M, Tosca A, Themelis G, Vazgiourake E, Balas C. In vivo
fluorescence kinetics and photodynamic therapy efficacy of delta-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
ROSSI
aminolevulinic acid-induced porphyrins in basal cell carcinomas and
actinic keratoses; implications for optimizating of photodynamic ther-
apy. Eur J Dermatol 2000;10:351-6.
25. Cappugi P, Campolmi P, Mavilia L, Prignano F, Rossi R. Topical 5-
aminolevulinic acid and photodynamic therapy in dermatology: a
minireview. J Chemother 2001;13:494-502.
26. Cappugi P, Rossi R, Mavilia L, Campolmi P. La terapia fotodinami-
ca nella pratica clinica. Manuale pratico e testo-atlante dermatologi-
co. 1° ed. Firenze: SEE editrice; 2005.
27. Cappugi P, Rossi R. Terapia fotodinamica. In: Campolmi P, Bonan P, Can-
narozzo G editors. Laser e sorgenti luminose in dermatologia. Manuale
di applicazioni pratiche. Milano: Masson Editore; 2003. p.138-47.
28. Szeimies RM, Karrer S, Radakovic-Fijan S, Tanew A, Calzavara-Pinton
PG, Zane C et al. Photodynamic therapy using topical methyl-5-aminole-
vulinate (Metvix) is as efficacious as cryotherapy in actinic keratosis, but
with superior cosmetic results and high patient satisfaction: a prospec-
tive, randomised study. J Am Acad Dermatol 2002;47:258-62.
29. Freeman M, Vinciullo C, Francis D, Spelman L, Nguyen R, Fergin P
et al. A comparison of photodynamic therapy using topical methyl
aminolevulinate (Metvix) with single cycle cryotherapy in patients with
actinic keratosis: a prospective, randomized study. J Dermatol Treat
2003;14: 99-106.
30. Horn M, Wolf P, Wulf HC, Warloe T, Fritsch C, Rhodes LE et al.
Topical methyl aminolaevulinate photodynamic therapy in patients with
basal cell carcinoma prone to complications and poor cosmetic out-
come with conventional treatment. Br J Dermatol 2003;149:1242-9.
31. Morton CA. Methyl aminolevulinate (Metvix) photodynamic thera-
py-practical pearls. J Dermatol Treat 2003;14:23-6.
32. Coors EA, Von der Driesh P. Topical photodynamic therapy for patients
with therapy-resistant lesions of cutaneous T-cell lymphoma. J Am
Acad Dermatol 2004;50:363-7.
33. Orenstein A, Haik J, Tamir J, Winkler E, Trau H, Malik Z et al. Pho-
todynamic therapy of cutaneous lymphoma using 5-aminolevulinic acid
topical application. Dermatol Surg 2000;26:765-9.
34. Rittenhouse-Diakun K, van Leegoed H, Morgan J, Hryhorenko E,
Paszkiewicz G, Whitaker JE et al. The role of trasferrin receptor
(CD71) in photodynamic therapy of activated and malignant lym-
phocytes using the hem precursor delta-aminolevulinic acid (ALA).
Photochem Photobiol 1995;61:523-8.
35. Edstrom DW, Portwit A, Ros AM. Photodynamic therapy with topi-
cal 5-aminolevulinic acid for mycosis fungoides: clinical and histo-
logical response. Acta Derm Venereol 2001;81:184-8.
36. Stables GI, Stringer MR, Robinson DJ. The treatment of cutaneous T
cell lymphoma by topical aminolevulinic acid photodynamic therapy.
Br J Dermatol 1997;137(S50):51.
37. Markham T, Sheahan K, Collins P. Topical 5-aminolevulinic acid
photodynamic therapy for tumor-stage mycosis fungoides. Br J Der-
matol 2001;144:1262-3.
38. Mori M, Mavilia L, Rossi R, Campolmi P, Cappugi P, Pimpinelli N.
La terapia fotodinamica nel trattamento dei linfomi primitivi cutanei.
G Ital Dermatol Venereol 2005;140:123-7.
39. Hongcharu W, Taylor CR, Chang Y, Aghassi D, Suthamjariya K,
Anderson RR. Topical ALA-photodynamic therapy for the treatment
of acne vulgaris. J Invest Dermatol 2002;115:183-92.
40. Hillemanns P, Untch M, Prove F, Baumgartner R, Hillemanns M,
Korell M. Photodynamic therapy of vulvar lichen sclerosus with 5-
aminolevulinic acid. Obstet Gynecol 1999;93:71-4.
41. Fabbrocini G, Di Costanzo MP, Riccardo AM, Quarto M, Colasanti
A, Roberti G et al. Photodynamic therapy with topical 5-aminolae-
vulinic acid for the treatment of plantar warts. J Photochem Photobi-
ol B Biology 2001;61:30-4.
42. Stender IM, Na R, Fogh H, Gluud C, Wulf HC. Photodynamic ther-
apy with 5-aminlaevulinic acid or placebo for recalcitrant foot and hand
warts: randomised double-blind trial. Lancet 2000;355:963-6.
43. Ross EV, Romero R, Kollias N, Crum N, Anderson RR. Selectivity of
protoporphyrin IX fluorescence for condylomata after topical appli-
cation of 5-aminolaevulinic acid: implications for photodynamic treat-
ment. Br J Dermatol 1997;137:736-42.
387

The relationship between parental and dermatologist assessment of disease severity
E. MAZZOTTI 1, C. DI PIETRO 2, S. MASTROENI 1, A. PROVINI 3, M. PARADISI 3, S. TABOLLI 2
Aim. The instruments to measure the severity of atopic der-
matitis (AD) are time consuming, so their use is limited in
routine clinical practice. The self administered eczema area
severity index (SA-EASI) was developed and validated for
a caregiver’s self assessment of the severity of child’s AD.
The aim of this study was to assess the relationship between
the SA-EASI and SCORing atopic dermatitis (SCORAD)
tools.
Methods. Thirty-five patients with AD admitted at the Pedi-
atric Unit of a Dermatological Hospital in Rome, Italy, and
their parents, participated in the study. The severity of the dis-
ease has been assessed at admission, by parents using the
SA-EASI, and by dermatologists using the SCORAD, inde-
pendently.
Results. Evidence of convergent validity was provided by
high correlation between SA-EASI and SCORAD (Rank
Spearman rs=0.71; P<0.001).
Conclusion. Both instruments are useful in daily clinical
practice and in the research on outcomes. The parents
received no training in the measurement of eczema and no
training in the use of the SA-EASI instrument itself. SA-
EASI, moreover promotes the involvement of families with an
affected child.
KEY WORDS: Atopic dermatitis - Child - Measures.
This report is part of a broader study examining the effects of a
patient/parental education programme on medical and psychosocial out-
comes.
This study was supported by a grant from the Italian Ministry of
Health.
Received: September 15, 2004.
Accepted for publication: July 6, 2005.
Address reprint requests to: Dott.ssa E. Mazzotti, Istituto Dermopatico
dell'Immacolata, IDI-IRCCS, Via dei monti di Creta 104, Rome, Italy.
E-mail: e.mazzotti@idi.it.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:389-95
Childhood atopic dermatitis
1Epidemiology Unit, Istituto Dermopatico dell'Immacolata
IDI-IRCCS, Rome, Italy
2Health Service Research Unit, Istituto Dermopatico
dell'Immacolata, IDI-IRCCS, Rome, Italy
3Pediatric Unit, Istituto Dermopatico dell'Immacolata, IDI-
IRCCS, Rome, Italy
A topic dermatitis (AD) is a chronic skin condition
with a significant quality of life (QoL) impact.
Psychological stress, quality of family life, financial
problems and social well-being are related to the sever-
ity of AD in children.1-4
Essential to the study of the impact of AD on QoL
is the measurement of disease severity.
Different scoring systems have been developed to
determine the severity of AD.5-11 Recently, more objec-
tive scores like that using permeability barrier function
and stratum corneum hydratation with computer-assist-
ed estimates for extent disease,12 or the specific soft-
ware to evaluate automatically the extension of the
involved area 13 have been developed.
Despite better precision and reproducibility of objec-
tive measures are known, clinicians and researchers use
more classical instruments.14-16
The SCORing atopic dermatitis (SCORAD),17 one
of the best validated systems,18-20 is based on objective
signs (e.g. extension) and subjective symptoms (e.g.
pruritus and sleep loss). The instrument is suited for
clinical trials, but is too time consuming for routine
clinical use.
389
MAZZOTTI
Generally, the use of self-assessed severity indices
in dermatology is restricted to adult patients.21, 22
Recently, an instrument for a caregiver’s self-assess-
ment of the severity of his/her child’s AD, the self-
administered eczema area and severity index (SA-
EASI) has been developed.23 The original study evi-
denced that “caregivers can accurately assess their
child’s cutaneous disease severity in a valid fashion
using the SA-EASI”.23
Aims of our study are to assess the performance of
the SA-EASI italian version and to measure the rela-
tionship between parental and dermatologist assess-
ment of AD severity.
Materials and methods
Subjects
A total of 35 inpatients, 16 males and 19 females,
aged 2 months to 17 years, were recruited from the
Pediatric Unit of a dermatological hospital in Rome,
Italy. Only patients with a diagnosis of AD were
enrolled in the study. The exclusion criteria were: oth-
er concomitant severe diseases; parents not available
to complete the SA-EASI.
Instruments
The SA-EASI is a one-page instrument allowing
caregivers of children with AD to measure disease
severity. To estimate the surface area involved a line-
drawing silhouette of the body (front and back) was
presented to the caregivers and they have to shade the
areas currently affected by AD. Based on the silhou-
ette shading, an investigator not directly involved in
patient evaluation assigned a value, corresponding to
0-100% body surface area (BSA) involvement, for
each of the following 4 areas: head, upper extremi-
ties, trunk, and lower extremities. To the BSA involved
for each of the 4 body regions was assigned a propor-
tional score as defined on a seven-point ordinal scale:
0, no eruption; 1, ≤9%; 2, 10-29%; 3, 30-49%; 4, 50-
69%; 5, 70-89%; 6, 90-100%. Each area score was
then multiplied by a factor assigned to the corre-
sponding body area on the SA-EASI scoring sheet.
The multiplier varied according to body region and
the child’s age. SA-EASI weights the involvement of
the head, upper extremities, trunk, and legs as 10%,
20%, 30%, and 40% of the total BSA, respectively,
390 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHILDHOOD ATOPIC DERMATIS
for children aged above 7 years, roughly consistent
with the “rule of nines”.24 For children <7 years old, a
modification was used: BSAs were 20% for the head,
20% for the upper extremities, 30% for the trunk and
30% for the lower extremities.24 Finally the 4 products
were summed to obtain the total area score. The sec-
ond part of the one-page SA-EASI instrument con-
sisted of five 100-mm visual analogue scales (VASs).
The VAS consists of a continuous line on which the
caregivers make a mark to show the average severity
of the AD lesions. The VASs enabled caregivers to
describe the redness, thickness, dryness, number of
scratches and itchness of an average AD lesion. On
each VAS, extremes and intermediate levels were
labelled with anchor marks at equivalent intervals
along the VAS line. For example, the VAS for redness
contains the following word descriptions: no redness,
slightly pink, pink, red, and dark red. Severity scoring
was calculated from an equation.23 The percentual val-
ue, no proportional score, of BSAs were used in this
study. Translation and adaptation were authorized by
the author. In the Italian version we changed the figure
from adult to child, and the intermediate levels on
each VAS were marked by signs of 3-mm (heigh) at
intervals of 10-mm.
The SCORAD is a scoring system based on the
assessment of severity by dermatologists. The complete
system is called SCORAD index 17 and also includes
the assessment of subjective symptoms (pruritus, sleep
loss) on a VAS. The extent of lesions is scored by
applying the rule of nine after drawing the lesions on
an evaluation form like that of SA-EASI. The intensity
is determined by grading each of the 6 items on a scale
from 0 to 3 (erythema, edema/papulation, oozing/crust,
excoriation, lichenification, and dryness). Each item
should be scored on the most representative area for a
given intensity item. Finally the total score is the sum
of extent/5+7*intensity/2 in a standardized way. Owing
to this formula extent accounts for about 25% and
intensity for about 75% of the total score. The range of
the objective SCORAD lies between 0 and 83. Based
on the objective SCORAD, the severity of AD can be
classified into mild (<15), moderate (between 15 to
40), and severe (>40). In this study we adopted the
version without the assessment of subjective symp-
toms (pruritus, sleep loss).17
Both instruments were completed immediatly after
the admission.
Agosto 2005
CHILDHOOD ATOPIC DERMATIS MAZZOTTI

Statistical analysis A second administration was proposed to parents

Correlation between total SCORAD and SA-EASI
score and between BSAs and severity was calculated
using the Rank Spearman’s correlation.
Stability was calculated using the Rank Spearman’s
correlation between 2 tools, delayed 48 h, for each
instrument.
BSA scores are based on the following formulas:
BSASA-EASI=(0.1*Ah)+(0.2*Au)+(0.3*At)+(0.4*Al)
BSASCORAD=(9*Ah)+(18*Au)+(37*At)+(36*Al)
(Ah, head area score; Au, upper extremities area
score; At, trunk area score; Al, legs area score)
All analysis were performed using PC-STATA.25
Results
The investigated population (N=35; 51% females)
with ages ranging from 2 months to 17 years
(mean±SD=7.7±5.07; median=8) was assessed. Chil-
dren were affected by moderate (61.6%) and severe
(38.4%) AD according to objective SCORAD. Mean
SCORAD score was 38.8+13.03 (median= 36.1), rang-
ing between 15.2 and 70.3.
A positive correlation was observed between total
SA-EASI score and objective SCORAD (Rank Spear-
man’s rs=0.71 P<0.0001) and also the extent, accord-
ing to the rule of nine, (rs=0.68 P<0.0001). Positive cor-
relations, ranged from 0.59 to 0.70, were observed
between singular body area (Ah, rs=0.61; Au, rs=0.61;
At, rs=0.59; Al, rs=0.70; P<0.0001). The intensity items
were poorly correlated. Between redness and erythe-
ma (rs=0.42 P=0.0013), scratches and excoriation
(rs=0.43 P=0.0010), between dryness (rs=0.29
P=0.061). Other correlations were less than 0.10.
TABLE I.—Psychometric features (mean, standard deviation-sd-,
median, range) of total SCORAD score and total SA-EASI score.
At admission (time 1) and at discharge (time 2). (N=23).
Time 1 Time 2
Total SCORAD score
Mean (sd) 36.9 (11.55) 16.0 (13.24)
Median 35 14
Range 21.8-68.5 0-57.4
Total SA-EASI score
Mean (sd) 10.2 (8.24) 2.5 (5.07)
Median 8.4 1.0
Range 0.16-30.24 0-22.57
*delta=time 1-time 2
of 23 children (13 females and 10 males) and to the der-
matologist. The lenght of 48 h was choosen for bal-
ancing the effect of recall bias. Patients showed mild
(52.4%), moderate (42.9%), and severe (4.8%) AD,
according to objective SCORAD, 48-72 h after admis-
sion (Table I).
The negative delta value observed for SCORAD
range is a result of disease worsening for a patient
between 2 times observations.
Discussion
To measure AD severity SA-EASI has been proved
to be equivalent to SCORAD.
The modified Italian SA-EASI version was easily
understandable and managed by all involved parents.
The results of this preliminary study are relevant
showing parents’ ability to assess, in a reliable and
similar way to dermatologist, the extension and sever-
ity of AD in their children using this tool. All correla-
tion showed a convergence and had similar magni-
tude. As expected the tools showed differences for
what concern specific characteristic of lesions, pecu-
liar of each instrument, however such discrepancies had
no influence about the total score.
Conclusions
The SA-EASI is identified as an instrument useful
to assess the AD manifestations (intensity and exten-
sion) in an easy and reliable way. It does not request
specific training or health personnel and can be used
by patient or caregiver to monitor the disease. The
SA-EASI could be considered a communication tool
between parents and physicians. The physician can
be able to follow the response to treatments.
The SCORAD allows the dermatologist to assess
Delta* the disease in ambulatory care, where time is limited,
and in chronic patients in their follow up.
19.5 (12.20) In comparison to SA-EASI, SCORAD is a more
18.8 precise instrument.
-10.11--45.38 The dermatologist with his personal experience
7.7 (7.1) can identify different clinical manifestations, obvi-
6.1 ously reported in a wider range of the severity
1.25-22.6 assessed.
The total scores of both instruments report different
patient status.

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 391

MAZZOTTI
In clinical practice it should be useful to adopt both
tools in a hospital setting, in day-hospital or in ambu-
latory setting, and, for patient/caregiver, at home.
SA-EASI moreover participate in the empowerment
of families with an affected child.
However, other studies with more patients are nec-
essary to confirm the tools capability to monitor
changes in the disease during the follow up.
Acknowledgments.—The authors thank all the families for par-
ticipating in the study and their perseverance in completing the
tools.
References
1. Aziah MS, Rosnah T, Mardziah A, Norzila M. Childhood atopic der-
matitis: a measurement of quality of life and family impact. Med J
Malaysia 2002;57:329-39.
2. Ben-Gashir MA, Seed PT, Hay RJ. Are quality of family life and dis-
ease severity related in childhood atopic dermatitis? J Eur Acad Der-
matol Venereol 2002;16:455-62.
3. Warschburger P, Buchholz HTH, Petermann F. Psychological adjust-
ment in parents of young children with atopic dermatitis: which
factors predict parental quality of life? Br J Dermatol 2004;150:304-
11.
4. Ben-Gashir MA, Seed PT, Hay RJ. Quality of life and disease sever-
ity are correlated in children with atopic dermatitis. Br J Dermatol
2004;150:284-90.
5. Clendenning WE, Clack WE, Ogawa M, Ishizaka K. Serum IgE stud-
ies in atopic dermatitis. J Invest Dermatol 1973;61:233-6.
6. Ring J, Senter T, Cornell RC, Arroyave CM, Tan EM. Plasma com-
plement and histamine changes in atopic dermatitis. Br J Dermatol
1979;100:521-30.
7. Zachary CB, MacDonald DM. Quantitative analysis of T lympho-
cytes subsets in atopic eczema, using monoclonal antibodies and flow
cytometry. Br J Dermatol 1983;108:411-22.
8. Queille-Roussel C, Raynaud F, Saurat JH. A prospective computerized
study of 500 cases of atopic dermatitis in childhood. Acta Derm
Venereol Suppl 1985;114:87-92.
9. Costa C, Rilliet A, Nicolet M, Saraut JH. Scoring atopic dermatitis:
the simpler the better? Acta Derm Venereol 1989;69:41-5.
10. Hanifin J. Standardized grading of subjects for clinical research stud-
ies in atopic dermatitis: workshop report. Acta Derm Venereol Sup-
pl 1989;144:13-4.
La dermatite atopica infantile. Confronto tra la misura di gravità del dermatologo
e quella del genitore
L a dermatite atopica (DA) è una patologia cronica che ha
un impatto significativo sulla qualità della vita (quality of
life, QoL), non soltanto del paziente ma anche dei suoi fami-
liari: a una maggiore gravità clinica della DA corrisponde un
più elevato livello di stress psicologico, maggiori difficoltà
392 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHILDHOOD ATOPIC DERMATIS
11. Abrams BB. Atopic dermatitis: elements in clinical study design and
analysis. Acta Derm Venereol Suppl 1989;144:15-9.
12. Sugarman JL, Fluhr JW, Fowler AJ, Bruckner T, Diepgen TL, Williams
ML. The objective severity assessment of atopic dermatitis score: an
objective measure using permeability barrier function and stratum
corneum hydration with computer assisted estimates for extent of
disease. Arch Dermatol 2003;139:1417-22.
13. Tripodi S, Panetta V, Pelosi S, Pelosi U, Boner AL. Measurement of
body surface area in atopic dermatitis using specific software (Sco-
radCard(c)). Pediatr Allergy Immunol 2004;15:89-92.
14. Frederiksson AJ, Peterssonn DC. Severe psoriasis: oral therapy with
a new retinoid. Dermatologica 1978;157:238-44.
15. Bahmer FA, Schafer J, Schubert HJ. Quantification of the extent and
the severity of atopic dermatitis: the ADASI score. Arch Dermatol
1991;127:1239-40.
16. Kezawa Z, Ikebe T, Ogura H, Odajima H, Kurosaka F, Sase K et al.
Clinical effect of hypoallergenic rice HRS-1 in a atopic dermatitis. Jpn
J Allergol 1991;40:633-42.
17. European Task Force on Atopic Dermatitis. Severity scoring of atopic
dermatitis: the SCORAD index. Consensus report of the European task
force on atopic dermatitis. Dermatology 1993;186:23-31.
18. Kunz B, Oranje AP, Labreze L, Stalder JF, Ring J, Taieb A. Clinical
validation and guidelines for the SCORAD index: consensus report of
the European Task Force on Atopic Dermatitis. Dermatology
1997;195:10-9.
19. Oranje AP, Stalder JF, Taieb A, Tasset C, de Longueville M. Scoring
of atopic dermatitis by SCORAD using a training atlas by investiga-
tors from different disciplines. ETAC Study Group. Early Treatment
of the Atopic Child. Pediatr Allergy Immunol 1997;8:28-34.
20. Wolkerstorfer A, De Waard Van Der Spek FB, Glazenburg EJ, Mul-
der PGH, Oranje AP. Scoring the severity of atopic dermatitis: three
item severity score as a rough system for daily practice and as a pre-
screening tool for studies. Acta Derm Venereol 1999;79:356-9.
21. Feldman SR, Fleischer AB Jr, Reboussin DM, Rapp SR, Exum ML,
Clark AR et al. The self-administered psoriasis area and severity
index is valid and reliable. J Invest Dermatol 1996;106:183-6.
22. Fleischer AB Jr, Rapp SR, Reboussin DM, Vanarthos JC, Feldman SR.
Patient measurement of psoriasis disease severity with a structured
instrument. J Invest Dermatol 1994;102:967-9.
23. Housman TS, Patel MJ, Camacho F, Feldman SR, Fleischer AB jr,
Balkrishnan R. Use of the Self Administered Eczema Area and Sever-
ity Index by parent caregivers: results of a validation study. Br J Der-
matol 2002;147:1192-8.
24. Hanifin JM, Thurston M, Omoto M, Cherill R, Tofte SJ, Graeber M,
the EASI Evaluator group. The eczema area and severuty index
(EASI): assessment of reliability in atopic dermatitis. Exp Dermatol
2001;10:11-8.
25. StataCorp. 1999. Statistical Software: Release 6.0. College Station, TX:
Stata Corporation.
nelle relazioni familiari e la necessità di disporre di maggio-
ri risorse, economiche e sociali, per fronteggiarla 1-4.
Per la valutazione dell’impatto che la patologia ha sui
diversi aspetti della QoL la misura della gravità diviene un
fattore cruciale.
Agosto 2005
CHILDHOOD ATOPIC DERMATIS
Sono stati proposti diversi metodi per la misura della gra-
vità della DA 5-11. Tra i contributi più recenti ci sono la misu-
ra del funzionamento della permeabilità della barriera e del-
l’idratazione dello strato corneo che si avvalgono di algoritmi
computerizzati per la stima dell’estensione del problema 12,
o la valutazione automatica, con software dedicati, dell’e-
stensione delle lesioni cutanee 13. Sebbene sia nota la mag-
gior precisione e la riproducibilità di misure oggettive, i cli-
nici e i ricercatori utilizzano strumenti più classici 14-16.
L’introduzione dello SCORing atopic dermatitis (SCO-
RAD, European Task Force on Atopic Dermatitis) 17, uno dei
sistemi meglio validati che combina insieme la valutazione
di alcuni criteri oggettivi (ad esempio l’estensione delle
lesioni) e soggettivi (ad esempio prurito, sonno perso), ha rap-
presentato un valido criterio di riferimento per la misura del-
la gravità della DA 18-20. Benché utilizzato nella ricerca cli-
nica, richiede, tuttavia, tempi troppo lunghi per la compila-
zione e ciò ne limita l’utilizzo nella pratica clinica di routi-
ne. Inoltre gli strumenti di valutazione di gravità in derma-
tologia sono stati generalmente rivolti a pazienti adulti 21,
22. Solo recentemente è stato sviluppato il self administe-
red eczema area and severity index (SA-EASI) 23 che con-
sente ai genitori la valutazione della gravità della DA dei
figli. I risultati riportati nello studio originale 23 hanno evi-
denziato che “i genitori, utilizzando il SA-EASI, possono
valutare in modo accurato la gravità della patologia cutanea
dei figli”.
Scopo di questo contributo è stato valutare la performan-
ce della versione in italiano del SA-EASI, confrontando la
misura della gravità della DA effettuata dal genitore con
quella effettuata dal dermatologo.
Materiali e metodi
Soggetti
Hanno partecipato allo studio 35 pazienti, 16 di sesso
maschile e 19 di sesso femminile, di età compresa tra 2 mesi
e 17 anni, ricoverati presso la divisione di dermatologia
pediatrica dell’IDI-IRCCS di Roma, e almeno uno dei loro
genitori. Sono stati inclusi pazienti con diagnosi di DA con-
fermata da un dermatologo esperto. Sono stati esclusi i
pazienti con patologie concomitanti gravi e/o i cui genitori
non erano disponibili o non erano in grado di compilare il SA-
EASI.
Strumenti
Il SA-EASI è uno strumento di una pagina che consente
al genitore del bambino affetto da DA di misurare la gra-
vità della patologia del figlio. È articolato in 2 sezioni, la
prima relativa alla localizzazione e all’estensione della der-
matite, la seconda relativa alle caratteristiche cliniche, ogget-
tive e soggettive, della dermatite stessa. Nella prima sezio-
ne sono rappresentate 2 silhouette di un bambino ideale, una
vista da una prospettiva frontale, l’altra posteriore. Il genitore
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
MAZZOTTI
deve tratteggiare con una penna le aree della silhouette cor-
rispondenti all’eczema sul corpo del proprio figlio.
La stima dell’area della superficie corporea (body surface
area, BSA) coinvolta è ottenuta applicando un algoritmo a par-
tire dalla valutazione percentuale (0-100) che un ricercatore,
che non ha visto il paziente, fa dell’area indicata dal genito-
re, separatamente per la rappresentazione anteriore e poste-
riore e per 4 distinti distretti corporei: testa, arti superiori,
tronco, arti inferiori. I valori percentuali vengono trasforma-
ti in punteggi proporzionali, definiti su una scala ordinale a 7
punti, dove 0 corrisponde a “nessun eczema”, 1 corrisponde
a una superficie coinvolta del ≤9%, 2 equivale a 10-29%, 3 a
30-49%, 4 a 50-69%, 5 a 70-89%, e 6 a 90-100%. Ogni pun-
teggio di area viene moltiplicato per uno specifico peso (0,1
per la testa, 0,2 per gli arti superiori, 0,3 per il tronco, 0,4
per gli arti inferiori) e i prodotti sono poi sommati per ottenere
il punteggio totale (BSA). Per i soggetti con età inferiore ai
7 anni di età viene applicato un algoritmo modificato in cui
i pesi corrispondenti ai distretti corporei sono rispettivamen-
te 0,2 per la testa, 0,2 per gli arti superiori, 0,3 per il tronco,
0,4 per gli arti inferiori. In entrambi i casi la funzione appli-
cata è consistente con la “regola del nove” 24.
La seconda sezione è costituita da 5 scale analogo visive
(visual analogue scales, VAS) di 100-mm. La VAS è costi-
tuita da una linea continua su cui il genitore deve segnare il
punto che corrisponde alla gravità media di una lesione del-
la DA. Sulla linea i livelli estremi e intermedi sono posti a
distanze equivalenti. Le scale VAS consentono al genitore di
descrivere l’intensità media di arrossamento (nessun rosso-
re, lievemente rosa, rosa, rossa, rosso scuro), di ispessimento
(da non ispessita a molto), di secchezza (da non secca a
estremamente secca), di lesioni da graffiamento (da nessun
graffio a molti graffi), di prurito (da nessun prurito a pruri-
to severo).
Il punteggio di gravità è ottenuto applicando un’equazio-
ne che combina i punteggi derivati dalle 2 sezioni 23.
In questo studio è stata utilizzata una versione in cui i
pesi specifici per distretto corporeo sono moltiplicati diret-
tamente per i punteggi percentuali di area.
L’autorizzazione per tradurre, adattare e utilizzare lo stru-
mento in lingua italiana è stata gentilmente concessa dal-
l’Autore. Dal riferimento originale l’immagine di un uomo
è stata modificata in quella di un bambino e le 5 scale, esat-
tamente di 100 mm, sono state graduate con linee di mede-
sima altezza, di 3 mm, collocate a una distanza di 10 mm l’u-
na dall’altra. È stata effettuata una retro traduzione per veri-
ficare la congruità della versione italiana con quella in ingle-
se.
Lo SCORAD 17 è uno strumento per la valutazione della
gravità clinica della DA completato dal dermatologo. È arti-
colato in 3 sezioni. La prima è inerente alla valutazione del-
l’estensione e alla localizzazione della dermatite. La secon-
da è relativa alla valutazione dell’intensità di: eritema, ede-
ma/papule, siero/crosta, escoriazioni, lichenificazione e sec-
chezza; quest’ultima è valutata sulle aree non coinvolte dal-
la dermatite, mentre le altre caratteristiche sono valutate sul-
l’area maggiormente interessata dall’eritema. La terza sezio-
393
MAZZOTTI
ne è relativa ai sintomi soggettivi di prurito e perdita di son-
no negli ultimi 3 giorni.
Nella prima sezione sono rappresentate 2 silhouette di un
bambino ideale, una vista da una prospettiva frontale, l’altra
posteriore. Il dermatologo deve tratteggiare con una penna le
aree della silhouette corrispondenti all’eczema sul corpo del
bambino.
La stima della BSA coinvolta è ottenuta applicando un
algoritmo a partire dalla valutazione percentuale (0-100)
che un ricercatore, che non ha visto il paziente, fa dell’area
indicata dal clinico, separatamente per la rappresentazione
anteriore e posteriore e per 4 distinti distretti corporei: testa,
arti superiori, tronco, arti inferiori.
Successivamente i punteggi di area sono moltiplicati per
uno specifico peso (9 per la testa, 18 per gli arti superiori, 36
per il tronco, 37 per gli arti inferiori). Valori diversi sono
proposti per i soggetti con meno di 2 anni di età 17. Infine i
prodotti sono sommati per ottenere il punteggio BSA.
Il punteggio totale è derivato dall’applicazione della for-
mula estensione/5+7intensità/2 in cui l’estensione pesa per
il 25% e l’intensità per il 75% del punteggio totale. Il range
teorico del punteggio è compreso tra 0 e 83. La gravità del-
la DA è classificata in lieve (<15), moderata (tra 15 e 40) e
grave (>40).
In questo studio è stata utilizzata la versione che esclude
il conteggio dei criteri soggettivi relativi al prurito e al son-
no perso 17.
Entrambe le valutazioni sono state effettuate al ricovero.
Il dermatologo (AP), dopo aver visitato il bambino e com-
pilato lo SCORAD, consegnava al genitore il SA-EASI.
Analisi statistica
Le correlazioni sono state calcolate tra i punteggi totali SA-
EASI e SCORAD, e separatamente tra estensione, distretti
corporei e gravità. Come indice di correlazione è stato cal-
colato il coefficiente ρ di Spearman (rs).
La stabilità è stata misurata, separatamente per i 2 stru-
menti, in 23 soggetti con una seconda somministrazione a
distanza di 48 h.
Le valutazioni, riportate in percentuale, sono state fatte
separatamente per distretto corporeo (testa, arti superiori,
tronco, arti inferiori) e piano d’osservazione (frontale, dor-
sale). Sono stati, quindi, applicati gli specifici algoritmi per
la misura delle BSA:
BSASA-EASI=(0,1*Ah)+(0,2*Au)+(0,3*At)+(0,4*Al)
BSASCORAD=(9*Ah)+(18*Au)+(37*At)+(36*Al)
(Ah= area testa, Au= area arti superiori, At= area tronco,
Al= area arti inferiori(
Per l’analisi statistica è stato utilizzato il software PC-
STATA 25.
Risultati
Trentacinque soggetti (51% di sesso femminile) di età
compresa tra 2 mesi e 17 anni (media= 7,7; ds= 5,07; media-
394 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHILDHOOD ATOPIC DERMATIS
na= 8) hanno partecipato allo studio. Il 61,6% sono risulta-
ti affetti da DA moderata (SCORAD tra 15 e 40) e il 38,4%
da DA grave (SCORAD >40). Il range di variazione del pun-
teggio totale SCORAD al momento del ricovero è compre-
so tra 15,2 e 70,3, con un punteggio medio di 38,8 (ds=13,03)
e mediano di 36,1.
Per valutare le caratteristiche del SA-EASI, come misu-
ra della gravità della DA, sono stati confrontati i punteggi otte-
nuti con quelli derivati dallo SCORAD.
Sono stati calcolati gli indici di correlazione tra i 2 pun-
teggi totali, tra i punteggi relativi alle misure di BSA e tra i
punteggi relativi ai 4 distretti corporei. Tutti gli indici sono
risultati soddisfacenti. Tra i punteggi totali la correlazione è
risultata 0,71 (P<0,0001), e solo leggermente inferiore tra le
2 misure di BSA (rs=0,68; P<0,0001). Le correlazioni tra
ogni coppia dei 4 distretti corporei variano tra 0,59 e 0,70
(superficie testa rs=0,61; superficie estremità superiori
rs=0,61; superficie tronco rs=0,59; superficie estremità infe-
riori rs=0,70; P<0,0001).
Dal confronto delle caratteristiche specifiche delle lesio-
ni emergono relazioni basse che suggeriscono una limitata
concordanza tra le misure; tra arrossamento ed eritema l’in-
dice di relazione è rs=0,42 (P=0,0013), tra graffi e escoria-
zione rs=0,43 (P=0,0010), tra secchezza rs=0,29. Per le carat-
teristiche di ispessimento e prurito, considerate dal SA-
EASI, e quelle di lichenificazione, presenza di crosta/siero
e infiammazione o formazione di papule dello SCORAD,
per le quali non vi è corrispondenza tra i 2 strumenti, le cor-
relazioni sono inferiori a 0,10.
Per valutare la capacità dei 2 strumenti di registrare un
cambiamento nello stato clinico del paziente è stato propo-
sto al dermatologo e ai genitori di un gruppo di 23 soggetti,
13 di sesso femminile e 10 di sesso maschile, di età compresa
tra 3 mesi e 17 anni (media 8,03, ds 5,22, mediana 8,11), di
compilare il SA-EASI e lo SCORAD al momento del rico-
vero e a distanza di 48 h o più. La durata dell’intervallo mini-
mo è stata scelta per bilanciare l’effetto della possibile distor-
sione dovuta al ricordo della prova precedente che può veri-
ficarsi quando l’intervallo tra le 2 è troppo breve.
Nella Tabella I sono riportate le caratteristiche dei punteggi
totali dei 2 strumenti al tempo 1 e al tempo 2. Nella Tabella
I, la differenza, delta, tra il punteggio al momento del rico-
vero e quello successivo evidenzia un miglioramento della
sintomatologia durante il ricovero, sia nei punteggi SCO-
RAD sia in quelli SA-EASI. Il punteggio differenziale nega-
tivo nel range dello SCORAD identifica un soggetto che tra
i 2 momenti della misura ha avuto un aggravamento dell’a-
rea interessata dalla patologia. Il 52,4% dei soggetti pre-
sentano a distanza di pochi giorni dal ricovero un quadro di
gravità lieve (SCORAD<15), il 42,9% ancora moderata, e il
4,8% grave.
Discussione
Il SA-EASI si è dimostrato uno strumento equivalente
allo SCORAD per misurare la gravità della DA infantile.
Agosto 2005
CHILDHOOD ATOPIC DERMATIS
La versione italiana adattata è risultata facilmente e imme-
diatamente comprensibile a tutti i genitori coinvolti. I risul-
tati di questo studio rappresentano un importante contribu-
to preliminare in quanto evidenziano la capacità del genito-
re di valutare in modo accurato, e sufficientemente sovrap-
ponibile a quello del dermatologo, l’estensione e la gravità
della DA del proprio figlio utilizzando uno strumento strut-
turato. La valutazione fatta dal genitore non si discosta da
quella fatta dal dermatologo, tutte le correlazioni indicano una
sostanziale convergenza e sono del medesimo ordine di gran-
dezza.
Come atteso i 2 strumenti mostrano delle differenze per
quanto riguarda le caratteristiche specifiche delle lesioni
peculiari di ogni strumento, tuttavia quest’assenza di paral-
lelismo non sembra influenzare i punteggi totali.
Conclusioni
In conclusione, sembra che i risultati di questo studio
indirizzino verso l’individuazione di uno strumento, il SA-
EASI, che ha il vantaggio - comune a molti strumenti auto-
somministrati - di non richiedere l’intervento di personale spe-
cializzato, che non necessita di addestramento, che consen-
te, attraverso una rapida e facile compilazione, di valutare le
manifestazioni di intensità ed estensione della patologia. Il
SA-EASI potrebbe essere lo strumento di comunicazione, sul-
la patologia, tra paziente (o genitore) e medico curante, e
rappresentare, per quest’ultimo, il mezzo per seguire l’an-
damento della patologia.
L’altro strumento, lo SCORAD, presenta caratteristiche
analoghe, e il suo impiego nella routine consentirebbe al cli-
nico una valutazione più puntuale e oggettiva, non soltanto
in quei contesti nei quali il tempo a disposizione del medi-
co è poco, ad esempio le visite ambulatoriali, ma anche nel
rapporto con il paziente cronico, seguito nel tempo.
Rispetto al SA-EASI, lo SCORAD si caratterizza come
una misura più accurata in quanto il dermatologo con la sua
esperienza professionale è in grado di differenziare le diver-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
MAZZOTTI
se presentazioni cliniche della patologia. Tuttavia il SA-
EASI presenta il vantaggio secondario di coinvolgere atti-
vamente, nella valutazione e nel monitoraggio, la famiglia del
bambino affetto da patologia cronica.
I punteggi di entrambi gli strumenti riflettono il diverso sta-
to clinico del paziente anche se saranno necessari altri stu-
di, su campioni più numerosi, per confermare la capacità di
registrare cambiamenti nella patologia, sia nel tempo, sia
come risposta ai trattamenti.
Riassunto
Obiettivo. Gli strumenti per misurare la gravità della der-
matite atopica (DA) richiedono tempo sia per la compila-
zione, sia per l’attribuzione del punteggio e sono, quindi,
poco utilizzati nella pratica clinica abituale. Il self admini-
stered eczema area severity index (SA-EASI) è stato svi-
luppato e validato per la valutazione, da parte dei genitori, del-
la gravità della DA dei figli. Lo scopo di questo studio era ana-
lizzare la relazione tra SA-EASI e SCORing atopic derma-
titis (SCORAD).
Metodi. Hanno partecipato allo studio 35 pazienti, rico-
verati presso la divisione pediatrica di un ospedale derma-
tologico di Roma, e almeno uno dei loro genitori. La gravità
della patologia è stata valutata, al momento del ricovero in
ospedale, separatamente dal genitore, con il SA-EASI, e dal
dermatologo, con lo SCORAD.
Risultati. La validità concorrente è espressa dalla corre-
lazione tra i punteggi totali dello SCORAD e del SA-EASI
(ρ di Spearman=0,71; P<0,001).
Conclusioni. Entrambi gli strumenti sono utili nella pra-
tica clinica quotidiana e nella ricerca sugli outcome. Il SA-
EASI ha il vantaggio di non richiedere l’intervento di per-
sonale specializzato e non necessita di addestramento; favo-
risce, inoltre, un maggior coinvolgimento dei familiari.
Parole chiave: Dermatite atopica - Estensione - Gravità -
Misura dell’accordo - SA-EASI - SCORAD - Validità con-
corrente.
395
 REVIEWS
G ITAL DERMATOL VENEREOL 2005;140:397-406

Extracellular matrix protein 1:

a newly discovered glycoprotein
with an important role in skin biology
I. CHAN 1, T. HAMADA 2, N. OYAMA 3, V. WESSAGOWIT 1, J.A. McGRATH 1

Extracellular matrix protein 1 (ECM1) is a glycoprotein found
in many tissues, including skin. First discovered in 1994, its
function in skin biology was largely unknown until 2002 when
it was identified as the candidate gene/protein for the autoso-
mal recessive disease, lipoid proteinosis. This inherited disor-
der is characterised clinically by skin and mucosal infiltra-
tion and scarring, and histologically by disruption or dupli-
cation of basement membrane, as well as widespread deposi-
tion of hyaline material in the dermis. Over 30 pathogenic
mutations in the ECM1 gene have been characterised, with
recurrent mutations, ancestral alleles, genotype-phenotype
correlation and new diagnostic techniques now established
for this rare genodermatosis. Further insight into the role of
ECM1 in human skin was revealed in 2003 with the discovery
of circulating autoantibodies against the ECM1 protein in the
sera of most patients with lichen sclerosus, a common chron-
ic inflammatory condition that shares some clinicopatholog-
ical features with lipoid proteinosis. These autoantibodies have
been characterised and the immunodominant epitope isolat-
ed, and a new ELISA test for lichen sclerosus is currently
being evaluated. Protein-protein interaction studies have iden-
tified that ECM1 binds to the major heparan sulphate pro-
teoglycan, perlecan, as well as to matrix metalloproteinase 9,
epidermal growth factor, and legumain. These findings, in
combination with the lipoid proteinosis and lichen sclerosus
data, suggest that ECM1 has a key role in several aspects of epi-
dermal differentiation, maintaining dermal architecture, and
regulating basement membrane composition. Clearly, the new-
Funding from the Charitable Foundation of Guy’s and St Thomas’
Hospitals and the British Skin Foundation for several of the original stu-
dies referred to in this review is gratefully acknowledged.
Address reprint requests to: J. McGrath, Genetic Skin Disease Group,
St John’s Institute of Dermatology, St Thomas’ Hospital, Lambeth Pala-
ce Road, London SE1 7EH, UK. E-mail: john.mcgrath@kcl.ac.uk
1Genetic Skin Disease Group
St John’s Institute of Dermatology
Division of Skin Sciences
Guy’s, King’s College
and St Thomas’ Hospitals’ Medical School,
St Thomas’ Hospital, London, UK
2Department of Dermatology
Kurume University School of Medicine, Kurume, Japan
3Department of Dermatology
Fukushima Medical University School of Medicine
Fukushima, Japan
ly discovered glycoprotein ECM1 has an important function
in skin biology.
KEY WORDS: Gene mutation - Autoantibodies - Lipoid proteinosis
- Lichen sclerosus - Epidermis - Dermis - Basement membrane.
The discovery of extracellular matrix protein 1
E xtracellular matrix protein 1 (ECM1) is a glyco-
protein that was first discovered in 1994. In a
study relevant to bone matrix biology, Mathieu et al.
analysed the proteins secreted by a clonal osteogenic
stromal cell line, MN7, derived from mouse bone mar-
row, using two-dimensional polyacrylamide gel elec-
trophoresis, Western blotting and microsequencing.1
Amongst several proteins identified, a novel glycosy-
lated 85-kDa protein with an average isoelectric point
of 5.7 was isolated.1 This protein was initially desig-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 397

CHAN
nated p85 based on its size, but was also named ECM1
because it was discovered amidst various other con-
nective tissue proteins including type I collagen,
osteonectin, cathepsin and sialo bone protein.1, 2 The
human homologue was later identified in 1997.3, 4
Most of the early functional studies on ECM1 con-
centrated on its role in tissues other than skin, and
involvement of ECM1 in bone and cartilage develop-
ment, angiogenesis and certain malignancies, was
demonstrated. For example, ECM1 was found to be a
negative regulator of endochondral bone formation,
inhibiting alkaline phosphatase activity and mineral-
isation.5 ECM1 was also shown to be able to stimulate
blood vessel endothelial cell proliferation (in culture),
and to promote angiogenesis (in chicken embryos).6
Presence of ECM1 was also demonstrated in the stro-
ma of 2 human breast cancer cell lines, MDA-435 and
LCC15.6 Additionally, increased ECM1 expression,
revealed by microarray experiments, has been report-
ed in cartilage formation, dendritic cell differentia-
tion and maturation, and in grade I, II and IV glioblas-
toma multiformes.7-9 Some experimental data for a
potential role for ECM1 in the skin, notably in epi-
dermal differentiation, were also postulated,2, 10 but
the key role of ECM1 in multiple aspects of cutaneous
biology was not immediately apparent.
Protein structure
of extracellular matrix protein 1
The ECM1 protein consists of 3 isoforms, ECM1a,
ECM1b and ECM1c, of 540, 415 and 559 amino acids,
respectively. ECM1 contains a signal peptide of 19
amino acids followed by 4 functional domains: a cys-
teine-free N-terminus, 2 tandem repeats and a C-ter-
minus.2, 3, 11 The latter 3 domains contain numerous cys-
teine residues, arranged in a specific manner. The cys-
teine distribution and structure in humans is almost
identical to its mouse counterpart, containing 28 cys-
teine residues, although there is just one amino acid less
in the human protein.3 Significantly, the cysteine-con-
taining domains all have the typical CC-(X7-10)C
arrangement that is capable of forming protein double
loops involved in protein-protein interactions.2, 3 ECM1
contains 1 double-loop domain within each of the 2 tan-
dem-repeats and 1 in the C-terminal domain.11 Dif-
ferent double loops can have varying binding affinities,
increasing the potential for interactions with a vari-
398 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
EXTRACELLULAR MATRIX PROTEIN 1
ety of biological ligands.12 The CC-(X7-10)C motif is
also present in the serum albumin family of proteins
and shows structural similarities to the Endo 16 cal-
cium-binding protein of sea urchin.2, 13 The specific
motif may enable ECM1 to serve as a transporter pro-
tein or to be involved in binding growth or differenti-
ation factors.10 Indeed, yeast-two-hybrid studies have
shown that ECM1 can bind to several other proteins,
including the major heparan sulphate proteoglycan,
perlecan, matrix metalloproteinase 9 (MMP-9, type
IV collagenase), epidermal growth factor (EGF), and
legumain.11, 14 The ECM1 protein also contains 3 N-
glycosylation sites for protein kinase C and several
phosphorylation sites for casein kinase II.3 ECM1 also
contains a calcium-binding domain, which is present
in the ECM1a and ECM1c isoforms but not in
ECM1b.2, 10
Gene structure and expression
of extracellular matrix protein 1
The ECM1 gene has been mapped to chromosome
1q21.2 3, 4 and it has 3 known splice variants, ECM1a,
ECM1b and ECM1c.3, 11 ECM1a is encoded by a 10-
exon gene, whereas ECM1b lacks exon 7 and ECM1c
contains an additional exon 5a within intron 5. Exon
5a is homologous to the sixth mouse exon that was
initially thought to be absent in the human gene.11
ECM1a is the most widely expressed splice variant. It
is found in various tissues including skin, liver, small
intestines, lung, ovary, prostate, testis, skeletal muscle,
pancreas and kidney, but gene expression is greatest in
placenta and heart. By contrast, ECM1b has a much
more restricted expression pattern, being detectable
only in tonsils and keratinocytes.3 The full pattern of
ECM1c expression has yet to be determined, but in
skin it accounts for approximately 15% of total ECM1
RNA.11 Apart from the functions identified for the
ECM1 protein, sequence analysis also predicts that
ECM1 is a positive regulator of the I-κB kinase/NF-κB
cascade. The precise localisation of the ECM1 gene
maps just centromeric to the epidermal differentiation
complex region,10 a locus that contains a cluster of 3
families of genes involved in epidermal differentia-
tion.3 Nevertheless, ECM1 may have a role in termi-
nal keratinocyte differentiation, as suggested by stud-
ies demonstrating expression of ECM1a within basal
keratinocytes and ECM1b in suprabasal cells.10
Agosto 2005
EXTRACELLULAR MATRIX PROTEIN 1 CHAN

Figure 1.—Skin and mucous membrane features of lipoid proteinosis in (A) a 36-year-old man with severe inflammation, erosions and thickening of
the oral mucosa and skin infiltration and scarring on his face, and (B) a 10-year-old boy with infiltration of the lips and tethering of the tongue with a
thickened frenulum and reduced tongue movement. There is also papular infiltration of his facial skin.

Extracellular matrix protein 1 gene mutations
in lipoid proteinosis
Lipoid proteinosis (OMIM 247100), also known as
Urbach-Wiethe disease or hyalinosis cutis et mucosae,
is a rare autosomal recessive disorder typified by gen-
eralised thickening and scarring of the skin and
mucosae (Figure 1).15, 16 Other characteristic clinical
features include beaded eyelid papules, waxy yellow
skin papules and nodules, and most notably, a hoarse
voice from infancy.16 Increased skin scaling and thick-
ening occurs in regions exposed to mechanical friction
including elbows, hands and knees. In 2002, genome-
wide linkage analysis using genomic DNA from con-
sanguineous families with lipoid proteinosis, was
reported.17 Screening mapped lipoid proteinosis to a
2.3-cM interval on the long arm of chromosome 1, at
1q21.2.17 A candidate gene approach was then used,
presupposing that, as in many other recessive geno-
dermatoses, the lipoid proteinosis gene product would
show reduced expression in dermal fibroblasts com-
pared to normal control fibroblasts. Following this
rationale, the gene for lipoid proteinosis was identified
as ECM1. Sequencing of genomic DNA from 6 con-
sanguineous families disclosed the presence of
homozygous loss-of-function mutations (nonsense,
frameshift or internal deletions) in all cases.17 Reduced
ECM1 protein expression in lipoid proteinosis skin
was also noted.17 Thereafter, the molecular basis of
lipoid proteinosis has been determined in over 50
patients world-wide and 31 different pathogenic muta-
tions (including unpublished data) have been identified
(Figure 2).18-22 Mutations have been identified in every
exon, apart from exon 5a. The majority of mutations
occur in exons 6 and 7, with 9 mutations occurring in
exon 7, and 10 mutations occurring in exon 6. Most of
the mutations are nonsense or frameshift mutations,
presumably resulting in truncation of the ECM1 pro-
tein, and/or low levels of the corresponding mRNA
through nonsense-mediated mRNA decay mecha-
nisms. In addition, there are 4 missense mutations:
V10G in exon 1 and F167I,18 F167L and L210P in
exon 6 (including unpublished data). All 4 of these
amino acid changes represent substitution of one
hydrophobic neutral amino acid by another, but the
relative sizes of the amino acids are different and there-
fore the functional conformation of ECM1 may be
altered. All these missense changes have been exclud-
ed as rare polymorphisms and the substituted amino
acids are highly conserved residues. The missense
mutation V10G occurs at the start of the ECM1 gene
in the region coding for the signal peptide, whereas
F167I, F167L and L210P all occur in the first tandem-
ly-duplicated domain of ECM1. A donor splice site
mutation, 80+1G>A, has also been identified (unpub-
lished data).
Initial genotype-phenotype correlation suggested
that mutations occurring outside exon 7 were associ-
ated with a slightly more severe mucocutaneous phe-
notype, but this has not been borne out in more detailed

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 399

CHAN EXTRACELLULAR MATRIX PROTEIN 1

Q95X/1432delA
F167L
1019delA C220G/R476X
501insC Q276X Q346X
Q32X Q114X 541 735delTG W359X 1253delC
80+1G>A del3 R243X
V10G ins16
R243X
5a

1 2 3 4 5 6 7 8 9 10

L210P
Q197X
542insAA/R243X
R53X W160X/F167I 1163-bp deletion
735delTG
243delG 507delT H1190insC
E248X
785delA
892delC

Figure 2.—Schematic representation of all known pathogenic mutations in the ECM1 gene in lipoid proteinosis. Double arrows indicate homozygous
mutations; joined arrows depict compound heterozygosity.

analyses.18 Specifically, considerable inter-individ-
ual variability has been shown, for example in 29
South African subjects, all with the same homozy-
gous mutation, Q276X, in exon 7.20 This mutant allele
was thought to have been propagated by a German
settler (from Cologne) to the Northern Cape in the
1650s.20, 23-25 Other mutated ancestral ECM1 alleles
have also been identified, for example 501insC in
Northern Europe and W359X in Scotland. The muta-
tion 507delT, however, appears to be a hotspot muta-
tion, having occurred on different genetic backgrounds
in 2 Thai brothers, a Canadian Iranian family, a Japan-
ese individual, and an Indian girl with lipoid pro-
teinosis.18, 21
New diagnostic test for lipoid proteinosis
Lipoid proteinosis can be difficult to diagnose in
early life. With time, it can usually be diagnosed clin-
ically but the early manifestations of lipoid proteinosis
are protean and may overlap with other diseases,
including subtypes of porphyria. Having identified
ECM1 as the lipoid proteinosis gene, however, it is
now possible to screen DNA in suspected cases to
establish the diagnosis, notwithstanding that this may
be time-consuming and not readily available in many
diagnostic laboratories. Alternatively, one comple-
mentary approach to the diagnosis of many recently
characterised genodermatoses has involved skin
immunohistochemistry. Indeed, many of the severe,
usually autosomal recessive, single gene disorders
typically involve loss-of-function mutations leading
to reduced or absent expression of the encoded protein.
Such changes can often be detected through dimin-
ished immunohistochemical labelling of skin sections
using an antibody to the corresponding protein. This
approach has proved to be very useful in the rapid
diagnosis of other genodermatoses such as epider-
molysis bullosa, lamellar ichthyosis, Netherton’s syn-
drome and Kindler syndrome.26, 27 To develop an
immunohistochemical test for lipoid proteinosis, a
rabbit polyclonal antibody to human ECM1 was raised
against the oligopeptide SGDTENAKGQGEQGSTG,
encoding the carboxyl-terminal of the human ECM1
protein.28 Immunolabelling with this antibody was
reduced in lipoid proteinosis skin, confirming its use-
fulness as a diagnostic probe (Figure 3). However, the
pattern of labelling was also able to provide clues as to
where the pathogenic mutation might lie.28 Specifi-
cally, attenuated but not absent labelling suggested
that the pathogenic mutation was in exon 7, whereas

400 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

EXTRACELLULAR MATRIX PROTEIN 1 CHAN

Figure 3.—Immunohistochemical labelling of (A) normal control skin (B) skin from a patient with lipoid proteinosis caused by the homozygous muta-
tion Q276X in exon 7 of the ECM1 gene, using a polyclonal anti-ECM1 antibody. In (A), there is intracellular and cell-surface labelling in the lower
epidermis, particularly in basal keratinocytes and to lesser extent, in suprabasal keratinocytes. (B) shows attenuated, though still detectable, immunos-
taining. (Bar = 50 µm).

absence of staining indicated a mutation elsewhere in atinocytes, a homogeneous appearance of collagen

ECM1. Thus, lipoid proteinosis can now be diagnosed
quickly and reliably by skin immnuohistochemistry.
Extracellular matrix protein 1 autoantibodies
in lichen sclerosus
Lichen sclerosus is a chronic inflammatory disorder
of skin of unknown aetiology.29, 30 Its features are vari-
able, including white papules and plaques, skin atro-
phy and scarring, mostly in genital skin but also involv-
ing extragenital sites (Figure 4). The prevalence of
lichen sclerosus is estimated to be up to 1 in 300 with
a ratio of affected females to males of approximately
10:1.30 There is a strong association with autoimmune
diseases such as vitiligo, alopecia areata, thyroid dis-
ease and pernicious anaemia and there is a positive
association with HLA class II antigen DQ7.30, 31 This
suggests that part of the disease aetiology or patho-
logical process in lichen sclerosus may involve gen-
eration of autoantibodies to one or more antigens in
skin. The possibility of ECM1 representing a putative
target for humoral immunity in lichen sclerosus is
highlighted by the dermatopathological abnormalities
in this disorder. Notably, the histology of lichen scle-
rosus includes hydropic generation of basal ker-
within the upper dermis (hyalinisation) and disrup-
tion of basement membranes. Collectively, many of
the alterations in the epidermis, dermis and dermal
blood vessels show some histological overlap with
lipoid proteinosis, raising the possibility that ECM1
may be a target antigen in lichen sclerosus. To inves-
tigate this further, immunoblotting, using a full-length
fusion protein for ECM1, was used to demonstrate
presence of circulating autoantibodies to ECM1 in the
sera of most patients with lichen sclerosus (74%, as
compared with 7% of controls).32 These antibodies
were present at low titre, since indirect immunofluo-
rescence microscopy was positive in only one case.
However, when the lichen sclerosus sera were affini-
ty purified (i.e. concentrated approximately 25 times),
a specific pattern of skin immunostaining was identi-
fied. This staining pattern was very similar to the
appearances of skin labelling with an antibody to
ECM1. Moreover, demonstration that the lichen scle-
rosus sera were targeting ECM1 was confirmed by
ablation of labelling following preabsorption of the
sera with recombinant ECM1 protein.32
Further immunoblotting studies with fragments of
ECM1 recombinant protein were able to show that
lichen sclerosus sera react with multiple ECM1 epi-
topes, the immunodominant epitope being between

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 401

CHAN
Figure 4.—Atrophic shiny, pale white plaques of extragenital lichen sclerosus on the trunk of (A) a 38-year old man and (B) a 65-year-old woman.
amino acids 359 and 559 within the distal second tan-
dem repeat and the carboxyl-terminus of ECM1.33, 34
The anti-ECM1 IgG subclass is predominantly IgG2,
with almost 90% of sera containing IgG2 anti-ECM1
autoantibodies, either alone or in combination with
other subclasses.33 These antibodies are not just an
epiphenomenon since they are not observed in other
sclerosing or basement membrane diseases.32 Fur-
thermore, passive transfer studies with affinity-purified
lichen sclerosus IgG autoantibodies injected intrader-
mally into the ears of neonatal mice have shown that,
compared to control injected sites, lichen sclerosus
IgG injection causes the mice to scratch their ears.
Macroscopically, at 28 days, there is swelling and ery-
thema, and microscopically there is oedema, a patchy
mononuclear inflammatory cell infiltrate in the dermis,
focal pigmentary incontinence in the superficial der-
mis and dilatation of some superficial blood vessels.34
These changes do not fully recapitulate the full histo-
logical features of lichen sclerosus (i.e. no hyalinosis)
but are fully consistent with early histological changes
seen in this disease.
New ELISA test for lichen sclerosus
High throughput diagnostic ELISA measurement of
serum autoantibodies has become an established part
of the investigation of several autoimmune skin dis-
orders, such as pemphigus and bullous pemphigoid.35
Moreover, the antibody titres determined by ELISA
in these diseases may have clinical implications for
optimal patient management. To assess whether there
402 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
EXTRACELLULAR MATRIX PROTEIN 1
might be any correlation between anti-ECM1 anti-
body titres and disease parameters in lichen sclerosus,
a diagnostic ELISA was recently reported.34 The pro-
tein for the ELISA in this study was based on the
immunodominant epitope, i.e. the distal second tan-
dem repeat and carboxyl-terminus of ECM1. This
ELISA test exhibited a high sensitivity of 80% (76
of 95 sera were positive) and a high specificity of
93.7% in discriminating lichen sclerosus from nor-
mal controls and other autoimmune basement mem-
brane or sclerosing diseases, thus establishing this
ELISA as a useful diagnostic test (Figure 5).34 Clin-
ically, higher ELISA titres correlated with more long-
standing disease, with cases that were refractory to
treatment, and with cases complicated by squamous
cell carcinoma.34 This ELISA, therefore, may be use-
ful in detecting individuals with lichen sclerosus who
need more aggressive immunosuppressive therapy or
who may be at risk for disease complications, such as
malignancy or extensive scarring. Detection of the
antibodies by ELISA may also provide a means of
assessing the potential benefits of new treatments for
lichen sclerosus, such as immunoadsorption thera-
pies in which recombinant ECM1 protein might be
used to remove circulating anti-ECM1 antibodies
from the sera of patients with lichen sclerosus. Cur-
rently, however, no prospective studies examining
how the anti-ECM1 antibody fluctuates with disease
activity and time in individual patients have been
reported, and these will be extremely important in
establishing the potential usefulness of the ECM1
antibody ELISA in clinical practice.
Agosto 2005
EXTRACELLULAR MATRIX PROTEIN 1 CHAN

1.2
EPIDERMAL
Arbitrary unit of ELISA (405 nm)

1.0 DIFFERENTIATION

0.8

0.6

0.4 ECM1 ADHESION

Cut off
(0.328)
0.2

0.0
Normal LS SLE BP SSc
(n=161) (n=95) (n=70) (n=72) (n=15) MMP-9
Figure 5.—ELISA for anti-ECM1 antibodies assessed in 95 patients with
lichen sclerosus (LS), as well as 318 control subjects, comprising 161
normal volunteers, and 70 systemic lupus erythematosus (SLE), 72 bul-
lous pemphigoid (BP), and 15 systemic sclerosis (SSc) individuals. These
data show that the assay is highly sensitive and specific for lichen sclero- ECM1 ECM1 PERLECAN
sus.
BINDING
The function of extracellular matrix protein 1 TO
in human skin COLLAGENS
ELASTIC
The discovery that loss-of-function mutations in
FIBRES
ECM1 result in lipoid proteinosis provided the first
GROWTH
clinical indication of the possible relevance of the BLOOD FACTORS
ECM1 protein to skin adhesion, epidermal differenti- VESSEL
ation, wound healing, scarring, angiogenesis and base- ENDOTHELIAL CELL
ment membrane integrity. Histologically, skin from PROLIFERATION
patients with lipoid proteinosis shows hyperkeratosis,
basement membrane thickening at the dermal-epider- Figure 6.—Illustration of the possible functions of ECM1 in human skin
mal junction and around blood vessels and adnexal biology.
epithelia, as well as presence of hyaline material in
the dermis.16 This suggests that a lack of ECM1 may
influence the normal pattern of epidermal differenti- ation and, therefore, the epidermal atrophy seen in
ation, as well as disrupting dermal physiology. It is lichen sclerosus may be due to changes in the dynam-
plausible that one of the main functions of ECM1 in the ics of normal keratinocyte maturation.10, 32 In the der-
dermis is to act as some form of biological glue, help- mis, the basement membrane thickening and hyaline
ing to regulate basement membrane and interstitial appearance to collagen could reflect perturbations in
collagen fibril macro-assembly and growth factor bind- the normal binding of ECM1 to proteoglycans, such as
ing.19 In lipoid proteinosis, the glue is defective or perlecan.11, 32 Notably, co-localisation between ECM1
simply missing, leading to dysregulation of dermal and perlecan has been demonstrated in skin basement
homoeostasis and clinical features of skin infiltration membranes, in dermal blood vessels, and surround-
and scarring.19 The clinical features seen in lichen ing adnexal epithelia and it has been shown that the car-
sclerosus may also help illustrate the function of ECM1 boxy-terminus of ECM1 interacts with the EGF-like
in human skin, i.e. antibodies to ECM1 disrupt the modules flanking the LG2 subdomain of perlecan
normal function of the protein and thereby illustrate domain V.11 Perlecan is also known to bind to
what the ECM1 protein normally protects against or fibronectin, laminin, type IV collagen, fibulin 2, dys-
prevents. ECM1 has a role in keratinocyte differenti- troglycan, platelet-derived growth factor 7 and fibrob-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 403

CHAN
ECM1 gene ECM1 protein
5a
Cysteine tandem tandem
1 23 45 6 7 8 9 10 N free repeat repeat terminal
domain 1 2
Signal peptide
Inherited mutations Acquired autoantibodies
Lipoid proteinosis Lichen sclerosus
Figure 7.—Illustration of the ECM1 gene and protein and the corre-
sponding disease associations. Loss-of-function mutations in the ECM1
gene result in lipoid proteinosis whereas autoantibodies to ECM1 are
found in the sera of most patients with lichen sclerosus. In both these dis-
orders, light microscopy reveals a hyaline appearance to the papillary der-
mis with dilated superficial blood vessels.
last-growth factor binding protein.11, 36 Disruption to
these normal associations could account for the hya-
line abnormalities seen in lichen sclerosus.32 Perhaps
disruption of normal expression of ECM1 in blood
vessels could also explain the ecchymoses seen in
lichen sclerosus. Moreover, the link between ECM1
expression and certain malignant tumours might pro-
vide a partial explanation for the increased incidence
of squamous cell carcinoma in lichen sclerosus.25
ECM1 also interacts with legumain, MMP-9, and
EGF.14 Legumain is a protease linked to aspects of
epidermal differentiation;37 EGF is a critical compo-
nent of several signalling cascades, including calci-
um response pathways;38 and MMP-9 is a metallo-
proteinase with a key role in basement membrane and
interstitial collagen remodelling as well as vasculari-
404 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
EXTRACELLULAR MATRIX PROTEIN 1
sation.39 Legumain and MMP-9 interact with ECM1
only at the 2 tandem repeat domains, while EGF only
C-
interacts at the cysteine-free and C-terminal regions of
C ECM1.14 Collectively, disruption of these protein-pro-
domain
tein interactions with ECM1 in either lipoid proteinosis
or lichen sclerosus may provide an explanation for
some of the clinicopathological abnormalities in both
these conditions (Figures 6, 7). Indeed, similar dis-
ruption of basement membrane at the dermal-epider-
mal junction and around dermal blood vessels has
been identified in both disorders.40
Conclusion
It is becoming clear that ECM1 has an important
role in the anatomy and biology of normal human skin.
Although a precise role has yet to be fully elucidated,
several clues to its function have been highlighted by
its disease associations in the rare genodermatosis,
lipoid proteinosis, and the common acquired inflam-
matory skin disorder, lichen sclerosus, and in its pro-
tein-protein interactions with several important regu-
lators of epidermal and dermal homeostasis.
Riassunto
Proteina 1 della matrice extracellulare: una nuova glico-
proteina fondamentale nella biologia cutanea
La proteina 1 della matrice extracellulare (ECM1) è una
glicoproteine presente in molti tessuti, compresa la cute.
Scoperta nel 1994, il suo ruolo nella biologia della cute è
stato in gran parte sconosciuto fino al 2002 quando è stato
identificato il gene/proteina responsabile della proteinosi
lipode, malattia a carattere autosomico recessivo. Dal pun-
to di vista clinico è caratterizzata da infiltrazioni e lesioni cica-
triziali a livello della cute e delle mucose, e da un punto di
vista istopatologico, da distruzione e duplicazione della
membrana basale con numerosi depositi di sostanza ialina nel
derma. Sono state identificate più di 30 mutazioni del gene
di ECM1, mutazioni ricorrenti, alleli ancestrali, correlazio-
ni fra genotipi e fenotipi; sono state pertanto messe a punto
nuove strategie diagnostiche. Un nuovo passo avanti nelle
conoscenze del ruolo di ECM1 nella biologia cutanea è sta-
to fatto nel 2003 con la scoperta di autoanticorpi circolanti
contro ECM1 nel siero di pazienti affetti da lichen sclerosus,
una malattia infiammatoria cronica che condivide molti
aspetti anatomopatologici e clinici con la proteinosi lipoide.
Sono stati caratterizzati gli autoanticorpi e isolati gli epito-
pi immunomodulanti; attualmente viene utilizzato un nuovo
test ELISA per la diagnosi del lichen sclerosus. Studi di
Agosto 2005
EXTRACELLULAR MATRIX PROTEIN 1
valutazione dell’interazione proteine/proteine hanno dimo-
strato che ECM1 lega a perlecan, proteglicano eparan solfato,
alle proteine della matrice metalloproteinasi 9, fattori di cre-
scita dell’epidermide e legumain. Questi risultati, insieme ai
dati di correlazione fra la proteinosi lipoide e il lichen scle-
rosus, suggeriscono che ECM1 gioca un ruolo fondamenta-
le in molti aspetti della differenziazione dell’epidermide,
nel mantenimento dell’architettura del derma e nel regolare
la composizione della membrana basale. Pertanto la glico-
proteina ECM1 ha un ruolo fondamentale nella biologia
cutanea.
PAROLE CHIAVE: Geni, mutazioni - Autoanticorpi - Proteino-
si lipoide - lichen sclerosus - Epidermide - Derma - Mem-
brana basale.
References
1. Mathieu E, Meheus L, Raymackers J, Merregaert J. Characterization
of the stromal osteogenic cell line MN7: identification of secreted
MN7 proteins using two-dimensional polyacrylamide gel elec-
trophoresis, western blotting and microsequencing. J Bone Miner
Res 1994;9:903-13.
2. Bhalerao J, Tylzanowski P, Filie JD, Kozak CA, Merregaert J. Mol-
ecular cloning, characterization and genetic mapping of the cDNA cod-
ing for a novel secretory protein of mouse. Demonstration of alternative
splicing in skin and cartilage. J Biol Chem 1995;270:16385-94.
3. Smits P, Ni J, Feng P, Wauters J, van Hul W, Boutaibi ME et al. The
human extracellular matrix gene 1 (ECM1): genomic structure, cDNA
cloning, expression pattern and chromosomal localization. Genomics
1997;45:487-95.
4. Johnson MR, Wilkin DJ, Vos HL, de Ortiz Luna RI, Dehejia AM,
Polymeropoulos MH et al. Characterization of the human extracellular
matrix protein 1 gene on chromosome 1q21. Matrix Biol 1997;16:
289-92.
5. Deckers MM, Smits P, Karperien M, Ni J, Tylzanowski P, Feng P et
al. Recombinant extracellular matrix protein 1 inhibits alkaline phos-
phatase activity and mineralization of mouse embryonic metatarsals
in vitro. Bone 2001;28:14-20.
6. Han Z, Ni J, Smits P, Underhill CB, Xie B, Chen Y et al. Extracellu-
lar matix protein 1 (ECM1) has angiogenic properties and is expressed
by breast tumor cells. FASEB J 2001;15:988-94.
7. Sekiya I, Vuoristo J, Larson B, Prockop DJ. In vitro cartilage forma-
tion by human adult stem cells from bone marrow defines the sequence
of cellular and molecular events during chondrogenesis. Proc Natl
Acad Sci USA 2002;99:4397-402.
8. Le Naour F, Hohenkirk L, Grolleau A, Misek DE, Lescure P, Geiger
JD et al. Profiling changes in gene expression during differentiation
and maturation of monocyte-derived dendritic cells using both oligonu-
cleotide microarrays and proteomics. J Biol Chem 2001;276:
17920-31.
9. Rickman D, Bobek M, Misek D, Kuick R, Blaivas M, Kurnit DM et
al. Distinctive molecular profiles of high-grade and low-grade gli-
nomas based on oligonucleotide microarray analysis. Cancer Res
2001;61:6885-91.
10. Smits P, Poumay Y, Karperien M, Tylzanowski P, Wauters J, Huyle-
broeck D et al. Differentiation-dependent alternative splicing and
expression of the extracellular matrix protein 1 gene in human ker-
atinocytes. J Invest Dermatol 2000;114:718-24.
11. Mongiat M, Fu J, Oldershaw R, Greenhalgh R, Gown AM, Iozzo RV.
Perlecan protein core interacts with extracellular matrix protein 1
(ECM1), a glycoprotein involved in bone formation and angiogene-
sis. J Biol Chem 2003;278:17491-9.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHAN
12. Kragh-Hansen U. Structure and ligand binding properties of human
serum albumin. Dan Med Bull 1990;37:57-84.
13. Godin RE, Urry LA, Ernst SG. Alternative splicing of the Endo16
transcript produces differentially expressed mRNAs during sea urchin
gastrulation. Dev Biol 1996;179:148-59.
14. Terlizzi J, Li K, Aho A, Fujimoto N, Oyama N, Hamada T et al. Char-
acterization of ECM1 protein interactions by yeast-two-hydrid system.
J Invest Dermatol 2004;22:A38.
15. Urbach E, Wiethe C. Lipoidosis cutis et mucosae. Virchows Arch
Path Anat 1929;273:285-319.
16. Hamada T. Lipoid proteinosis. Clin Exp Dermatol 2002;27:624-9.
17. Hamada T, McLean WHI, Ramsay M, Ashton GH, Nanda A, Jenkins
T et al. Lipoid proteinosis maps to 1q21 and is caused by mutations
in the extracellular matrix protein 1 gene (ECM1). Hum Mol Genet
2002;11:833-40.
18. Hamada T, Wessagowit V, South AP, Ashton GH, Chan I, Oyama N
et al. Extracellular matrix protein 1 gene (ECM1) mutations in lipoid
proteinosis and genotype-phenotype correlation. J Invest Dermatol
2003;120:345-50.
19. Chan I, El-Zurghany A, Zendah B, Benghazil M, Oyama N, Hamada
T et al. Molecular basis of lipoid proteinosis in a Libyan family. Clin
Exp Dermatol 2003;28:545-8.
20. van Hougenhouck-Tulleken W, Chan I, Hamada T, Thornton H, Jenk-
ins T, McLean WH et al. Clinical and molecular chracterization of
lipoid proteinosis in Namaqualand, South Africa. Br J Dermatol
2004;151:413-23.
21. Chan I, Sethuraman G, Sharma VK, Bruning E, Hamada T, McGrath
JA. Molecular basis of lipoid proteinosis in two Indian siblings. J
Dermatol 2004;31:764-6.
22. Chan I, Bingewar G, Patil K, Nayak C, Wadhwa SL, McGrath JA.
An Indian child with lipoid proteinosis resulting from a recurrent
frameshift mutation (507delT) in the extracellular matrix protein 1
(ECM1) gene. Br J Dermatol 2004;151:726-7.
23. Heyl T. Genealogical study of lipoid proteinosis in South Africa. Br
J Dermatol 1970;83:338-40.
24. Gordon H, Gordon W, Botha V. Lipoid proteinosis in an inbred
Namaqualand community. Lancet 1969;1:1032-5.
25. Strassberger E. The Rhenish Mission Society in South Africa 1830-
1950. South Africa: C. Struik (Pty) LTD; 1969.
26. McGrath JA, Eady RA. The role of immunohistochemistry in the
diagnosis of the non-lethal forms of junctional epidermolysis bul-
losa. J Dermatol Sci 1997;14:68-75.
27. Ashton GHS, McLean WHI, South AP, Oyama N, Smith, FJD, Al-
Suwaid R et al. Recurrent mutations in Kindlin-1, a novel keratinocyte
focal contact protein, in the autosomal recessive skin fragility and
photosensitivity disorder, Kindler syndrome. J Invest Dermatol
2004;122:78-83.
28. Chan I, Oyama N, Hamada T, Bhogal BS, Black MM, Hamada T et
al. Rapid diagnosis of lipoid proteinosis using an anti-extracellu-
lar matrix protein 1 (ECM1) antibody. J Dermatol Sci 2004;35:
151-3.
29. Powell J, Wojnarowska F. lichen sclerosus. Lancet 1999;353:1777-83.
30. Tasker GL, Wojnarowska F. lichen sclerosus. Clin Exp Dermatol
2003;28:28-33.
31. Neill SM, Tatnall FM, Cox NH. Guidelines for the management of
lichen sclerosus. Br J Dermatol 2002;147:640-9.
32. Oyama N, Chan I, Neill SM, Hamada T, South AP, Wessagowit V et
al. Autoantibodies to extracellular matrix protein 1 in lichen sclero-
sus. Lancet 2003;362:118-23.
33. Chan I, Oyama N, Neill S, Wojnarowska F, Black MM, McGrath JA.
Characterization of autoantibodies to extracellular matrix protein
1 (ECM1) in lichen sclerosus. Clin Exp Dermatol 2004;29:
499-504.
34. Oyama N, Chan I, Neill S, South AP, Wojnarowska F, Kawakami Y et
al. Development of antigen-specific ELISA for circulating autoanti-
bodies to extracellular matrix protein 1 (ECM1) in lichen sclerosus.
J Clin Invest 2004;113:1550-9.
405
CHAN
35. Ishii K, Amagai M, Hall RP, Hashimoto T, Takayanagi A, Gamou S
et al. Characterization of autoantibodies in pemphigus using antigen-
specific enzyme-linked immunosorbent assays with baculovirus-
expressed recombinant desmogleins. J Immunol 1997;159:2010-7.
36. Dunlevy JR, Hassell JR. Heparan sulphate proteoglycans in base-
ment membranes: perlecans, agrin and collagen XVIII. In: Iozzo RV
editor. Proteoglycans; structure, biology and molecular interactions.
New York: Marcel Dekker Inc; 2000.p.275-336.
37. Zeeuwen PL. Epidermal differentiation: the role of proteases and
their inhibitors. Eur J Cell Biol 2004;83:761-73.
406 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
EXTRACELLULAR MATRIX PROTEIN 1
38. Uyemura T, Takagi H, Yanagida T, Sako Y. Single-molecule analysis
of epidermal growth factor signaling that leads to ultrasensitive cal-
cium response. Biophys J 2005;88:3720-30. Epub 2005 Mar 4.
39. Sung HJ, Johnson CE, Lessner SM, Magid R, Drury DN, Galis ZS.
Matrix metalloproteinase 9 facilitates collagen remodeling and angio-
genesis for vascular constructs. Tissue Eng 2005;11:267-76.
40. Kowalewski C, Kozlowska A, Chan I, Gorska M, Wozniak K, Jablon-
ska S et al. Three-dimensional imaging reveals major changes in skin
microvasculature in lipoid proteinosis and lichen sclerosus. J Der-
matol Sci 2005;38:215-24.
Agosto 2005

The histologic diagnosis of early mycosis fungoides:
frequent problems, sporadic solutions
Recent advances in the diagnosis of mycosis fungoides, includ-
ing T cell receptor gene rearrangement studies, have improved
our ability to diagnose this challenging condition. However, it
is clear that clonality alone does not equate with malignancy.
Despite the aide of ancillary studies, mycosis fungoides remains
one of the most difficult dermatologic diagnoses to establish. The
histopathologic diagnosis of mycosis fungoides remains with-
in the realm of clinical/histologic/molecular correlation. As
such, histologic analysis of early mycosis fungoides remains a
key factor in the diagnosis of this challenging condition. It is well
known that mycosis fungoides can mimic and be mimicked by
a variety of other dermatologic conditions, most of which are
inflammatory. A wide variety of studies have well characterized
the histologic findings of mycosis fungoides, however, we know
much less about which histologic criteria are most specific for
this disorder. This article will address myriad of difficulties
encountered in the histopathologic diagnosis of mycosis fun-
goides and then review individual criteria used to establish
this disorder, with emphasis on criterion specificity.
KEY WORDS: Mycosis fungoides, diagnosis - Mycosis fungoides,
histology - Mycosis fungoides, classification.
What is the most difficult diagnosis to establish in
dermatopathology?
P atch Stage mycosis fungoides (MF) is arguably
the most difficult dermatopathologic diagnosis to
establish histopathologically. There is literature to sub-
stantiate this impression. Approximately a decade ago,
Address reprint requests to: E. J. Glusac, MD, Yale University School
of Medicine, Dermatopathology Laboratory, 5031 LMP, P.O. Box 208059,
New Haven, CT 06520-8059. E-mail: Earl.Glusac@yale.edu
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:407-16
E. J. GLUSAC
Department of Pathology and Dermatology
Yale University School of Medicine, New Haven, CT, USA
studies performed by members of European Organi-
zation for Research and Treatment of Cancer (EORTC)
demonstrated accuracy in the diagnosis in MF likened
to a coin toss.1 Three expert pathologists reviewed 73
MF biopsies admixed with controls on 2 different
occasions. MF was identified correctly on both occa-
sions 50% of the time.1 Forty percent of control cas-
es were identified correctly on both occasions. Other
studies seemed to appear more optimistic but, on clos-
er review, were equally disheartening. In a study sim-
ilar to that of the EORTC, Olerud et al.2 were able to
correctly diagnose or suggest MF 92% of the time
(60% of MF biopsies diagnostic of MF; 32% consis-
tent with or suspicious for MF). Review of the control
group in this study revealed significant problems, how-
ever. Fifty two percent of control biopsies were called
suspicious for MF, and 6% were called MF outright.
These results suggest that if we lower our threshold for
the diagnosis of MF, we will significantly over-diag-
nose this condition. Given the implications of over-
treatment, insurance coverage issues, psychological
well being of patients, and a variety of other matters,
it is arguably more important not to over-diagnose MF
than to under-diagnose it.
Some authors who have addressed discordance in the
diagnosis of MF have argued that similar inaccura-
cies are seen in most diseases and in most organ sys-
407
GLUSAC THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS
tems.3 A variety of references are typically cited in
support of this premise. One frequently cited refer-
ence, authored by Rosai, addressed diagnostic dis-
crepancies in the histopathology of the intraductal
breast carcinoma.4 However, a careful review of this
manuscript reveals disagreement between benign and
malignant diagnoses (the type of discrepancy we see
in MF studies) in less than 1/4 of cases. Frequently
cited papers in pulmonary and in gynecologic pathol-
ogy demonstrate even less significant discrepancies
when compared with MF.5, 6 And, for all the difficul-
ties encountered in histopathologic interpretation of
melanocytic lesions, interrelator agreement is still
much better with these lesions than with MF.7
Why so difficult?
There are at least 8 important reasons why MF is so
difficult to diagnose.
1) MF is a clinicopathologic diagnosis.8-10 The his-
tory we often receive is rule out MF. It is often unclear
in such cases whether the clinician believes the patient
likely has MF or, rather, an inflammatory disease, with
MF at the bottom of the differential diagnosis. A vari-
ety of other conclusions are also possible.
2) Treatment alters the histopathologic features of
MF. At the time of biopsy, patients have often failed
standard treatments for a presumed exanthem, includ-
ing topical steroids. It should be noted that topical
steroids appear to diminish or erase the epidermotropic
features of MF.11
3) MF has a wide variety of clinical and histopatho-
logic variants, granulomatous, folliculotropic, verru-
cous, bullous, hyperpigmented and hypopigmented to
name a few.12 While the diagnosis of standard patch
stage MF is difficult enough, the existence of a variety
of challenging variants complicates the matter.
4) Clinically, early MF often looks more like a rash
than a neoplasm.13 Certainly, other neoplasms can
occasionally resemble exanthems; sporadic examples
of Bowen’s disease (squamous cell carcinoma in situ)
show an appearance that resembles eczema. Fortu-
nately, these 2 diseases do not resemble one another
under a microscope. Most neoplasms are mimicked
by other neoplasms, e.g. malignant melanoma and
pigmented basal cell carcinoma. Fortunately, these 2,
at least, do not resemble each other microscopically.
408 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
5) Mycosis fungoides very often looks like a rash
histologically as well. As such, a clinical misdiagnosis
may be supported by a congruous if inaccurate histo-
logic impression. This can readily occur, as the cells of
early MF often do not show significant morphologic dif-
ferences from those of inflammatory conditions.11, 14
6) Mycosis fungoides does not resemble merely a
few different inflammatory processes under the micro-
scope; it resembles many. Shapiro et al., in an analy-
sis of 222 MF biopsies, demonstrated that virtually
every inflammatory pattern developed by Wallace
Clark and A. Bernard Ackerman can be seen in MF.8
The most common patterns encountered are psoriasi-
form, lichenoid and psoriasiform/lichenoid. Less com-
mon patterns include superficial perivascular, super-
ficial perivascular and interstitial, vacuolar, psoriasi-
form/spongiotic, spongiotic/psoriasiform/lichenoid,
nodular, superficial and deep perivascular and inter-
stitial, diffuse and folliculitic. Rare patterns include
spongiotic, vasculitic, vesicular and panniculitic. With
this in mind, I am sometimes asked how one becomes
good at diagnosing MF. My answer is “by becoming
good at everything in the differential diagnosis”
(inflammatory skin disorders mostly). Often, one can
only rule out MF by making another diagnosis.
7) MF resembles inflammatory disorders not to a
small degree but, frequently, to a large degree. Lym-
phocytes typically infiltrate the basal layer in MF,
interacting with Langerhans cells.11, 14, 15 This feature
is seen in a wide variety of other inflammatory con-
ditions and is, in fact, typical of some of them, includ-
ing lichen sclerosis et atrophicus.16, 17
8) There is no agreed upon absolute criteria for the
diagnosis of early MF. The concept of MF underwent
a paradigm shift in the late 1970’s.11, 14, 18 Previously
thought of as a rare, relentless, fatal lymphoma, MF
became accepted as a lymphoma which usually
behaves in an indolent fashion, marked by patches,
with progression to tumor stage disease or fatal disease
in a minority of patients. It is really a tale of 2 dis-
eases. In fact, were a particular British author able to
write about this disorder he might have said (my
changes in italics): “It was the best of diseases, it was
the worst of diseases, it was the diagnosis of wisdom,
it was the diagnosis of foolishness... we has all crite-
ria before us, we had nothing before us, we were all
going direct to Heaven, we were all going direct the
other way”.19
Histopathologists generally demonstrate a high
Agosto 2005
THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS
degree of accuracy in the diagnosis of tumor stage
MF. Clinically and histopathologically, the presence of
a neoplasm is evident at this stage. And whether these
tumors retain their epidermotropic capacity or whether
they lose it, as they often do, a diagnosis of tumor
stage MF can generally be made. And whether the
tumor is comprised of small to medium size convoluted
lymphocytes or large transformed ones, a diagnosis
of the tumor stage of this lymphoma can usually be
established.18 Plaque stage disease, though not as sim-
ple, is generally not so laden with difficulties as patch
stage disease.20 One sees, by definition, involvement
of the reticular dermis in plaque stage disease. This
is usually accompanied by the epidermal and papillary
dermal changes typical of MF. In difficult examples of
plaque stage disease, ancillary studies, including
immunohistochemistry and gene rearrangement stud-
ies, are often helpful.21-23
The diagnosis of patch stage disease, as mentioned,
is entirely a different matter. Regarding the literature
on histopathologic diagnosis of early MF, it is fair to
say that there is good news and bad news. The good
news is that some studies have conveyed that accura-
cy in the diagnosis of MF does increase with experi-
ence.2, 3 The bad news is that most studies suggest that
we have not yet established adequate criteria to diag-
nose or dismiss MF consistently.1-3
Magic bullet?
The next logical question to ask is “If histopathol-
ogy is inadequate to diagnose early MF, what ancillary
studies can identify this challenging condition?” The
first to come to mind is immunophenotyping. It is
readily accessible and commonly used, but how use-
ful is it? Early hopes centered on increased CD4:CD8
ratio, but it has been subsequently demonstrated that
most inflammatory disorders are CD4 predominant,
and, of course, some examples of cutaneous T cell
lymphoma are CD8 positive. Immunophenotyping is
clearly helpful in plaque and tumor stage MF, where
loss of the common T cell antigens CD2, CD3 and/or
CD5 may be seen.21, 22 Such losses are not seen, how-
ever, in patch stage disease. Loss of Leu-8 was origi-
nally thought to be helpful in patch stage disease, but
its value has now been dismissed.21, 22, 24 Loss of CD7
(Leu-9) is sometimes still touted as useful,25 but loss
of CD7 can be seen as frequently in inflammatory
conditions as in early MF.21, 22, 24 Additionally, it is
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GLUSAC
important to bear in mind that immunohistochemical
analysis of T cell infiltrates relies upon identifying
antigen loss. It must be kept in mind that, even with
adequate controls, it is more difficult to be certain
about absence of staining than positive staining. At
Yale University and at some other MF centers,
immunohistochemistry is generally reserved until after
a diagnosis of cutaneous T cell lymphoma has been
established. It is then employed to identify aggressive
subsets of cutaneous T cell lymphoma, such as gam-
ma/delta lymphoma or aggressive variants of CD8
positive lymphoma.
Less controversial is investigation for T cell recep-
tor gene rearrangements (TCR) via polymerase chain
reaction (PCR). It is important to keep a variety of
caveats in mind, however, in the interpretation of TCR
results. It should be noted that a significant percentage
of patients with indubitable patch stage MF will not
demonstrate a clone via PCR.23, 26-28 Furthermore,
investigation of patients without indubitable MF is
fraught with even greater difficulty. Patients described
as having parapsoriasis,29 pre-MF 26 or as borderline 26
can show clonality rates of significantly less than 50%.
It is also important be aware that clones can be iden-
tified in disorders that we do not categorize as malig-
nant. A few of these include cutaneous lymphoid hyper-
plasia,29, 30 pityriasis lichenoides et varioliformis acu-
ta,31-33 pigmented purpuric eruption,34 lichen sclero-
sis,35 and even lichen planus.36
As such, it is fair to say that there is no magic bul-
let. It is also likely fair to say that the gold standard for
the diagnosis of MF remains in the realm of clini-
cal/histologic/molecular correlation. As such,
histopathology remains a key factor.
With this in mind, and given the coin toss like sta-
tus of histopathologic diagnosis of MF, it is fair to ask:
“Do we know the histologic criteria for MF?” I think
that we do. There have been many excellent descrip-
tive studies regarding MF.8, 14, 37, 38 From such studies
we know a great deal about histologic features of MF,
but we know less regarding which criteria are most
specific. To be more accurate in the diagnosis of MF,
we must know the relative specificity of various criteria
employed. We can only establish specificity with con-
trolled, blinded studies of MF versus controls. Impor-
tantly, there must also be a gold standard against which
the histopathologic features of MF can be analyzed
as independent variables. Given that the current gold
standard remains within the realm of clinical/histo-
409
GLUSAC THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS

attempted to address the specificity of various criteria

Figure 1.—Lymphocytes within the epidermis which are larger than those
within the dermis has been shown to be a specific criterion for MF.
logic/molecular correlation, it is difficult to avoid cir-
cular reasoning in histologic studies of MF e.g. biop-
sies are placed into an MF arm of a study, at least in
part, due to the fact that they originally exhibited his-
tologic features of MF.18
With these limitations in mind, at least 3 studies have
for MF. Each involved blinded reviews of MF and con-
trol cases. I will refer to these as the Stanford study,15
the EORTC study 39 and the International Society for
Cutaneous Lymphoma (ISCL) study (Burg G, Cerroni
L, Glusac EJ, Guitart J, Haeffner AC, Sander CA et
al. International Society for Cutaneous Lymphoma--
Early Mycosis Fungoides Study Project; Zurich, May
1999, manuscript sumitted). The Stanford study was the
largest of the 3, involving 64 MF biopsies and 47 con-
trols.15 This study emulated a practice-like scenario.
The MF cases were biopsies sent in to rule out MF that
subsequently proved to be MF, as judged by the clini-
cian involved in the study, via analysis of disease pro-
gression and of ancillary tests (immunophenotyping
and/or gene rearrangement studies). Control cases were
biopsies sent in to rule out MF from patients who proved
to have another disorder as judged by these same means.
The ISCL study, with 33 MF biopsies and 33 control
biopsies had, arguably, the purest MF group (Burg et al.
submitted). It employed early biopsies from patients
who subsequently progressed to tumor stage disease
and/or died of disease. The control group included a
wide variety of inflammatory disorders. I had the good

TABLE I.—Criteria for mycosis fungoides.

Stanford 15 EORTC 39 ISCL (submitted)

MF CTL MF CTL MF CTL

7-9 µ Convoluted
Lymphocytes __ __ 100%** 8% __ __
LargeER
Interepidermal Lymphocytes 20%* 0% __ __ 17% 3%
Convoluted lymphocytes 67%* 32% __ __ 53% 12%
Haloed lymphocytes 59%* 13% __ __ 13% 0%
Pautrier’s
Microabscesses 37%* 2% 4%** 0% 17% 7%
Disproportionate
Exocytosis 58%* 28% __ __ 37% 6%
Basilar***
Lymphocytes 67%* 23% 46%** 0% 17% 3%
Pagetoid
Dystribution — — 33%** 0% 0% 0%
Papillary dermal fibrosis 61%* 49% 33%** 100% 67% 63%
* 2+ or greater (of 4)
** tiny collections of 4 cells in 42%
*** Defined variously, see text
Stanford = Stanford University study
ISCL = International Society for Cutaneous Lymphoma study
EORTC = European Society for Research and Treatment of Cancer study
MF = Mycosis Fungoides
CTL = Control Cases

410 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS GLUSAC

fortune to be involved in the Stanford 15 and ISCL stud-

ies (Burg et al. submitted). The third study, performed
by the EORTC, involved 24 MF biopsies.39 This study
evaluated initial biopsies from patients who subse-
quently presented with indubitable MF via subsequent
biopsy and clinical course. The EORTC study had a
strong control group of 13 distinct MF mimics. In the
remainder of the article, I will focus on the findings of
these 3 studies.

Criteria for mycosis fungoides

LargER intraepidermal lymphocytes

Lymphocytes in the epidermis which are larger Figure 2.—Strictly defined, Pautrier’s microabscesses such as these have
than those within the dermis (largER intraepidermal been shown to be a specific feature of MF.
lymphocytes) were evaluated by the Stanford 15 and
ISCL (Burg et al. submitted) studies (Figure 1). This
finding was seen in 17% to 20% of MF biopsies and
in 0 to 3% of controls (Table I). This criterion is a very
interesting one. It highlights the fact that, early in
the course of MF, neoplastic cells tend to home to
the epidermis. It also suggests that much of what we
see in a MF biopsy is reaction pattern to tumor rather
than tumor per se. This connotes that a variety of
other frequently employed criteria may represent
reaction to pattern to tumor in large part.18 Papillary
dermal fibrosis and band-like infiltrate come to mind.
The presence of LargER intraepidermal lymphocytes
is not a sensitive criterion; however, it is an important
one. It is important because it is a relatively specific
criterion for MF, a disease with few specific crite-
ria. Figure 3.—Pseudo-Pautrier’s microabscesses are composed primarily of
Langerhans cells. They often show a flask-shaped configuration, opening
on to the epidermal surface.
Pautrier’s microabscesses
Pautrier’s microabscesses, another important crite- in the epidermis. Such collections were seen in 42% of
rion, is also relatively specific but insensitive for MF. early MF biopsies, approximating the percentage of
How often one identifies Pautrier’s microabscesses Pautrier’s microabscesses seen in the Stanford study
depends on how one defines the term. The EORTC (37%), which accepted 4 cell clusters.15 Most other
study defined it as “sharply marginated clusters of studies of Pautrier’s microabscesses, including the
atypical lymphoid cells... that were closely opposed to ISCL study (Burg et al. study submitted), have iden-
one another with uniform cytologic features... with tified them in approximately 20% of early MF biop-
no plasma or fibrin deposition or significant cytopathic sies.8, 37 It is important to bear in mind that collections
changes in the surrounding keratinocytes” 39 (Figure 2). of cells within the epidermis resembling Pautrier’s
Defined so rigorously, Pautrier’s microabscesses were microabscesses can be seen in inflammatory condi-
found in only 4% of early MF biopsies by the EORTC tions.40-42 These are usually comprised of Langerhans
group.39 It should be noted that this study also had a cat- cells. Many such pseudo-Pautrier’s microabscesses
egory termed “tiny collections” of up to 4 cells with- have a flask shaped appearance, opening onto the epi-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 411

GLUSAC THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS
Figure 4.—Disproportionate exocytosis (lymphocytes in the epidermis
associated with relative paucity of spongiosis) is a subjective finding
indicative of MF.
Figure 5.—Four or more contiguous lymphocytes within the basal layer
has been shown to be a feature indicative of MF.
dermal surface 40 (Figure 3). Such collections will
label with CD1a and S100.41 They are typical of spon-
giotic processes but can also be seen in MF.41
Disproportionate exocytosis
Disproportionate exocytosis describes intraepider-
mal lymphocytes, associated with a relative paucity
of spongiosis 14 (Figure 4). It is a criterion that is uni-
formly relied upon in scanning any seemingly inflam-
matory skin biopsy. Though important, this criterion
is difficult if not impossible to quantify. Both the Stan-
ford study 15 and ISCL studies(Burg et al. study sub-
412 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
Figure 6.—Lymphocytes at all levels of the epidermis (pagetoid distribu-
tion) is indicative of MF but is rarely seen in early biopsies of this condi-
tion.
mitted) subjectively evaluated this criterion, and both
found it to be useful.
Basilar lymphocytes
Basilar lymphocytes have been likened to strings
of pearls or toy soldiers. Neoplastic lymphocytes in MF
typically infiltrate the basal layer, in contiguity with
Langerhans cells.11 When basilar lymphocytes are
florid, a diagnosis of MF is likely.14 But how much is
enough? In the Stanford study, 1-5 lymphocytes in the
basal layer per 20X field was a statistically signifi-
cant and relatively sensitive if poorly specific dis-
criminator.15 In the ISCL study 4 contiguous lym-
phocytes within the basal layer (Figure 5) was an insen-
sitive criterion (17% of MF cases), but it was almost
perfectly specific (Burg et al. study submitted). The
EORTC study found that several contiguous rete (not
further specified) involved by basilar lymphocytes
was seen in approximately half of MF cases and, sur-
prisingly, in no control specimens.39
Pagetoid distribution
A pagetoid distribution of lymphocytes (Figure 6)
(lymphocytes seen at all levels of the epidermis) was
not seen in any MF case or any controls in the ISCL
study (Burg et al. submitted). This is not surprising, as
a pagetoid distribution is generally considered a find-
ing of more advanced MF. Again, however, there is
unexpected data from the EORTC study, where page-
Agosto 2005
THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS
Figure 7.—Medium-large convoluted lymphocytes that approximate the
width of basilar keratinocytes are features strongly supportive of MF.
toid distribution was seen in 1/3 of MF biopsies and in
no control specimens.39
Medium-large convoluted lymphocytes
The EORTC study evaluated for the presence of
medium to large (7-9 µ) lymphocytes with convo-
luted outlines (Figure 7). Such cells approximate the
width of basilar keratinocyte nuclei. In the EORTC
study, such cells were present within the epidermis of
all 24 MF biopsies and within the dermis of 22 of
24 MF biopsies.39 This feature was found in only one
of 13 control specimens, and, then, only in the epi-
dermis of that case. This data is very striking, and it
is curious. It would appear to imply that a diagnosis
of MF can be made routinely and reliably. Other stud-
ies performed by members of this same group1 and
other groups 2 do not appear to support that suppo-
sition. Nonetheless, medium-large convoluted lym-
phocytes is an important criterion. Its combination of
size and convolution has crystalized existing ele-
ments in the literature and provided a useful histologic
benchmark (width of basilar keratinocyte nuclei) for
comparison.
The Stanford 15 and ISCL studies (Burg et al. sub-
mitted) did not evaluate for the above criterion, but,
rather, for convoluted nuclei alone. Each found con-
voluted nuclei to be a significant, if imperfect dis-
criminator. It should be noted that the evaluation
of nuclear convolutions is a highly subjective
endeavor 43 and requires excellent histopathologic
sections.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GLUSAC
Figure 8.—Lymphocytes with halos or vacuoles around them are a partially
artifactual feature that is supportive of MF.
Haloed lymphocytes
Haloed lymphocytes are defined as lymphocytes
within the epidermis with a vacuole around them evi-
dent at relative low magnification (Figure 8). The cause
of halos is not precisely known. They are not typical-
ly seen in frozen sections of MF and do not contain
mucin.37, 44 They are thought to be, in large part, an arti-
factual phenomenon. They are possibly the result of
contraction of the more abundant cytoplasm of the
neoplastic lymphocytes and/or poor cohesion of neo-
plastic lymphocytes to surrounding keratinocytes.14,
38 Haloing does vary amongst laboratories. In the Stan-
ford study, performed on biopsies processed in the
Stanford University histopathology laboratory, it was
the strongest histologic discriminator between MF
and control cases.15 In the ISCL study (Burg et al.
study submitted), performed on biopsies from a vari-
ety of European laboratories, it was an insensitive
(13%) but completely specific discriminator.
Papillary dermal fibrosis
While papillary dermal fibrosis (Figure 9) has been
touted as a key feature of patch stage MF, none of the
3 studies in question found it to be a useful discrimi-
nator between MF and control cases. It may be a mark-
er simply of disease chronicity. In fact, the EORTC
study found papillary dermal fibrosis much more fre-
quently in control cases than early MF biopsies.39 It
should be noted that the criterion assessed in each of
these studies was simply papillary dermal fibrosis, as
413
GLUSAC THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS
Figure 9.—Thickened, wiry collagen bundles within the papillary dermis
are frequently seen in late patch state MF; however, this feature has not been
shown to be a significant discriminator between MF and controls.
opposed to lymphocytes splayed between thickened
papillary dermal collagen bundles. The latter criterion
may still be a useful discriminator, especially in late
patch stage MF.
Criteria clusters?
It is well known that no single histologic criterion can
establish a diagnosis of MF. It is poorly known at this
time whether specific criteria clusters can be useful
in the evaluation of MF. The Stanford study did find
that moderate disproportionate exocytosis in combi-
nation with at least one haloed lymphocyte for 20X
field was specific for MF.15 While I doubt that this
feature is specific for MF, it is likely a criteria cluster
that merits attention.
Sézary syndrome
Regarding the leukemic variant of cutaneous T cell
lymphoma, Sézary syndrome it is fair to say that his-
tologic diagnosis goes from difficult to more difficult.
Buechner and Winkelmann’s classic treatise on this
condition showed that only 15% of cases showed an
epidermotropic pattern.45 Shapiro et al. demonstrated
increased spongiosis and diminished epidermotropism
as compared to standard MF.8 Other studies have
demonstrated diminished disproportionate exocyto-
sis,46 diminished basilar lymphocytes,46 fewer convo-
luted lymphocytes,46 increased acanthosis 47 and dimin-
ished Pautrier’s microabscesses 47 as compared to MF.
414 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
Trotter et al. performed, arguably, the most thorough
study of this condition.48 Their study involved 41
patients with Sézary syndrome, each with a clonal
proliferation in the blood as identified by southern
blot analysis. Only 38% of biopsies in this study
showed an epidermotropic pattern with atypical lym-
phocytes. One third of patients showed chronic spon-
giotic changes and no histologic evidence of lym-
phoma on original or repeat biopsies. Of note, this
group showed no better survival than other patients
within the study. The authors suggest that, at least in
some patients with Sézary syndrome, the exanthem
may represent a non-specific chronic spongiotic
response to a primarily leukemic process.48
In this review, I have presented a variety of statistics.
In doing so, I do not mean to imply that the diagnosis
of MF can be made via an equation-like scheme. These
have been attempted in the past but are problematic.49-
51 Nonetheless, it is useful to be aware of the sensitiv-
ity and specificity of various criterion in MF, in order
to integrate this knowledge into the necessarily gestalt
fashion in which we must all establish a diagnosis.
The gold standard for the diagnosis of MF arguably
remains within the realm of clinical/histologic/mole-
cular correlation. Familiarity with sensitivity and speci-
ficity of criteria for MF, in conjunction with extensive
knowledge of disorders within the differential diagnosis
of MF (inflammatory dermatopathology), should help
us improve our accuracy in the histopathologic diag-
nosis of early biopsies of this condition.
Riassunto
Diagnosi istopatologica della micosi fungoide: frequenti i
problemi, rare le soluzioni
Recenti studi sulla diagnosi della micosi fungoide, compresi
quelli riguardanti il riarrangiamento del gene del recettore
delle cellule T, hanno migliorato le possibilità diagnostiche di
questa patologia che ancora oggigiorno lancia numerose sfi-
de al dermatologo. Tuttavia è chiaro il concetto che clonalità
non significa necessariamente neoplasia maligna. Attual-
mente la micosi fungoide costituisce una delle patologie più
difficili da diagnosticare. L’analisi istopatologica permette
di eseguire studi e correlazioni da un punto di vista clinico,
istologico e molecolare e come tale rimane fondamentale
nella diagnosi di questa patologia. La micosi fungoide si può
manifestare con caratteristiche simili ad altre patologie der-
matologiche generalmente di tipo infiammatorio. Molti stu-
di hanno riportato le caratteristiche istologiche della micosi
fungoide, tuttavia non sappiamo ancora quali siano i criteri
istologici più specifici per poter fare diagnosi.
Agosto 2005
THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS
Questa review prenderà in considerazione tutte le diffi-
coltà incontrate nella diagnosi istopatologica della malat-
tia e i criteri individuali utilizzati, siano essi più specifici o
meno.
PAROLE CHIAVE: Micosi fungoide, diagnosi - Micosi fungoi-
de, anatomia patologica, Micosi fungoide, classificazione.
References
1. Santucci M, Burg G, Feller AC. Interrater and intrarater reliability of
histologic criteria in early cutaneous T-cell lymphoma. Dermatol Clin
1994;12:323-7.
2. Olerud JE, Kulin PA, Chew DE, Carlsen RA, Hammar SP, Weir TW
et al. Cutaneous T-cell lymphoma. Evaluation of pretreatment skin
biopsy specimens by a panel of pathologists. Arch Dermatol
1992;128:501-7.
3. Santucci M, Biggeri A, Feller AC, Burg G. Accuracy, concordance, and
reproducibility of histologic diagnosis in cutaneous T-cell lymphomaan
EORTC Cutaneous Lymphoma Project Group Study. European Orga-
nization for Research and Treatment of Cancer. Arch Dermatol
2000;136:497-502.
4. Rosai J. Borderline epithelial lesions of the breast. Am J Surg Pathol
1991;15:209-21.
5. Feinstein AR, Gelfman NA, Yesner R. Observer variability in the
histopathologic diagnosis of lung cancer. Am Rev Respir Dis
1970;101:671-84.
6. Ismail SM, Colclough AB, Dinnen JS, Eakins D, Evans DM, Gradwell
E et al. Observer variation in histopathological diagnosis and grading
of cervical intraepithelial neoplasia. BMJ 1989;298:1030-1.
7. Weinstock MA, Barnhill RL, Rhodes AR, Brodsky GL. Reliability of
the histopathologic diagnosis of melanocytic dysplasia. The dys-
plastic nevus panel. Arch Dermatol 1997;133:953-8.
8. Shapiro PE, Pinto FJ. The histologic spectrum of mycosis fun-
goides/Sezary syndrome (cutaneous T-cell lymphoma). A review of
222 biopsies, including newly described patterns and the earliest
pathologic changes. Am J Surg Pathol 1994;18:645-67.
9. Glusac EJ, Shapiro PE, McNiff JM. Cutaneous T-cell lymphoma.
Refinement in the application of controversial histologic criteria. Der-
matol Clin 1999;17:601-14.
10. Glusac EJ. Criterion by criterion, mycosis fungoides. Am J Der-
matopathol 2003;25:264-9.
11. Ming M, LeBoit PE. Can dermatopathologists reliably make the diag-
nosis of mycosis fungoides? If not, who can? Arch Dermatol
2000;136:543-6.
12. LeBoit PE. Variants of mycosis fungoides and related cutaneous T-cell
lymphomas. Semin Diagn Pathol 1991;8:773-81.
13. Zackheim HS, McCalmont TH. Mycosis fungoides: the great imita-
tor. J Am Acad Dermatol 2002;47:914-8.
14. Sanchez JL, Ackerman AB. The patch stage of mycosis fungoides. Am
J Dermatopathol 1979;1:5-26.
15. Smoller BR, Bishop K, Glusac EJ, Kim YH, Hendrickson M. Reassess-
ment of histologic parameters in the diagnosis of mycosis fungoides.
Am J Surg Pathol 1995;19:1423-30.
16. Fung MA, LeBoit PE. Light microscopic criteria for the diagnosis
of early vulvar lichen sclerosus. A comparison with lichen planus.
Am J Surg Pathol 1998;22:473-8.
17. Citarella L, Massone C, Kerl H, Cerroni L. Lichen sclerosus with
histopathologic features simulating early mycosis fungoides. Am J Der-
matopathol 2003;25:463-5.
18. Glusac EJ. Of cells and architecture: new approaches to old criteria in
mycosis fungoides. J Cutan Pathol 2001;28:169-3.
19. Dickens CA. A tale of two cities. Reader’s Digest ed. Pleasantville, NY:
The Readers Digest Association, Inc; 1859.
20. Lefeber WP, Robinson JK, Clendenning WE, Dunn JL, Colton T.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GLUSAC
Attempts to enhance light microscopic diagnosis of cutaneous T-cell
lymphoma (Mycosis Fungoides). Arch Dermatol 1981;117:408-11.
21. Ralfkiaer E. Immunohistological markers for the diagnosis of cuta-
neous lymphomas. Semin Diagn Pathol 1991;8:62-72.
22. Ralfkiaer E. Controversies and discussion on early diagnosis of cuta-
neous T-cell lymphoma. Phenotyping. Dermatol Clin 1994;12:329-34.
23. Bachelez H, Bioul L, Flageul B, Baccard M, Moulonguet-Michar I,
Verola O et al. Detection of clonal T-cell receptor gamma gene
rearrangements with the use of the polymerase chain reaction in cuta-
neous lesions of mycosis fungoides and Sezary syndrome. Arch Der-
matol 1995;131:1027-31.
24. Payne CM, Spier CM, Grogan TM, Richter LC, Bjore CG, Cromey
DW et al. Nuclear contour irregularity correlates with Leu-9-, Leu-8-
cells in benign lymphoid infiltrates of skin. An ultrastructural mor-
phometric and quantitative immunophenotypic analysis suggesting the
normal T-cell counterpart to the malignant mycosis fungoides/Ã(c)zary
cell. Am J Dermatopathol 1988;10:377-89.
25. Ormsby A, Bergfeld WF, Tubbs RR, Hsi ED. Evaluation of a new
paraffin-reactive CD7 T-cell deletion marker and a polymerase chain
reaction-based T-cell receptor gene rearrangement assay: implica-
tions for diagnosis of mycosis fungoides in community clinical prac-
tice. J Am Acad Dermatol 2001;45:405-13.
26. Aston-Key M, Diss TC, Du MQ. The value of the polymerase chain
reaction in the diagnosis of cutaneous T-cell infiltrates. Am J Surg
Pathol 1997;21:743-7.
27. Tok J, Szabolcs J, Silvers DN, Zhong J, Matsushima AY. Detection of
clonal T-cell receptor gamma chain gene rearrangements by poly-
merase chain reaction and denaturing gradient gel electrophoresis
(PCR/DGGE) in archival specimens from patients with early cutaneous
T-cell lymphoma: correlation of histologic findings with PCR/DGGE.
J Am Acad Dermatol 1998;38:453-60.
28. Bergman R, Faclieru D, Sahar D, Sander CA, Kerner H, Ben-Aryeh
Y et al. Immunophenotyping and T-cell receptor gamma gene
rearrangement analysis as an adjunct to the histopathologic diagno-
sis of mycosis fungoides. J Am Acad Dermatol 1998;39:554-9.
29. Staib G, Sterry W. Use of polymerase chain reaction in the detection
of clones in lymphoproliferative diseases of the skin. Cancer Res
1995;139:239-47.
30. Wood GS, Tung RM, Haeffner AC, Crooks CF, Liao S, Orozco R et
al. Detection of clonal T-cell receptor gamma gene rearrangements in
early mycosis fungoides/Sezary syndrome by polymerase chain reac-
tion and denaturing gradient gel electrophoresis(PCR/DGGE). J Invest
Dermatol 1994;103:34-41.
31. Weiss LM, Wood GS, Ellisen LW, Reynolds TC, Sklar J. Clonal T-cell
populations in pityriasis lichenoides et varioliformis acuta (Mucha-
Habermann disease). Am J Pathol 1987;126:417-21.
32. Dereure O, Levi E, Kadin ME. T-Cell clonality in pityriasis lichenoides
et varioliformis acuta. Arch Dermatol 2000;136:1483-6.
33. Weinberg JM, Kristal L, Chooback L, Honig PJ, Kramer EM, Lessin
SR. The clonal nature of pityriasis lichenoides. Arch Dermatol
2002;138:1063-67.
34. Toro JR, Sander CA, LeBoit PE. Persistent pigmented purpuric der-
matitis and mycosis fungoides: simulant, precursor, or both? A study
by light microscopy and molecular methods. Am J Dermatopathol
1997;19:108-18.
35. Lukowsky A, Munche JM, Sterry W, Audring H. Detection of expand-
ed T cell clones in skin biopsy samples of patients with lichen scle-
rosus et atrophicus by T cell receptor-y. J Invest Dermatol
2000;115:254-9.
36. Schiller PI, Flaig MJ, Puchta U, Kind P, Sander CA. Detection of
clonal T cells in lichen planus. Arch Dermatol 2000;292:568-9.
37. Nickoloff BJ. Light-microscopic assessment of 100 patients with
patch/plaque-stage mycosis fungoides. Am J Dermatopathol
1988;10:469-77.
38. Smith NP. Histologic criteria for early diagnosis of cutaneous T-cell
lymphoma. Dermatol Clin 1994;12:315-22.
39. Santucci M, Biggeri A, Feller AC, Massi D, Burg G. Efficacy of his-
415
GLUSAC THE HISTOLOGIC DIAGNOSIS OF EARLY MYCOSIS FUNGOIDES: FREQUENT PROBLEMS, SPORADIC SOLUTIONS
tologic criteria for diagnosing early mycosis fungoides. An EORTC
Cutaneous Lymphoma Study Group Investigation. Am J Surg Pathol
2000;24:40-50.
40. LeBoit PE, Epstein BA. A vase-like shape characterizes the epidermal-
mononuclear cell collections seen in spongiotic dermatitis. Am J Der-
matopathol 1990;12:612-6.
41. Candiago E, Marocolo D, Manganoni MA, Leali C, Facchetti F. Non-
lymphoid intraepidermal mononuclear cell collections (Pseudo-Pautri-
er Abscesses) A morphologic and immunophenotypical characteri-
zation. Am J Dermatol 2000;22:1-6.
42. Burkert KL, Huhn K, Menezes DW, Murphy GF. Langerhans cell
microgranulomas (pseudo-pautrier abscesses): morphologic diversi-
ty, diagnostic implications and pathogenetic mechanisms. J Cutan
Pathol 2002;29:511-6.
43. Yeh YA, Hudson AR, Prieto VG, Shea CR, Smoller BR. Reassessment
of lymphocytic atypia in the diagnosis of mycosis fungoides. Mod
Pathol 2001;14:285-8.
44. El Darouti M, Marzouk SA, Horn TD. Failure of detection of mucin
in the clear halos around the epidermotropic lymphocytes in mycosis
fungoides. J Cutan Pathol 2000;27:183-5.
45. Buechner SA, Winkelmann RK. Sezary syndrome: a clinicopatho-
logic study of 39 cases. Arch Dermatol 1983;119: 979-86.
416 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
46. Kohler S, Kim YH, Smoller BR. Histologic criteria for the diagnosis
of erythrodermic mycosis fungoides and Sezary syndrome: a critical
reappraisal. J Cutan Pathol 1997;24:292-7.
47. Kamarashev J, Burg G, Kempf M, Hess Smid M, Dummer R. Compar-
ative analysis of histological and immunohistological features in myco-
sis fungoides and Sezary syndrome. J Cutan Pathol 1998;25:407-12.
48. Trotter MJ, Whittaker SJ, Orchard GE, Smith NP. Cutaneous
histopathology of Sezary syndrome: a study of 41 cases with a proven
circulating T-cell clone. J Cutan Pathol 1997;24:286-91.
49. Cooper KD. A scoring system based on differentially weighted criteria
for establishing a standardized threshold for the diagnosis of early
mycosis fungoides. In: Lambert WC, Giannotti B, van Vloten WA
editors. Basic mechanisms of physiologic and aberrant lymphopro-
liferation in the skin. NATO ASI series A. New York: Plenum Press;
1994. p. 291.
50. Hoppe RT, Wood GS, Abel EA. Mycosis fungoides and Sézary syn-
drome: pathology, staging, and treatment. Curr Probl Cancer
1990;14:293-71.
51. Guitart J, Kennedy J, Ronan S, Chimiel JS, Hsiegh YC, Variakojs D.
Histologic criteria for the diagnosis of early mycosis fungoides: pro-
posal for a grading system to standardize the pathology reporting. J
Cutan Pathol 2001;28:174-83.
Agosto 2005

Late Lyme disease, Chronic Lyme disease
and post Lyme disease syndrome
Lyme borreliosis (LB) is a tick-borne spirochetosis caused by
Borrelia burgdorfei (Bb) and transmitted by the bite of infect-
ed hard-bodies ticks of the genus Ixodes. LB is a multisystemic
disease involving skin, joints, nervous system, but also heart and
eyes could be involved. The diagnosis of LB is primarily based
on clinical and epidemiological criteria, but also serological
tests, histological evaluation and cultivation can provide useful
supporting evidence. Late involvement of skin, nervous sys-
tem, joints can also arise after a long period from the thick
bite, even after its apparent eradication. In this review we
describe different clinical syndromes related to Lyme disease.
The hypothesis on the pathogenesis, the similarities with oth-
er diseases, the clinical and diagnostic fundamentals for the
differential diagnosis and the proposed treatment are pre-
sented.
KEY WORDS: Lyme Borreliosis - Late Lyme disease - Chronic
Lyme disease, therapy - Resistant -Lyme arthritis - Post -Lyme
disease.
L yme borreliosis (LB) is a tick-borne spirochetosis
caused by Borrelia burgdorfei (Bb) and trans-
mitted by the bite of infected hard ticks of the genus
Ixodes. LB is a multisystemic disease involving skin,
joints, nervous system, but also heart and eyes could
be involved.1, 2 LB is clinically subdivided into 3 stages,
otherwise the disease could be classified in early and
late LB.3 The diagnosis of LB is primarily based on
clinical and epidemiological criteria, but also sero-
Address reprint requests to: G. Trevisan, Unità Operativa di Dermato-
logia, Dipartimento di Scienze Cliniche, Morfologiche e Tecnologiche, Uni-
versità degli Studi di Trieste, Ospedale di Cattinara, Strada di Fiume 447,
34149 Trieste, Italy. E-mail: trevisan@univ.trieste.it
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:417-33
Clinical and laboratory analysis
G. TREVISAN 1, S. ORTENZIO 1, S. BONIN 1, 2
1Operative Unit of Dermatology
Department of Clinical, Morphological
and Technological Sciences
University of Trieste, Cattinara Hospital, Trieste, Italy
2International Centre for Genetic Engineering
and Biotechnology (ICGEB), Trieste, Italy
logical tests, histological evaluation and cultivation
can provide useful supporting evidence. In routine
patients management the diagnosis of LB is usually
without problem for the early manifestation of the dis-
ease.3
Late involvement of skin, nervous system, joints
can also arise after a long period from the thick bite,4
even after its apparent eradication.5 In some of these
cases, it is possible to find higher IgM values.6 The
pathogenetic mechanism of antibodies persistence is
not clear. Treatment with antibiotics is beneficial for
all clinical manifestations of LB. However, in the late
stages of borreliosis, symptoms may persist despite
extensive and repeated antibiotic treatment,7 even with-
out objective signs of infection or biological mark-
ers.8 This event could be explained by an intracellular
persistence of the Bb in tissues. According to this the-
ory, it escapes to the host immunity system. The intra-
cellular location of Bb could also explain the persis-
tence of the Borrelia in the skin and joints.9-11 Some
patients, after the antibiotic treatment, present inflam-
mation of one or more joints.3, 12 Borrelia seems to be
able to trigger some postinfectious syndromes, whose
417
TREVISAN LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
symptoms can persist also in absence of live spiro-
chetes and thus do not respond to antibiotics.
In the last years the attention is focused on several
clinical features related to LB. These clinical mani-
festations could be similar to LB but also non-specif-
ic complaints and could be associated to the spiro-
chete characteristics or to the immune response to it.
These syndromes are:
— late Lyme disease;
— chronic Lyme disease;
— treatment-resistant Lyme disease;
— post-Lyme disease;
— chronic fatigue syndrome.
Late Lyme disease
Late Lyme disease occurs in 7 months or more after
the bite of the infected thick. Cutaneous manifesta-
tions are mainly atrophosclerodermia, while extracu-
taneous manifestations involve mainly joints and ner-
vous systems.13
Skin manifestations
Acrodermatitis chronica atrophicans (Pick-Herxheimer
disease)
Acrodermatitis chronica atrophicans (ACA) is a rel-
atively frequent chronic skin manifestation of LB.
ACA develops insidiously. It occurs from few months
up to 1 year after the thick bite. Initially it is charac-
terised by red-bluish discolourations, usually on exten-
sor surface of extremities. The lesion could be uni or
bilateral.14 The disease evolution is chronic. The lesions
enlarge very slowly over months to years, after which
the edema slowly vanishes and atrophy become grad-
ually prominent. The skin becomes thin and wrinkled
and the discoluration becomes violet. Sometimes the
atrophy lesions develop and dermis, subcutis and mus-
cles are affected. It has been also reported the presence
of chronic ulcers and malignant transformation of the
atrophic skin. Lesions can occur with itching or burn-
ing sensation, but also without any symptoms.14, 15
Other cutaneous manifestations are: lichen sclero-
sus et atrophicus,16, 17 general sclerodermia, atropho-
dermia of Pierini-Pasini, of nodular panniculitis of
Pfeifer-Weber- Christian.15, 18
Atrophosclerodermic dermatitis such as lichen scle-
rosus et atrophicus, morphea, circumscribed sclero-
418 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
dermia, linear scleroderma, idiopathic atrophoderma
of Pierini-Pasini, Parry-Romberg syndrome, Busckhe
scleredema and eosinophil fasciitis of Schulmann has
been reported as late Lyme disease manifestations,
although the associations have not yet been estab-
lished satisfactorily. In some cases of ACA borrelial iso-
lation was recovered from skin biopsy specimens of
ACA lesion of more than 10 years duration.
In particular, morphea has been related several times
to late stages of LB. Morphea, also known as localised
scleroderma, is characterised by the induration of cutis
and subcutis due to collagen deposition.16 Morphea
is classified according to clinical characteristics and the
depth of cutaneous tissue involved. Some of these are
plaque, generalised, linear and morphea profunda.
Contrary to systemic sclerosis, in the localised scle-
roderma there is no involvement of internal organs,
digital sclerosis or the presence of Raynaud's phe-
nomenon. Bb DNA was detected by PCR amplification
in skin biopsies of some morphea patients.9
Musculoskeletal system:chronic arthritis
(duration > 1 year)
Lyme arthritis is often regarded as a onset manifes-
tation of LB in North America. In Europe it is less fre-
quent, but the clinical features of Lyme arthritis are
similar both in Europe and North America. The onset
could appear from few weeks to years after the thick
bite. The course of Lyme arthritis is very variable, it is
usually recurrent and can last for several years. The
arthritis could become chronic or maintain intermittent
attacks lasting from a few weeks to months.3 In the
beginning the attacks of arthritis are frequent and short,
then they may be longer. Patients present usually gen-
eral tiredness.15
Features:15
— mono or oligoarthritis;
— asymmetric;
— frequent and intermittent attacks.
Localization:15
— large joints are predominantly affected, most
often the knee.
Clinical features:15
— swelling;
— cutaneous nodules;
— loss of functionality;
— no rigidity in the morning.
Agosto 2005
LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
Examination:15
Rx of the affected joints could present:
— articular effusion;
— osteoporosis;
— erosion of caput osseum;
— calcification of peri-articular soft tissues;
— subarticular bone cysts.
In absence of antibiotic treatment a irreversible ero-
sion of cartilage and bone could occur.
Duration of the illness: in the absence of antibiotic
treatment, the arthritis can last for almost 4 years.
Duration and severity of the arthritis seems to be relat-
ed to genetic features. Patients with HLA-DR4 and
/or DR2 haplotype develop frequently chronic arthri-
tis characterised by erosion and treatment failure.15,
19, 20 Most patients present negative tests for the detec-
tion of Borrelial DNA in synovial fluid, even if the
arthritis persists.4 Lyme arthritis is among the most
common chronic arthritis in children (about 33%).
Nervous system
Encephalopathy with cognitive dysfunction: typi-
cally it is subacute or chronic. Representative mani-
festations are subtle memory and cognitive dysfunction.
Physical examination is usually normal. Patients could
be fretful and somnolent.15
Chronic encephalomyelitis: patients develop fea-
tures that can resemble those seen in focal sclerosis.
Unifocal or multifocal inflammatory disease at the
central nervous system (brain, optic nerve, encephal-
ic trunk, cerebellum) is normally slowly progressive
and involves white matter more than grey-matter (rare).
Typical manifestations are temporary or permanent
focal deficit like hemiparesis, paraparesis, ataxia and
aphasia.15 MRI may suggest a white-matter disease.
Multi-infarctual encephalopathy: acute focal neu-
rological deficit 16 could be present, some of these
could be temporary (like a transient ischemic attack)
or permanent (like a stroke).
Axonal poli-neuropathy:21 most of the patients with
late LB develop mild sensitive neuropathy (mostly
associated with arthritis). Typical manifestations are
peripheral intermittent paresthesia at the extremities.
Electromyography shows a picture of normal nerve
conduction.
At the investigation cerebrospinal fluid (CSF) is
typically normal in patients with late LB, sometimes
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TREVISAN
TABLE I.—Differential diagnosis.
ACA:
Acrocyanosis
Phlebitis
Venous insufficiency
LES
Chronic neuroborreliosis:
Meningitis (viral or bacterial origin)
Alzheimer’s disease
Amyotrophic lateral sclerosis
Parkinson's disease
Face ague
Lyme arthritis:
Rheumatoid arthritis
Juvenile rheumatoid arthritis
Reactive arthritis (Reiter’s syndrome)
LES
CSF may show lymphocytic pleocytosis 10, 22 with
intrathecal antibody production. Borrelia culture from
CSF is positive in about 5% of the cases, while PCR
is positive in about 50% of the cases.
Heart
Myocardiopathy 23 with heart failure is very rare
and could be the only reason for fatal outcome in
patients with LB.
Eyes
Ocular problems in LB seem to be very rare. Eyes
can be affected primarily as a result of the inflamma-
tion of the ocular tissues such as conjunctivitis, keratitis,
iridocyclitis, retinal vasculitis, chorioiditis and optic
neuropathy.3, 24 Prolonged intraocular inflammation
can result in blindness.
Chronic Lyme disease and differential diagnosis
A summary scheme of differential diagnosis is
shown in Table I.
Signs that enable a reliable clinical diagnosis are
related to early LB, they include erythema migrans, the
typical LB skin lesion, Bell's palsy or arthritis (in par-
ticular mono and oligo-articular). On the contrary, the
diagnosis of late LB with its chronic symptoms results
sometimes ambiguous.25, 26
Some symptoms, such as persistent fatigue, arthral-
419
TREVISAN LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
gias, fibromyalgias and other alteration of the nervous
system (neurocognitive and neuropsychiatric), are
related to chronic Lyme disease.27, 28 However, case
reports of patients have reported several non-specific
complaints such as paresthesia, tremor, palpitation,
tachycardia, equilibrium alteration, sweating, visual and
gastrointestinal (irritable colon) disorders, frequent
urination.25 Some investigators believe that Bb per-
sistence could be the causative agent. The pathogen-
ic mechanism of the chronic Lyme disease is not yet
understood.
There are controversial opinions about chronic and
late Lyme disease. Some authors, in fact, consider
them the same clinical event, while others describe
them as 2 distinct clinical aspect of the disease.25
To support the second theory there is the presence in
patients with late LB arthritis of characteristic fea-
tures (swelling, cutaneous nodules,...), while in chron-
ic LB these signs are absent. Moreover, patients with
chronic LB are mostly not responsive to antibiotic
treatment. Patients with late LB and high IgG present
a better response to treatment in comparison with
patients with lower IgG and symptoms of chronic LB.
Typically, chronic LB patients are characterised by
high IgM value and low IgG.6 Several reports support
the increasing evidence that IgM reactivity is com-
mon in chronic, active disease. It is known that IgM
reactivity may represent reactivation of latent disease
or persistent infections in other chronic infections (e.g.
cytomegalovirus and toxoplasma) and it is likely the
case of LB.6 On the other side, the detection of IgM and
IgG antibodies to individual Bb antigens (24kDa, 31
kDa, 34 kDa, but mainly 41 kDa) can provide a sup-
porting evidence of an active phase of LB, but IgM
response could also be induced by several condition of
cross reactivity, including ehrlichial, cytomegalovirus
and toxoplasma infections leading to false positive
results. Some patients with chronic Lyme disease who
responded to treatment, decreased the level of IgM
and increased the level of IgG. The mechanism that
leads to IgM persistence and lower IgG production
remains unclear.6
The entity of chronic Lyme disease has been the
subject of great controversy. Some authors have not
approved the chronic form of Lyme disease assuming
that the ongoing long-lasting symptoms could be relat-
ed to psychiatric problems. The fact that a chronic LB
exists is supported by published reports of epidemio-
logic studies that are related to the incidence of chron-
420 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
ic neuroborreliosis in 30-50% of patients developing
fibromyalgia and chronic fatigue.29 Chronic LB may
be present not only in different organs but also in dif-
ferent patterns. The pathophysiology of the chronic
symptoms is not well understood, hypothesis are rang-
ing from persisting infection to autoimmunity or a
combination of the 2.
Chronic Lyme disease is not fatal, but debilitating,
characterised by persistent symptoms with cyclic
recrudement of the disease. The variety of symptoms
could be related to genetic factors that can contribute
to the development of chronic disease. The incidence
of asymptomatic patients has not been reported but
there are evidences that some individuals, asympto-
matic for months or years after the infection, can man-
ifest chronic symptoms of the disease owing to pro-
voking events such as trauma, pregnancy or psycho-
logical stress.25 There are many theories about the
mechanism leading to chronic LB, it is known that
viable Bb can persist for decades and cause late skin
manifestation of ACA. Thus, the immunopathogenet-
ic findings in ACA can serve as a model for studying
the chronic course of LB. Recent findings indicate
that the most important cells for antigen presentation,
the epidermal Langherans cells, are invaded by Bb in
early LB. Therefore, Langherans cells were stained
immunohisochemically with different markers to inves-
tigate their functional activity. The number of Langher-
ans cells CD1a positive was reduced in erythema
migrans but was normal or slightly elevated in ACA.
In both diseases there was also a marked downregu-
lation of major hisocompatibility complex class II
molecules on Langherans cells. This phenomenon
might be a mechanism that protects against the pre-
sentation of autoantigens and may be the cause of
impaired capacity of Langherans cells to eliminate Bb
antigens, thus explaining the chronicity of LB.30
Other authors have postulated the autoimmune the-
ory 31 for chronic LB. T cell recognition of self antigens
is a key event in the pathogenesis of autoimmune dis-
eases.32 To date, the initial events that trigger autore-
active T cells are unknown. The molecular mimicry
hypothesis predicts that during an infection T cells
that recognize both a microbial antigen and a related
self peptide become activated and cause autoimmune
disease. The hypothesis that the T cell response to one
or more antigens of Bb is different in patients with
treatment-responsive or treatment-resistant Lyme
arthritis was tested. Results from this study demon-
Agosto 2005
LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
strated that treatment resistant patients presented some
alterations in T cell antigen recognition associated
with an up-regulation of HLA molecules.
Other theories have been made on the pathogenesis
of chronic LB ranging from the intacellular infection
persistence to coinfection with other microorganisms
(e.g. Babesia microti) or the existence of Bb species
resistant to antibiotic treatment.33 Most of the clini-
cal manifestations of LB are due to the local presence
of the causative agent, Bb in the affected tissue. How-
ever, the precise means of tissue damage are not well
understood and there is not proof that the organism, live
or dead, is always present. An understanding of the
complex interactions between the organism and the
host can explain manifestations of the disease and the
persistence of symptoms and signs after antibiotic
treatment.34
Whether chronic LB represents continual infection
or it is a post-Lyme disorder is currently unknown.
Some authors have reported about post-Lyme syn-
drome (PLS) in patients who have developed persistent
symptoms after antibiotic treatment. They include
physic and mental fatigue, myalgias, athralgias, pares-
thesias or dysesthesias or memory and mood distur-
bances.8, 35 Related symptoms have been detected both
in seropositive and seronegative patients. Even for PLS
the mechanism of symptoms persistence is not under-
stood. There are also limited informations regarding
to the utility of extended antibiotic treatments for this
disorder. Some authors reported that treatment over
several months appears to be required to achieve sig-
nificant improvement in most patients, but other results
have shown that treatment with long course of antibi-
otics did not improve symptoms more than placebo.36
Patients with PLS report the following features, as
summarised by the Centers for disease control and
prevention (CDC):36
— diagnosis of LB in the past;
— treatment with standard courses of antibiotics
for established acute LB;37
— long-lasting symptoms (for months or even
years).
In particular, PLS is characterised by:
— encephalopathy with memory or mood distur-
bances (at a short date);
— athralgias;
— musculoskeletal pain localised in the back and
cervicalgias;
— chronic fatigue.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TREVISAN
Some risk factors to develop PLS have been report-
ed, such as delayed treatment (over 1 year), high lev-
els of serum immunoglobulin G antibodies and the
presence of multiple bands in Western blots that seems
to be related to aphasia.
Athralgias, mostly persistent knee synovitis, in some
cases is possibly related to the triggering of intrasyn-
ovial autoimmunity. For these patients, there is no evi-
dence of Borrelia infection by culture or detection of
Bb DNA in blood or spinal and synovial fluids. The
IgG positivity is a common feature with chronic LB,
so serological analysis is not a good tool to distin-
guish PLS from chronic LB.38 Different studies have
reported on the antibiotic treatment failure in patients
with PLS.36 In these cases hydroxychloroquine treat-
ment seems to be effective.6, 39
Another clinical syndrome associated to LB is the
treatment-resistant Lyme arthritis. In about 10% of
patients with Lyme arthritis joint inflammation per-
sists for months or even several years after the appar-
ent eradication of the spirochete, Bb, from the joint
with antibiotic treatment.5, 12, 40 A model of molecular
mimicry has been proposed affecting genetically sus-
ceptible individuals to explain this treatment-resistant
course. The majority of patients with treatment-resis-
tant Lyme arthritis have HLA-DRB1*0401 or related
alleles, and the severity and duration of their arthritis
correlate with cellular and humoral immune respons-
es to outer-surface protein A (OspA) of the spirochete.
Using an algorithm, the immunodominant epitope of
OspA presented by the DRB1*0401 molecule was
predicted to be located at aa 165-173. In a search of the
Genetics Computer Group gene bank, only one human
protein was identified, lymphocyte function associat-
ed antigen-1 (hLFA-1), that had sequence homology
with OspA(165-173)and predicted binding in the
DRB1*0401 molecule. Synovial fluid T cells from
most patients with treatment-resistant arthritis respond-
ed to both OspA and hLFA-1, whereas those from
patients with other forms of chronic inflammatory
arthritis did not. Molecular mimicry between a dom-
inant T cell epitope of OspA and hLFA-1 may be an
important factor in the persistence of joint inflamma-
tion in genetically susceptible patients with treatment-
resistant Lyme arthritis.12, 40, 41
Several studies have been performed to identify pos-
sible sites of bacterial persistence in patient with treat-
ment resistant Lyme arthritis. Among them, PCR analy-
sis in DNA obtained from urine, synovial fluid and
421
TREVISAN LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
membrane demonstrated that Bb DNA was not
detectable in synovial fluid after antibiotic treatment.
However, in patients with ongoing or recurring Lyme
arthritis after antibiotic treatment a negative Bb PCR
in synovial fluid or urine does not exclude a persisting
infection. In these patients, in fact synovial membrane
could be positive for Bb PCR detection.11
It has also been reported in literature about B garinii
seronegative arthritis.42
Chronic fatigue syndrome
Chronic fatigue syndrome (CFS) is a poorly under-
stood condition characterized by debilitating fatigue
and associated symptoms lasting at least 6 months.
Studies indicate that the illness is not simply a mani-
festation of an underlying psychiatric disorder, but
rather is an illness characterized by activation of the
immune system, various abnormalities of several hypo-
thalamic-pituitary axes, and reactivation of certain
infectious agents.43 Many searchers have investigated
factors associated with CFS. Infections may play a
part in ongoing symptomatology in a minority of
patients but most of the agents proposed, such as EBV,
enteroviruses, fungi (Candida albicans) and bacteria
(Chlamydia pneumoniae) have not been found to be
important in the etiology of the disease.
The main feature of the disease is a debilitating
fatigue reducing activity to less than 50% of the
patient's premorbid activity for at least 6 months. The
disease could be persistent or recurrent.44
In addition, other symptoms could be present (at
least 4):
— neurophysiological disorders such as lack of con-
centration and forgetfulness;
— pharyngitis;
— painful cervical or axillary lymphadenopathy;
— mild fever or chills;
— myalgias or muscle discomfort or pain;
— arthralgias;
— headache;
— sleep disturbance;
— prolonged generalised fatigue after usual levels
of activity.
Exclusion criteria: chronic ongoing psychiatric ill-
ness that preceded the development of chronic fatigue
and other diagnosis to explain symptoms of CFS.
Many physical illnesses may have fatigue as a symp-
422 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TABLE II.—Chronic fatigue syndrome: differential diagnosis.
Autoimmune diseases
Localised infections
Chronic inflammatory diseases (Sarcoidosis)
Malignant tumors
Chronic or sub-acute infections (Lyme)
Neuromuscular diseases
Endocrine diseases
Parasitosis
Mycosis
Psychiatric disturbances
HIV infection
Side effects of long-lasting therapies
tom. It is clearly important to exclude common cause
of prolonged fatigue such as anemia. A good history
and examination may point out the way to other poten-
tial causes such as arthritis and diabetes mellitus.
Numerous tests have been suggested for a patient pre-
senting prolonged fatigue. Referrals to a specialist
infectious diseases clinic have not, however, found
batteries of tests to be particularly helpful. In particu-
lar, for a differential diagnosis with LB the exclusion
criteria are mainly related to laboratory tests such as
serology and PCR analysis.
Chronic fatigue syndrome: differential diagnosis
In Table II is shown the differential diagnosis of
CFS.44, 45
According to the clinical aspect of the disease it is
difficult to distinguish among chronic LB, fibromyal-
gias and CFS, because each disorder describes symp-
tomatic pain and fatigue.46 In comparison with LB,
CFS and fibromyalgias present more generalised and
debilitating symptoms, comprising weakness, strong
headache, widespread muscle pain, arthralgia, mus-
culoskeletal pain, symmetrical painful sites in char-
acteristic areas, sleep disturbance and lack of con-
centration.47 On the other side, these patients don't
present arthritis and previous diagnosis of LB, more-
over they have normal neurological tests and they are
more anxious and depressed in comparison with chron-
ic neuroborreliosis patients.
Patients with CFS and PLS share many features.
The neuropsychiatric differences have been examined
in these disorders to enhance understanding of how
mood, fatigue, and cognitive performance interrelate
in chronic illness. Despite the symptoms overlap,
Agosto 2005
LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
patients with PLS show greater cognitive deficits than
patients with CFS compared with healthy controls.
This is particularly apparent among patients with PLS
who lack premorbid psychiatric illness.48
Diagnosis
Diagnosis of LB is mainly based on clinical and
epidemiological criteria, supported by laboratory
tests.3, 15 In routine patient management with typical
early skin lesion the diagnosis of LB in endemic area
is purely clinical. In these cases laboratory testing is not
necessary. For the other manifestations of LB labora-
tory support is essential to confirm the infection. In
more complicated cases it is necessary to evaluate
more aspects. They include the clinical picture, other
preceding or concomitant diseases of the Lyme bor-
reliosis complex, serodiagnostic results, CSF find-
ings, demonstration of intrathecal specific antibody
synthesis, results of PCR analysis, response to ade-
quate antibiotic therapy and exclusion of other dis-
eases. The significance of each of these criteria depends
on the clinical involvement and on the stage of Lyme
borreliosis. Laboratory diagnosis is possible with direct
or indirect methods. Indirect methods to assess Bb
antibodies in serum, synovial and cerebral fluids are
serological tests. These comprise ELISA, immuno-
fluorescence assays (IFA) and Western blotting. A
two-step serological approach has been proposed to
increase specificity. A positive or equivocal first test
(ELISA or IFA) is followed on the same serum sam-
ple by an immunoblot test which can detect IgM and
IgG antibodies to individual Bb antigens.
Direct laboratory diagnosis is related to histological
and immunohistochemical techniques, cultivation and
hybridiastion using fresh tissues and biological flu-
ids.15 Routine laboratory tests, including VES are usu-
ally normal. VES value is an important aspect for the
differential diagnosis with LES and rheumatoid arthri-
tis. Seronegativity does not exclude LB diagnosis,49 as
reported for Lyme arthritis.41 In suspected cases for
late LB, a positive serology is fundamental.50 In about
75% of patients a negative ELISA and a positive West-
ern blot was reported. Most patients with Lyme arthri-
tis are IgG positive both in ELISA and Western blot.
The appearance and evolution of IgM and IgG anti-
bodies to Bb was investigated in patients with erythe-
ma migrans. The first immune response resulted against
flagellin (41kDa) and OspC (24 kDa) in fact the most
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TREVISAN
frequent IgM bands were of 24 kDa (OspC) and 41
kDa. No IgG is typically detected in this phase of the
disease. IgM to antigens of 60 and 66 kDa could be
revealed with the further sero-conversion to IgG anti-
bodies against 24 and 41 kDa antigens. With clinical
manifestation of LB the presence of antibodies against
24 and 41 kDa antigens may be of assistance in con-
firming the diagnosis.6, 29, 51 The antigen of 41 kDa is
not a characteristic only of Bb, but 24 kDa. Other anti-
gens are characteristic of Bb such as 39, 83 or 93 kDa,
The immune response to the recombinant outer sur-
face protein A (OspA- 31 kDa) occurs about 1 year
after the infection. With the resolution of symptoms the
IgM level usually disappears or decreases but some-
times IgG can persist.6, 29.
Some patients could be symptomatic despite nega-
tive western blot.52 The absence of immune response
could be explained by the intracellular localisation of
Bb, that in this way evades the immune system.25, 53, 54
Recently, a sensitive and specific ELISA was intro-
duced in which the antigen is a 26-mer peptide with-
in the sixth invariant region (IR6) of the Vlse (Variable
major protein like sequence Expressed) outer-surface
lipoprotein of Bb.55 The outer cell membrane of Bb
contains many polypeptides, the most extensively stud-
ied are the outer surface proteins Osp. OspA and B
proteins are expressed in vector, but not in vertebrate
host. OspC and VlsE are expressed in vivo in vertebrate
host.56 VlsE is an outer surface lipoprotein of Bb that
undergoes antigenic variation through an elaborate
gene conversion mechanism and is thought to play a
major role in the immune response to Borellia in Lyme
disease. The surface localization of the variable amino
acid segments appears to protect the conserved regions
from interaction with antibodies and hence may con-
tribute to immune evasion.57, 58
VlsE has a predicted molecular mass of 34. Two
invariable domains, one at the amino terminus and the
other at the carboxyl terminus, encompass together
approximately one-half of this molecule's length. The
remainder is composed of a central variable domain
that contains 6 variable regions (VRs) and 6 invari-
able regions (IRs). These 2 types of regions are inter-
spersed with each other, and each constitutes about
one-half of the variable domain's length.
The coding sequence of VlsE contains 1 vls cas-
sette region in the middle and 2 noncassette regions.
Most sequence differences among the vls cassettes
are confined within 6 highly variable regions.57, 58 DNA
segments of the silent cassettes are able to recombine
423
TREVISAN LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
in an apparently random manner into the vlsE cassette
region throughout the course of infection. Sequence
results are consistent, with roughly 6 to 11 recombi-
nation events with multiple silent vls cassettes during
the first 28 days of infection. The promiscuous recom-
bination events at the vlsE site lead to extensive genet-
ic and antigenic variation in VlsE variants.57, 58
Recombination events between the expressed and
silent vmp genes lead to antigenic variation and thus
evasion of the host immune response during the course
of mammalian infection. Unlike the invariable regions
of variable major protein, which are not antigenic in
natural infections, the most conserved of the IRs, IR6,
is immunodominant in Lyme disease patients. IR6 is
exposed on the surface of VlsE, as assessed by
immunoprecipitation experiments, but is inaccessible
to Ab on the spirochete's outer membrane.58 VlsE thus
significantly departs from the antigenic variation par-
adigm, whereby immunodominance is only manifest
in variable portions. IR6 may contribute to divert the
Ab response from other, perhaps protective regions
of VlsE.
The major limitation in serologic tests is that they do
not reliably distinguish between active and past infec-
tion. With these tests the IgG and sometimes IgM
response may persist for years after successful antibi-
otic treatment. A recent study reported that IgG anti-
VlsE antibody titres wane rapidly after successful
antibiotic treatment in both humans and experimental
animal.59 However, other authors reported that the
Anti-VlsE response persisted for months or years after
antibiotic treatment so their presence cannot be equat-
ed with spirochetal persistence in LB.60
Direct methods of laboratory diagnosis are particu-
larly useful in the early phases of the disease when
antibody titre is absent or low. Ideally, the detection of
Bb by cultivation is the standard goal to prove the
infection, but the sensitivity of this method is inade-
quate for diagnosis. Positive Bb colture results some-
times from skin biopsies, serum, synovial and cerebral
fluids.15 PCR is a good tool for Bb detection espe-
cially for differential diagnosis of suspected LB.15
Therapy
Antibiotic therapy is more effective in the early
phases of the disease. The results of treatment depends
not only on the location, extent, and duration of clin-
ical manifestations but also on several other factors
424 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
including the choice of antibiotic, dosage, duration of
treatment, potential or adverse effects and compli-
ance.15 Amoxicillin and doxycycline are the treatment
of first intention in early LB,3 but not for late disease.
Some patients treated with standard courses of these
antibiotics recover only partly. For other patients symp-
toms persist during and after treatment. In these cas-
es longer duration of treatment has been proposed.36
Lyme arthritis could be treated with oral or par-
enteral antibiotic therapy (amoxicillin and doxycy-
cline).4, 61 Treatment with these 2 antibiotics lasting
even 28 days was reported with positive response.
Nervous system involvement and Lyme carditis are
usually treated with ceftriaxone and penicillin G or
cefuroxime intravenously.
Treatment for children consists of conventional
antimicrobial therapy - either orally administered
amoxicillin, doxycycline (major than 12 years old),
erythromycin, or penicillin or intravenously adminis-
tered ceftriaxone, better if children are older than 8
years.3, 4
In most patients with Lyme arthritis, antibiotic ther-
apy is curative, but patients with persistent symptoms
after the first cycle of antibiotic treatment could be
retreated for 4 weeks with the previous oral therapy,
otherwise for 2 weeks intravenously with ceftriaxone
or cefotaxime. For patients treatment-resistant arthri-
tis, in which antibiotic therapy is ineffective; arthro-
scopic synovectomy may be considered to reduce artic-
ular inflammation.4
In cases of neuroborreliosis with central nervous
system involvement a lumbar puncture is recommended
before antibiotic treatment with ceftriaxone and peni-
cillin G or cefuroxime intravenously (2 g/die for 2-4
weeks).3, 4
The appropriate treatment of patients with chronic
Lyme disese is a controversial clinical issue. With the
possibility that chronic LB is due to persistent infec-
tion, the use of antibiotics with intracellular penetra-
tion capability such as macrolides and tetracycline is
proposed.6, 39 The location of Bb in vivo is unknown,
but increasing number of microbes with reactivation
potential are located in lysosomes or other acidic endo-
somes. Macrolide alone has limited activity in acid
environment. Usually, macrolides have not been used
in the treatment of LB. A recent study reported the
use of macrolides in conjunction with lysosomotrop-
ic agents which can alter the pH of acidic intacellular
compartements such as lysosome. With this assum-
Agosto 2005
LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
tion, patients were treated with hydroxychloroquine as
lysosomotropic agent.39, 62 Treatment over several
months appears to be required to achieve significant
improvement in most patients with chronic LB. In
recent observations the duration of treatment appears
to be closer to 12-18 months. However, some patients
with chronic LB have no benefits from the use of this
therapy.
Even if late, chronic and post manifestations of
Lyme disease have been identified, difficulties in diag-
nosis of late stages of Lyme disease persist due to low
sensitivity of serological testing and late inclusion of
Lyme disease in the differential diagnosis. A special
attention to clinical spectrum and laboratory criteria is
required in these cases. Based on experimental evi-
dence and experience more additional work is needed
to improve the understanding of the underlying patho-
physiology of the disease, its diagnosis and treatment.
Acknowledgements.—The authors wish to thank Dr. S. Mier-
tusova for the English revision of the manuscript.
References
1. Trevisan G, Cinco M. Lyme disease: a general survey. Int J Dermatol
1990;29:1-8.
2. Steere AC. Lyme disease. N Engl J Med 2001;345:115-25.
3. Hengge UR, Tannapfei A, Tyring SK. Lyme borreliosis. Lancet
2003;3:489-500.
4. Wormser GP, Nadelman RB, Dattwyler RJ, Dennis DT, Shapiro ED,
Steere AC et al. Practice guidelines for the treatment of Lyme disease.
Clin Infect Dis 2000;31 Suppl 1:1-14.
5. Carlson D, Hernandez J, Bloom BJ, Coburn J, Aversa JM, Steere AC.
Lack of Borrelia burgdorferi DNA in synovial samples from patients
with antibiotic treatment - resistant Lyme arthritis. Arthritis Rheum
1999;42:2705-9.
6. Donta S. The existence of chronic Lyme disease. Curr Treat Options
Inf Dis 2001;3:261-2.
7. Preac-Mursic V, Weber K, Pfister HW, Wilske B, Gross B, Baumann
A et al. Survival of Borrelia burgdorferi in antibiotically treated
patients with Lyme borreliosis. J Infection 1989;17:355-9.
8. Steiner I. Treating post Lyme disease. Neurology 2003;60:1888-9.
9. Trevisan G, Stinco G, Nobile C, Bonin S, Stanta G. Detection of Bor-
relia burgdorferi in skin biopsies from patients with morphea by poly-
merase chain reaction. J Eur Acad Dermatol Venereol 1996;6:15.
10. Keller TL, Halperin JJ, Whitman M. PCR detection of Borrelia
burgdorferi DNA in cerebrospinal fluid of Lyme neuroborreliosis
patients. Neurology 1992;42:32-42.
11. Priem S, Burmester GR, Kamradt T, Wolbart K, Rittig MG, Krause A.
Detection of Borrelia burgdorferi by polymerase chain reaction in
synovial membrane, but not in synovial fluid from patients with per-
sisting lyme arthritis after antibiotic theraphy. Ann Rheum Dis
1998;57:118-21.
12. Steere AC, Gross D, Meyer AL. Autoimmune mechanisms in antibiot-
ic treatment - resistant Lyme arthritis. J Autoimmun 2001;16:263-8.
13. Wahlberg P, Granlund H, Nyman D, Panelius J, Seppala I. Late Lyme
borreliosis: epidemiology, diagnosis and clinical features. Ann Med
1993;25:349-52.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TREVISAN
14. Trevisan G, Cattonar P, Nobile C, Perkan V. Dermatological mani-
festations of Lyme borreliosis. Acta Dermatovenereologica APA
1996;5:101.
15. Trevisan G, Stinco G. Dermatologia di importazione. In: Veraldi S,
Rizzitelli G, Caputo R editors. Infezione da batteri: Borreliosi. Milano:
Poletto; 2000. p. 2-21.
16. Kaya G, Berset M, Prins C, Chavaz P, Saurat JH. Chronic Borrelio-
sis presenting with morphea and lichen sclerosus et atrophicus - like
cutaneous lesions. A case report. Dermatology. 2001;202:373-5.
17. Trevisan G, Menni S, Stinco G, Nobile C, Pistritto G, Perin R. Lichen
slerosus et atrophicus and Borrelia burgdorferi infection. Eur J Pedi-
at Dermatol 1994;4:159.
18. Trevisan G. Atypical dermatological manifestations of Lyme borre-
liosis. Acta Dermatovenereologica 2001;10.
19. Kristoferistsch W, Mayr WR, Partsch H, Neumann R, Stanek G. HLA-
DR in Lyme borreliosis. Lancet 1986;2:278.
20. Steere AC, Dwyer E, Winchester R. Association of chronic Lyme dis-
ease with HLA-DR4 and HLA-DR2 alleles. N Engl J Med
1990;323:219.
21. Duray PH. Clinical pathologic correlations of Lyme disease. Rev
Infect Dis 1989; 11 Suppl 6: S1487-93.
22. Ogrinc K, Lotric-Furlan S, Maraspin V, Cimperman J, Ruzic- Sabljio,
Strle F et al. Cerebrospinal fluid findings in patients with symptoms
suggesting chronic Lyme borrliosis. Ien Klin Wochenschr
2002;114:535-8.
23. Cox J, Krajden M. Cardiovascular manifestations of Lyme disease. Am
Heart J 1991;122:1449-55.
24. Bertuch AW, Rocco E, Schwartz EG. Lyme disease: ocular manifes-
tations. Ann Ophtalmol 1988;20:376-8.
25. Donta S. Late and chronic Lyme disease. Med Clin N Am 2002;86:
341-9.
26. Vrethem M, Hellblom L, Widlund M, Ahl M, Danielsson O, Ernerud
J et al. Chronic symptoms are common in patients with neuroborre-
liosis: a questionnaire follow- up study. Acta Neurol Scand
2002;106:205-8.
27. Elkins LE, Pollina DA, Scheffer SR, Krupp LB. Psychological states
and neuropsychological performances in chronic Lyme disease. Appl
Neuropsychol 1999;6:19-26.
28. Westervel HJ, McCaffrey RJ. Neuropsychological functioning in
chronic Lyme disease. Neuropsychol Rev 2002;12:153-77.
29. Seltzer EG, Gerber MA, Cartter ML, Freudigman K, Shapiro ED.
Long-term outcomes of persons with Lyme disease. JAMA 2000;
283:609-16.
30. Silberer M, Koszik F, Stingl G, Aberer E. Downregulation of class II
molecules on epidermal Langerhans cells in Lyme borreliosis. Br J
Derm 2000;143;786-94.
31. Behar SM, Porcelli SA. Mechanisms of autoimmune disease induc-
tion: the role of the immune response to microbial pathogens. Arthri-
tis Rheum 1995;38:458-76.
32. Hemmer B, Gran B, Zhao Y, Marques A, Pascal J, Tzou A et al. Iden-
tification of candidate T - cell epitopes and molecular mimicry in
chronic Lyme disease. Nat Med 1999;5:1346-7.
33. Sweeney CJ, Ghassemi M, Agger WA, Persing DH. Coinfection with
Babesia microti and Borrelia burgdorferi in a western Wisconsin res-
ident. Mayo Clin Proc 1998;73:338-41.
34. Sigal LH. Immunologic mechanisms in Lyme neuroborreliosis: the
potential role of autoimmunity and molecular mimicry. Semin Neu-
rol 1997;17:63-8.
35. Klempner MS. Controlled trials of antibiotic tratment in patients with
post- treatment chronic lyme disease. Vector Borne Zoonotic Dis
2002;2:255-63.
36. Krupp LB, Hyman LG, Grimson R, Coyle PK, Melville P, Ahnn S et
al. Study and treatment of post Lyme disease (STOP - LD). Neurol-
ogy 2003;60:1923-30.
37. Datttwyler RJ, Halperin JJ, Volkman DJ, Luft BJ. Treatment of late
Lyme boreliosis-randomized comparison of ceftriaxone vs Penicillin.
Lancet 1988;2:1191-4.
425
TREVISAN LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
38. Weinstein A, Britchkov M. Lyme arthritis and post - Lyme disease syn-
drome. Curr Opin Rheumatol 2002;14:383-7.
39. Donta S. Macrolide therapy of chronic Lyme disease. Med Sci Monit
2003;9:PI136-42.
40. Guerau de Arellano M, Huber BT. Development of autoimmunity in
Lyme arthritis. Curr Opinion Rheumatol 2002;14:388-93.
41. Akin E, Aversa J, Steere A. Expression of adhesion molecules in syn-
ovia of patients with treatment-resistant lyme arthritis. Infect Immun
2001;69:1774-80.
42. Dejmkova H, Hulinska D, Tezgova D, Pavelka K, Gatterova J, Vavrik
P. Seronegative Lyme arthritis caused by Borrelia garinii. Clin
Rheumatol 2002;21:330-4.
43. Craig T, Kakumanu S. Chronic fatigue syndrome: evaluation and
treatment. Am Fam Physician 2002;65:1083-90.
44. Fukuda K, Staus SE, Hickie I, Sharpe MC, Dobbins JG, Komaroff A.
The chronic fatigue syndrome: a comprehensive approach to its def-
inition and study. International chronic fatigue syndrome study group.
Ann Intern Med 1994;121:953-9.
45. Abbey S, Garfinkel P. Chronic fatigue syndrome and depression:
cause, effect or covariate. Rev Infect Dis 1991;13 Suppl 1: S73-83.
46. Hsu VM, Patella SJ, Sigal LH: Chronic Lyme disease as the incorrect
diagnosis in patients with fibromyalgia. Arthritis Rheum 1993;36:
1493-500.
47. Fallon BF. Differential diagnosis in Lyme disease. 12th Internation-
al Conference on Lyme disease and other spirochetal and tick-borne
disorders, 1999, 2-10 April, New York, NY.
48. Gaudino EA, Coyle PK, Krupp LB. Post-Lyme disease syndrome
and chronic fatigue syndrome. Neuropsychiatric similarities and dif-
ferences. Arch Neurol 1997;54:1372-6.
49. Dattwyler RJ, Volkman DJ, Luft BJ, Halperin JJ, Thomas J, Golight-
ly MG. Seronegative Lyme disease. Dissociation of specific T- and B-
lymphocyte responses to Borrelia burgdorferi. N Engl J Med
1988;319:1441-6.
50. Dorward DW, Schwan TG, Garon CF. Immune capture and detection
of Borrelia burgdorferi antigens in urine, blood, or tissues from infect-
ed ticks, mice, dogs, and humans. J Clin Microbiol 1991;29:1162-70.
51. Karlsson M. Aspects of the diagnosis of Lyme borreliosis. Scand J
Infect Dis Suppl 1990;67:1-59.
Malattia di Lyme tardiva e cronica e sindrome post malattia di Lyme.
Diagnosi clinica e di laboratorio
L a borreliosi di Lyme (BL) è un’infezione multisistemi-
ca causata da Borrelia burgdorferi (Bb), una spiroche-
ta trasmessa all’uomo dal morso di una zecca dura del gene-
re Ixodes, che interessa principalmente la cute, le articola-
zioni, il sistema nervoso, il cuore e l’occhio 1, 2.
La BL è in genere suddivisa in 3 stadi clinici, oppure in 1
stadio precoce e 1 tardivo 3.
I criteri clinici ed epidemiologici, supportati da indagini
sierologiche, istologiche e colturali, costituiscono le basi
fondamentali per porre una diagnosi certa della malattia di
Lyme, diagnosi che risulta essere relativamente semplice
nelle forme precoci associate alla malattia 3.
Tuttavia, manifestazioni cutanee, nervose, osteoarticola-
426 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
52. Donta S. Tetracycline therapy for chronic Lyme disease. Clin Infect
Dis 1997;25 Suppl 1:S52- 6.
53. Nanagara R, Duray PH, Schumacher HR Jr. Ultrastructural demon-
stration of spirochetal antigens in synovial fluid and synovial mem-
brane in chronic Lyme disease: possible factors contributing to per-
sistence of organisms. Hum Pathol 1996;27:1025-34.
54. Donta St. Treatment of chronic Lyme disease with macrolide antibi-
otics. 8th International Conference on Lyme Borreliosis, 1999, June
20-24, Munich, Germany.
55. Liang FT, Steere AC, Marques AR, Johnson BJ, Miller JN, Philipp MT.
Sensitive and specific serodiagnosis of Lyme disease by enzime-
linked immunosorbent assay with a peptide based on an immun-
odominant conserved region of Borrelia burgdorferi VlsE. J Clin
Microbiol 1995;37:3990-6.
56. Ohnishi J, Piesman J, De Silva AM. Antigenic and genetic hetero-
geneity of Borrelia burgdorferi populations transmitted by icks. Proc
Natl Acad Sci 2001;98:670-5.
57. Lawrenz MB, Hardham JM, Owens RT, Nowakowski J, Steere AC,
Wormser GP et al. Human antibody responses to VlsE antigenic vari-
ation protein of Borrelia burgdorferi. J Clin Microbiol 1999;37:3997-
4004.
58. Eicken C, Sharma V, Klabunde T, Lawrenz MB, Hardham JM, Nor-
ris SJ et al. Crystal structure of Lyme disease variable surface antigen
VlsE of Borrelia burgdorferi. J Biol Chem 2002;277:21691-6.
59. Philipp MT, Bowers LC, Fawcett PT, Jacobs MB, Liang FT, Mar-
ques AR et al. Antibody response to IR6, a conserved immunodom-
inant region of the VlsE lipoprotein, wanes rapidly after antibiotic
treatment of Borrelia burgdorferi infection in experimental animals
and in humans. J Infect Dis 2001;184:870-8. Epub 2001 Aug 30.
60. Peltoma M, McHugh G, Steere AC. Persistence of the antibody
response to the VlsE sixth invariant region( IR6) peptide of Borrelia
burgdorferi after successful antibiotic treatment of Lyme disease. J
Infect Dis 2003;187:1178-86.
61. Cimmino MA, Accardo S. Long term treatment of chronic Lyme
arthritis with benzathine penicillin. Ann Rheum Dis 1992;51:1007-8.
62. Maurin M, Benoliel AM, Bongrand P, Raoult D. Phagolysosolal alka-
linization and the bactericidal effect of antibiotics: the Coxiella bur-
netii paradigm. J Infect Dis 1992;166:1097-102.
ri si possono manifestare anche dopo un lungo periodo dal-
l’avvenuta trasmissione della spirocheta 4 o, addirittura, in
seguito a un’apparente eradicazione della stessa 5. In alcuni
casi, si associa un riscontro laboratoristico di elevati valori
di IgM, espressione, come è noto, di attività di malattia, sen-
za che sia stata fatta chiarezza sul probabile meccanismo
patogenetico che causa tale persistente attività 6.
In altri pazienti si segnala la persistenza dei sintomi, nono-
stante un ciclo completo di terapia antibiotica 7 adeguata e
nonostante l’assenza di segni obiettivi o di marker biologi-
ci 8, il tutto legato, probabilmente, a una persistenza intra-
cellulare della Bb che riesce così a eludere la risposta immu-
nitaria dell’ospite o allo stesso tropismo della Bb che può per-
Agosto 2005
LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
sistere in siti privilegiati (cute 9, liquor 10, liquido sinovia-
le 11) ed eludere la risposta immunitaria. Una piccola per-
centuale di pazienti geneticamente predisposti, poi, si presenta
resistente al trattamento, manifestando un’infiammazione
articolare persistente nonostante l’apparente eradicazione
della spirocheta 3, 12.
Negli ultimi anni si è posta l’attenzione, quindi, su un
complesso di sindromi cliniche correlate alla BL che pre-
sentano varie similitudini e sulla cui patogenesi sono state for-
mulate e vagliate varie ipotesi, legate ora alle caratteristi-
che proprie della spirocheta ora alla risposta immunitaria
nei confronti della stessa.
Tali sindromi sono:
— malattia di Lyme tardiva;
— malattia di Lyme cronica;
— artrite di Lyme resistente al trattamento;
— post malattia di Lyme;
— sindrome della stanchezza cronica.
Malattia di Lyme tardiva
La malattia di Lyme tardiva, definita anche “late Lyme
disease” compare almeno 7 mesi dopo l’infezione: le mani-
festazioni cutanee sono prevalentemente atrofo-scleroder-
miche e l’interessamento extracutaneo è di tipo sia articola-
re che neurologico 13.
Manifestazioni cutanee
Acrodermatite cronica atrofica (o malattia di Pick - Herxhei-
mer)
L’acrodermatite cronica atrofica (ACA) è una manife-
stazione cutanea classica della forma cronica e compare
dopo mesi o anni. Inizia, insidiosamente, con una fase
infiammatoria precoce sotto forma di placche eritemato-cia-
notiche, infiltrate a livello delle superfici estensorie delle
estremità, specie in prossimità delle articolazioni uni o
bilateralmente 14.
Il decorso della malattia è cronico, le lesioni iniziali ten-
dono progressivamente ad allargarsi coinvolgendo l’intera
superficie acrale e tendono a divenire atrofiche. La cute
diviene progressivamente liscia, sottile, trasparente e ane-
lastica. L’atrofia può coinvolgere anche il sottocute e il tes-
suto muscolare sottostante con grave compromissione degli
arti colpiti. Sono state descritte ulcerazioni croniche e tra-
sformazioni maligne della cute atrofica. Possono essere pre-
senti prurito o bruciore, ma la malattia può anche decorrere
senza alcun sintomo 14, 15.
Altre manifestazioni cutanee di tale forma sono: lichen
sclero-atrofico 16, 17, sclerodermia generalizzata, atrofoder-
mia di Pierini-Pasini, panniculite nodulare di Pfeifer-Weber-
Chrstian 15, 18.
Le dermatiti atrofosclerodermiche come il lichen sclero-
sus et atrophicus, la morfea, la sclerodermia in placche, la
sclerodermia lineare, l’atrophoderma profundum di Pieri-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TREVISAN
ni-Pasini, l’emiatrofia facciale di Parry-Romberg, lo scle-
redema di Busckhe e la fascite eosinofila di Schulmann sono
state ripetutamente segnalate, quali manifestazioni tardive del-
la BL, ma, mentre per l’ACA il rapporto con l’infezione da
Bb sembra costante, con, in alcuni casi, l’isolamento della Bb
da biopsie cutanee anche a 10 anni dall’insorgere della BL,
per gli altri quadri atrofo-sclerodermici la correlazione con
la BL è molto discussa 18.
La morfea, in particolare, è stata più volte associata alla
forma tardiva della BL; conosciuta anche come sclerodermia
localizzata, è caratterizzata da un indurimento localizzato del-
la cute e del tessuto sottocutaneo dovuto a un’eccessiva depo-
sizione di collagene 16. I vari tipi di morfea si classificano in
base alle caratteristiche cliniche e al grado di profondità del tes-
suto coinvolto. Essi includono forme a placca, generalizzate,
lineari e forme più profonde. A differenza della sclerosi siste-
mica, nella forma localizzata non si hanno mai sclerodattilia,
il fenomeno di Raynaud o coinvolgimento di organi interni. In
alcuni pazienti, è stato possibile individuare tramite metodi-
che di polymerase chain reaction (PCR) il DNA della Bb in bio-
psie cutanee di pazienti con morfea 9.
Sistema muscolo-scheletrico: artrite cronica
(durata >1 anno)
È la manifestazione clinica più caratteristica e più fre-
quente nel Nord America, meno frequente in Europa 3.
Può comparire qualche settimana ma, addirittura, anche
dopo anni dall’inoculazione della Bb. Il disturbo può dive-
nire cronico o intermittente, con attacchi che durano da un
paio di settimane a qualche mese e poi si risolvono 3. L’in-
tensità degli attacchi diminuisce con il tempo. Non è pre-
sente iperpiressia ma è comune un senso di affaticamento
generale 15.
Caratteristiche 15:
— mono-oligosettorialità;
— asimmetria;
— attacchi frequenti.
Sede 15:
— l’articolazione del ginocchio risulta essere la più col-
pita; in genere sono interessate le grosse articolazioni.
Clinica 15:
— tumefazione;
— marcata impotenza funzionale;
— noduli cutanei;
— assenza di rigidità al mattino.
Esami strumentali 15:
Radiografie delle articolazioni interessate dal processo
patologico possono evidenziare:
— versamento articolare;
— osteoporosi;
— erosioni capi ossei;
— calcificazioni nei tessuti molli periarticolari;
— cisti ossee sottoarticolari.
427
TREVISAN LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
Se non viene trattata, può portare a una irreversibile ero-
sione della cartilagine e dell’osso con conseguente danno
permanente.
Durata della malattia: se non è trattata, persiste per alme-
no 4 anni, fattori di tipo genetico influenzerebbero la dura-
ta e la severità dell’artrite. Soggetti con HLA-DR4 e/o DR
15, 19, 20 sembrano sviluppare con maggiore frequenza un’ar-
trite cronica erosiva scarsamente responsiva alla terapia anti-
biotica e gli stessi possono avere una PCR negativa per la
ricerca del DNA della Bb nel liquido sinoviale, nonostante
la persistenza dell’infiammazione 4, per motivi che saranno
descritti successivamente.
L’artrite cronica viene considerata l’espressione più fre-
quente di Lyme nei bambini (fino al 33%).
Sistema nervoso
Encefalopatia con soli difetti cognitivi. Manifestazioni
tipiche sono disturbi della memoria (a breve termine) e del-
le funzioni intellettive. È presente irritabilità e sonnolenza 15.
Encefalomielite cronica oscillante. Può simulare una scle-
rosi a placche per l’interessamento multifocale del sistema
nervoso centrale (cervello, nervi ottici, tronco encefalico e cer-
velletto) il decorso e l’aspetto alla risonanza magnetica (mul-
tipli focolai nella sostanza bianca periventricolare). È carat-
terizzata da improvvisi deficit focali, transitori o perma-
nenti, quali emiparesi, paraparesi, atassia e afasia 15.
Encefalopatia multi-infartuale: possono essere presenti
difetti neurologici focali 16 acuti transitori (tipo attacco ische-
mico transitorio) o permanenti (tipo ictus).
Polineuropatia assonale 21: molti pazienti affetti da bor-
reliosi tardiva (di solito in associazione con l’artrite) pre-
sentano una lieve neuropatia sensitiva, che si manifesta cli-
nicamente con parestesie intermittenti agli arti. Gli studi
elettromiografici rivelano un quadro multinevritico con velo-
cità di conduzione nervosa generalmente nella norma.
Esami effettuati sul liquido cerebrospinale mostrano nel-
la neuroborreliosi cronica un liquor generalmente normale o
con pleiocitosi 10, 22, con presenza di anticorpi anti Bb risul-
tanti da una sintesi intratecale. Metodiche di PCR possono
consentire l’isolamento genomico della spirocheta.
Cuore
La miocardiopatia dilatativa 23, considerata una compli-
canza molto rara, potrebbe essere la sola ragione di un out-
come fatale nei pazienti con LB.
Occhio
Come conseguenza di una neuroborreliosi o di una con-
dizione infiammatoria dell’occhio possono manifestarsi:
cheratiti, episcleriti, iridocicliti, neurite ottica 3, 24. L’in-
fiammazione intraoculare di lunga durata può portare a cecità.
428 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
Lyme cronico e diagnosi differenziale
Nella Tabella I viene mostrato uno schema riassunto del-
le diagnosi differenziali.
Mentre può essere relativamente semplice riconoscere
disturbi associati alla malattia di Lyme quali l’eritema cro-
nico migrante, la paralisi di Bell o l’artrite (specie se
monoarticolare o oligoarticolare), può essere piuttosto dif-
ficoltoso riconoscere sintomi cronici associati alla malat-
tia di Lyme 25, 26.
Si definisce malattia di Lyme cronica un complesso di
sintomi cronici correlati alla malattia di Lyme che consisto-
no in senso di affaticamento generale con artralgie (senza
alcun segno obiettivo), fibromialgie e altre disfunzioni del
sistema nervoso (neurocognitive e neuropsichiatriche) 27,
28. Possono essere presenti, ma più raramente, sintomi asso-
ciati, di tipo neurologico, quali parestesie, tremori, palpita-
zioni, tachicardia, alterazioni dell’equilibrio, sudorazione
(a volte intensa), disturbi visivi, disturbi gastro-intestinali
(sindrome del colon irritabile) e aumento della frequenza
urinaria 25.
Responsabile del corteo sintomatologico sarebbe la per-
sistenza dell’infezione da Bb.
Rimane ancora da distinguere, soprattutto per ciò che
riguarda il processo patogenetico, tale forma dalla cosid-
detta “late Lyme disease” descritta in precedenza; le 2 sin-
dromi, infatti, non sono ancora perfettamente differenziate,
anzi, spesso considerate sinonimi, costituirebbero, in realtà,
2 entità cliniche diverse 25.
Le differenze principali consisterebbero nel fatto che la
malattia di Lyme tardiva è caratterizzata, contrariamente alla
forma cronica, da segni obiettivi di artrite (tumefazione,
noduli cutanei...) che non sono presenti nell’altra forma e
che la forma tardiva, oltre a essere più facile da diagnosticare,
risulta molto più responsiva al trattamento rispetto alla cro-
nica.
Si è constatato, infatti, che pazienti affetti da malattia di
Lyme tardiva con un’importante risposta anticorpale di tipo
IgG rispondevano meglio alla terapia se confrontati con
pazienti con sintomi cronici e con bassa risposta anticorpa-
le IgG che necessitavano di terapia più prolungata. In gene-
re, in molti pazienti affetti dalla forma cronica si possono ritro-
vare valori alti e persistenti di IgM associati a un titolo anti-
corpale IgG piuttosto limitato 6.
La risposta anticorpale di tipo IgM contro le proteine spe-
cifiche di Bb (23Kd, 31Kd, 34Kd e soprattutto 41Kd) è indi-
cativa di una fase attiva della malattia di Lyme. L’elevata
presenza di titoli IgM nella forma cronica della malattia non
è caratteristica solo della malattia di Lyme; ci sono esempi
di altre infezioni che possono presentare un nuovo innalza-
mento delle IgM dopo l’infezione primaria in caso, ad esem-
pio, di riattivazione della malattia (toxoplasmosi, citome-
galovirus) 6. Nei casi di malattia di Lyme cronica che rispon-
de al trattamento, si può assistere a un progressivo calo del
titolo IgM a favore delle IgG. Rimane, tuttavia, ancora da sta-
bilire il meccanismo che determina la persistente reattività
delle IgM e la limitata reattività delle IgG, è molto probabi-
Agosto 2005
LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
le che sia coinvolto il processo di “switch” dalla risposta
immunitaria di tipo IgM a quella di tipo IgG 6.
Alcuni Autori negano l’esistenza di una forma cronica di
malattia di Lyme, asserendo, piuttosto, che tali pazienti sof-
frano di disturbi psichiatrici; in realtà, studi epidemiologici
confermano l’esistenza di una forma cronica di neuroborre-
liosi che si sviluppa dal 30% al 50% di pazienti che svilup-
pano una serie di disturbi spesso difficilmente distinguibili
dalla fibromialgia e dalla sindrome della fatica cronica, con-
siderando anche il fatto che entrambi questi disturbi posso-
no essere conseguenti alla malattia di Lyme 29.
I sintomi della malattia di Lyme cronica sono in genere per-
sistenti con delle fasi di peggioramento che si verificano
ciclicamente. Alcuni pazienti sono più sintomatici di altri, par-
ticolare che fa pensare a una differenza geneticamente deter-
minata nella risposta all’infezione e al suo coinvolgimento
sistemico.
Si può, quindi, affermare che la malattia di Lyme cronica
non è di certo una malattia distruttiva o fatale ma può esse-
re fortemente debilitante.
L’incidenza dell’infezione asintomatica non è stata valu-
tata, ma è importante precisare che molti pazienti, sebbene
asintomatici per lungo tempo, possono andare incontro a
una riaccensione dell’infezione e, quindi, a una comparsa
di sintomi della malattia di Lyme cronica mesi o anni dopo
aver contratto l’infezione e ciò a causa di eventi scatenanti
quali trauma, gravidanza, o stress di tipo psicologico 25.
Non è noto il meccanismo responsabile della riattivazio-
ne della malattia e numerose sono le teorie sulla patogene-
si della malattia di Lyme cronica. Le caratteristiche immu-
nopatogenetiche presenti nell’ACA forniscono un modello
di studio interessante per meglio comprendere la patogene-
si di tale disturbo.
Recenti studi indicano che le cellule di Langherans del-
l’epidermide sono invase dalla Bb già nelle forme precoci di
malattia di Lyme. Tali cellule sono state studiate immunoi-
stochimicamente con differenti marker per valutare la loro
attività funzionale. Si è visto che le cellule di Langherans atti-
ve sono ridotte nell’eritema migrante ma sono normali o
addirittura elevate nell’ACA. In entrambe le manifestazioni
cliniche, c’è, inoltre, una downregulation di molecole di
classe II del complesso maggiore di istocompatibilità sulle
cellule di Langherans. Questo fenomeno di downregulation
potrebbe essere una sorta di meccanismo protettivo che evi-
ta la presentazione di autoantigeni ma impedisce alle cellu-
le di Langherans di eliminare Bb, da qui deriverebbe l’in-
sorgenza della forma cronica 30.
Secondo la teoria dell’autoimmunità 31, che è descritta
meglio successivamente, esisterebbero dei linfociti T, nati dal-
l’interazione con Bb, che mostrano un’alterazione nella loro
capacità di riconoscimento dell’antigene 32 e potrebbero
essere coinvolti nei sintomi neurologici e muscolo-scheletrici
nelle varie fasi della malattia. Ci sarebbe, infatti, una dele-
zione incompleta di tali linfociti autoreattivi nel timo e ciò,
aggiunto alla up-regulation di molecole HLA, di co-stimo-
latori, di co-recettori provocherebbe il danno.
Altre ipotesi che tentano di spiegare la patogenesi della
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TREVISAN
malattia di Lyme cronica citano l’esistenza di forme parti-
colarmente attive di Bb che sfuggono al controllo degli anti-
biotici o suggeriscono che la forma cronica sia espressione
di un danno causato dalla risposta dell’ospite nei confronti
della spirocheta, o ancora che essa sia dovuta a una confezione
di Bb con un altro microrganismo trasmesso dalla zecca Ixo-
des (Babesia microti, specie di Erlichia) 33.
È noto che le manifestazioni cliniche della malattia di
Lyme sono dovute alla presenza, a livello locale, dell’agen-
te causale ma, a tutt’oggi, non è perfettamente noto il mec-
canismo di danno tissutale che deriva da un’interazione mol-
to complessa tra il batterio in questione e il sistema immu-
nitario dell’ospite 34.
Alcuni pazienti mostrano una persistenza di sintomi, qua-
li stanchezza fisica e mentale, mialgie, artralgie senza artri-
te, disestesie/parestesie, cefalea, vertigini, disturbi della
memoria nonostante l’adeguato trattamento antibiotico 8, 35.
Si parla, in questo caso, di “post-Lyme syndrome” (PLS).
La persistenza di tali sintomi è stata riportata sia in pazien-
ti privi di anticorpi IgG anti Borrelia sia in pazienti con tito-
lo anticorpale positivo.
Il meccanismo patogenetico che causa la persistenza di
tali sintomi non è stato ancora chiarito e, anzi, la questione rima-
ne piuttosto controversa. A causa dell’associazione di tali sin-
tomi con l’infezione borreliosica alcuni pazienti sono stati
sottoposti a terapie antibiotiche prolungate, spesso con il risul-
tato che, alla sospensione del trattamento, avveniva la ricom-
parsa della sintomatologia precedentemente descritta 36.
Un paziente si definisce affetto da PLS quando sono
rispettati i seguenti criteri definiti dai Centers for disease Con-
trol and Prevention (CDC) 36:
— documentata infezione di Lyme in passato;
— ciclo completo di terapia antibiotica adeguata 37;
— persistenza per mesi o anni dei sintomi precedente-
mente descritti.
In particolare, la PLS si presenta con:
— encefalopatia con disturbi della memoria a breve ter-
mine;
— dolore muscolare con particolare coinvolgimento del
dorso e della regione cervicale;
— artralgie.
Probabilmente i sintomi articolari, soprattutto alle ginoc-
chia, sono causati dall’instaurarsi di un meccanismo autoim-
munitario intrasinoviale. È importante considerare che, in tale
forma, i sintomi non si associano ad alcun segno obiettivo né,
tantomeno, a qualsiasi marker biologico. La ricerca della Bor-
relia nel sangue o nel liquor cerebri o sinoviae, tramite esa-
me culturale o tecniche di amplificazione del DNA (PCR) del-
la Borrelia risulta, infatti, negativa, mentre la sierologia può
mantenersi positiva per alcuni anni dopo la guarigione del-
l’infezione borreliosica e, quindi, non consente di distin-
guere accuratamente tra la forma cronica e la PLS 38.
Da vari studi descritti in letteratura emerge che, nella PLS,
il trattamento antibiotico sistemico per via sia endovenosa che
orale è inefficace 36 e, quindi, non necessario, mentre può
essere efficace la terapia con idrossi-clorochina 6, 39.
429
TREVISAN LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
Un’altra sindrome clinica associata alla malattia di Lyme
è la cosiddetta “treatment-resistant Lyme arthritis”, un’artrite
di Lyme che si manifesta nel 10% dei pazienti ed è caratte-
rizzata da un’infiammazione articolare persistente che non
risponde alla terapia antibiotica 12, 40.
Tale infiammazione persiste per mesi o anni dopo l’ap-
parente eradicazione della spirocheta 5.
Studi di biologia molecolare hanno mostrato che tali
pazienti hanno una predisposizione genetica dovuta a un
aplotipo HLA - DRB1* 0401 e relativi alleli e che la seve-
rità e la durata di tale artrite sarebbe altamente correlata alla
risposta immunitaria sia di tipo umorale sia cellulo-media-
ta nei confronti della proteina di superficie della Bb (Osp
A). Ciò deriva dal fatto che, in questi pazienti, tale proteina,
in particolare la sequenza di amminoacidi OspA165 - 173,
ha una sequenza altamente omologa a un peptide contenuto
in una proteina presente normalmente nell’uomo, la hLFA -
1α (lymphocyte function associated antigen - 1alpha). I
linfociti T del liquido sinoviale di questi pazienti con artri-
te resistente al trattamento e geneticamente predisposti deter-
minano una risposta immunitaria nei confronti sia di OspA
sia di hLFA - 1, ciò suggerisce che hLFA - 1α possa agire
da parziale agonista e che abbia un ruolo importante per la
persistenza dell’infiammazione articolare 12, 40, 41.
Sono stati effettuati vari studi per individuare i possibili siti
di persistenza della Bb in pazienti con forme resistenti al
trattamento, comprese analisi di PCR per il DNA della spi-
rocheta effettuate sulla membrana sinoviale in pazienti che
mostravano risultati negativi per la PCR condotta su liquido
sinoviale dopo il trattamento antibiotico. Ebbene, in alcuni
casi, la PCR condotta sulla membrana sinoviale dava esito
positivo suggerendo che, in forme di artrite di Lyme resi-
stenti al trattamento antibiotico, la PCR negativa su liquido
sinoviale non esclude la presenza intrarticolare della Bb 11.
In letteratura sono state descritte anche forme di artrite
sieronegativa da B. garinii 42.
La sindrome da stanchezza cronica
La sindrome da stanchezza cronica (chronic fatigue syn-
drome, CFS), nota anche come sindrome da stanchezza cro-
nica e immunodeficienza, è una malattia caratterizzata da
una persistente fatica cronica accompagnata da una serie di
sintomi sistemici di tipo reumatologico, cognitivo e simil-
infettivo 43.
Non è nota ancora la causa di tale sindrome, varie ricer-
che hanno confermato un possibile ruolo eziologico di alcu-
ne infezioni virali (compresi il virus dell’Epstein barr, ente-
rovirus, virus poliomielitico), funghi (in particolare Candi-
da albicans), batteri (Chlamydia pneumoniae), ma non c’è
evidenza che essa possa essere provocata da un determina-
to e specifico agente eziologico.
Colpisce tutte le razze, le etnie, senza distinzione di clas-
se sociale; si rileva, comunque, una maggiore prevalenza
per il sesso femminile tra la terza e la quarta decade di vita.
La principale caratteristica della malattia è una stanchez-
430 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
za cronica che limita le attività precedenti la malattia del
50% per un periodo di almeno 6 mesi. La malattia può esse-
re persistente o recidivante 44.
Inoltre possono essere presenti altri sintomi (almeno 4):
— compromissione della memoria e della concentrazio-
ne;
— faringite;
— dolorabilità dei linfonodi (cervicali o ascellari);
— febbricola (o sensazione di febbre/brividi);
— dolori muscolari (mialgia);
— dolori articolari multipli (artralgia);
— cefalee di nuovo esordio;
— disturbi del sonno (ipersonnia o insonnia);
— stanchezza successiva all’esercizio fisico.
Criteri di esclusione: altra causa o diagnosi responsabile
della stanchezza o dei sintomi accusati.
In molti casi la valutazione della CFS si concentra sulla
ricerca di una causa infettiva o di un’altra causa specifica
per questa malattia.
Molto importante è la diagnosi differenziale della malat-
tia, che spesso è molto laboriosa, in quanto la stanchezza e
gli altri sintomi generali si manifestano in molte altre malat-
tie; in generale, si tratta di una diagnosi di esclusione, dopo
aver effettuato un’anamnesi completa, un’attenta visita medi-
ca e, soprattutto, una serie di test laboratoristici che escludano
altre cause 43. Nel caso della diagnosi differenziale con la
malattia di Lyme esami sierologici e test diretti (PCR) ci
permettono di porre diagnosi di esclusione.
Diagnosi differenziale nella sindrome da stanchezza
cronica
Nella Tabella II è mostrata la diagnosi differenziale del-
la CFS 444, 45.
Soltanto sulla base clinica è arduo distinguere la malattia
di Lyme cronica e la CFS o fibromialgia 46, questa difficoltà
è dovuta al fatto che una piccola percentuale di pazienti in
effetti sviluppa dolore cronico o la sindrome di affatica-
mento in associazione con o subito dopo la malattia di Lyme.
Rispetto alla malattia di Lyme, la sindrome da fatica croni-
ca o la fibromialgia tendono a produrre sintomi più genera-
lizzati e disabilitanti, comprendenti notevole fatica, forte
mal di testa, dolore muscolare scheletrico diffuso, punti
dolorosi simmetrici in siti caratteristici, dolore e rigidità in
molte articolazioni, disestesia diffusa, difficoltà di concen-
trazione e disturbi del sonno 47. I pazienti con la CFS o la
fibromialgia non presentano evidenza di infiammazione arti-
colare; hanno risultati normali nei test neurologici e hanno
un grado di ansietà e depressione maggiore rispetto ai pazien-
ti con neuroborreliosi cronica, non hanno storia di infezio-
ne di Lyme.
La CFS mostra anche sintomi molto simili alla post-Lyme
disease e, in alcuni studi, effettuati per valutare le differen-
ze di ordine neuropsichiatrico, si è visto che i pazienti con
PLS mostrano una deficienza cognitiva maggiore rispetto
Agosto 2005
LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
ai pazienti con CFS, e ciò avviene soprattutto e ovviamente
in pazienti con PLS e con pregressi disordini di tipo psi-
chiatrico 48.
Diagnosi
La diagnosi della malattia di Lyme si basa principalmen-
te su criteri clinici ed epidemiologici con il supporto di inda-
gini sierologiche, istologiche e colturali 3, 15.
Quando l’affezione segue il suo decorso tipico, l’accer-
tamento diagnostico risulta solitamente agevole; le indagi-
ni laboratoristiche non fanno altro che confermare l’ipotesi
del clinico. Tuttavia, come nelle forme precedentemente
descritte, la diagnosi può risultare estremamente difficile e
raramente certa. Essa potrà essere il risultato della somma di
molti indizi, quali provenienza da zona endemica, dato anam-
nestico positivo per puntura di zecca, sintomatologia clini-
ca, che cercano di cogliere i criteri di diagnosi differenziale
delle forme prima descritte, correlazione clinico temporale
delle manifestazioni patologiche, sfumature del quadro cli-
nico, esclusione di altre affezioni, positività, tecniche di
amplificazione genica.
Nella diagnostica di laboratorio della malattia di Lyme
ci si avvale di metodiche dirette o indirette.
La diagnosi indiretta si basa sulla ricerca degli anticorpi
anti Borrelia nel siero, nel liquido cefalorachidiano e nel
liquido sinoviale di soggetti affetti, con test di primo livello
(ELISA o test immunoenzimatico) e test di secondo livello
(Western-Blot); la diagnosi diretta si avvale di metodiche
istologiche, immunoistochimiche, colturali e di ibridazione
genetica, effettuate su tessuti e liquidi biologici 15.
I test laboratoristici di routine sono in genere normali, la
VES è normale (questo è un importante elemento per la dia-
gnosi differenziale con l’artrite reumatoide e il lupus erite-
matoso sistemico).
La negatività dei test sierologici non esclude la BL 49 e
sono riportate, come descritto precedentemente, segnala-
zioni di artrite di Lyme sieronegativa 41. Tuttavia, quando
si sospetta che una manifestazione clinica sia espressione
di una malattia di Lyme in fase tardiva, la positività sierolo-
gica è un elemento fondamentale 50.
Più del 75% dei pazienti con malattia cronica di Lyme ha
un test ELISA negativo e un Western blot positivo. In gene-
re, i pazienti con artrite oligoarticolare possono presentare
un’intensa e positiva risposta IgG sia con l’ELISA che con
il Western blot.
Da analisi di Western blot risulta che, nella malattia di
Lyme, la prima reazione immunologica si ha nei confronti del-
la proteina flagellare 41-Kd e la proteina OspC 23. In gene-
re, nella fase dell’eritema cronico migrante, si ha un’inten-
sa reazione IgM contro le proteine 23-Kd e 41-Kd e nessu-
na reazione di tipo IgG. Nelle settimane successive persiste
la reazione IgM, a volte accompagnata da reazioni minori
contro le proteine 60-Kd e la 66 Kd, e perciò segue un innal-
zamento delle IgG contro le proteine 23-Kd e 41-Kd. In pre-
senza di un caratteristico quadro clinico, la reazione immu-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TREVISAN
nitaria contro la 23-Kd e la 41-Kd è considerata diagnosti-
ca della malattia di Lyme 6, 29, 51.
La proteina 41 Kd non è caratteristica solo delle Bb, a
differenza della 23-Kd. Altre proteine caratteristiche sono 35-
Kd, 37-Kd, 39-Kd, 83-Kd, e 93-Kd.
Reazioni immunitarie contro la 31-Kd non compaiono
fino a un anno o più dopo aver contratto l’infezione. Con la
risoluzione dei sintomi, scompare o si attenua la reazione
di tipo IgM, mentre la risposta di tipo IgG può persistere
dopo la scomparsa dei sintomi, ma, comunque, in genere, si
attenua o scompare dopo un adeguato trattamento antibioti-
co 6, 29.
Alcuni pazienti possono manifestare i sintomi nonostan-
te il Western blot risulti negativo 52. Questo fatto si spiega con-
siderando che, a volte, la Borrelia rimane all’interno delle cel-
lule, senza avere una fase extracellulare e ciò impedisce che
si generi la risposta immunitaria dell’organismo nei con-
fronti della spirocheta 25, 53, 54.
I test serologici di ultima generazione sono volti a sem-
plificare la serologia, pur mantenendo elevato il valore pre-
dittivo in termini di sensibilità e specificità al punto da poter
sostituire, da solo, la procedura a 2 test (ELISA e Western
blot) finora ritenuta la più predittiva.
Recentemente si è utilizzato, infatti, un nuovo antigene, in
metodologia immuno enzimatica, rappresentato da un cor-
to peptide, detto VlsE 55.
È noto che la Bb presenta sulla membrana esterna una
serie di lipoproteine, le Osps, che vengono espresse in manie-
ra differenziale nell’ospite mammifero e nel vettore. Le pro-
teine OspA e OspB sono espresse nel vettore ma non nel
mammifero, le OspC e le Vlse sono espresse in vivo nel
mammifero 56.
VlsE (variable major protein-like sequence, expressed) è
una proteina di superficie della Bb che riveste un importan-
te ruolo nella diagnosi sierologica della malattia di Lyme in
quanto gioca un ruolo fondamentale nella sopravvivenza
della spirocheta nell’uomo.
Dopo la penetrazione nell’organismo ospite la Bb subisce
importanti modificazioni a livello della VlsE, che le per-
mettono di sfuggire al riconoscimento e all’attivazione del
sistema immunitario 57, 58.
La VlsE è divisa in varie parti: regioni conservate, che
formano il dominio transmembrana e ancorano la proteina alla
membrana batterica, e regioni variabili, che subiscono costan-
temente delle ricombinazioni. Per la produzione della Vlse
il DNA di Bb contiene da 15 a 20 sequenze (cosiddette varia-
ble major protein-like sequenze [vls]) che contengono un
gran numero di informazioni genetiche 57, 58.
Ognuna è composta da 12 sezioni: 6 costanti e 6 variabi-
li. Dalla ricombinazione di elementi differenti nascono pro-
teine di superficie che differiscono nella loro regione varia-
bile. Il DNA che si assembla e che si usa per la sintesi delle
proteine è detto vlsE (E = expressed). Esso contiene anche
i domini transmembrana della VlsE.
VlsE è esclusivamente espressa in vivo nei mammiferi,
quindi non è presente nelle colture di Bb 58.
Le regioni costanti sono mascherate da quelle variabili e sono
431
TREVISAN LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
protette dal diretto attacco del sistema immunitario. Quando
la Bb viene processata dalle cellule che presentano l’antige-
ne, l’intera VlsE viene presentata al sistema immune con la for-
mazione di anticorpi sia contro le regioni conservate che con-
tro quelle variabili. Questi anticorpi in vivo non possono lega-
re la Bb a causa dell’effetto maschera, ma sono particolar-
mente importanti nella diagnosi in quanto agenti contro una
regione fortemente conservata della proteina 58.
Il limite maggiore dei test sierologici è che non consento-
no di distinguere tra infezioni recenti e passate, soprattutto
perché, spesso, il titolo delle IgG, e talvolta delle IgM, può man-
tenersi elevato anche in seguito al trattamento antibiotico.
Un recente studio ha dimostrato che il titolo degli anticorpi
della classe IgG anti- VlsE cala rapidamente dopo la terapia
antibiotica; ciò suggerisce che una diminuzione di tale tito-
lo anticorpale entro 6 mesi dall’infezione e dopo terapia
antibiotica possa essere considerato come un marker indi-
cativo dell’eradicazione della spirocheta (con una funzione
assai simile al VDRL nella sifilide) 59.
Altri studi che valutano la risposta anticorpale al peptide
VlsE dopo terapia antibiotica, sia 6 mesi che anni dopo, nel
siero di pazienti con manifestazioni cliniche di Lyme precoce
o cronica, mostrano che la persistenza di titoli anticorpali
anti Vlse per mesi o anni dopo la terapia antibiotica non può
essere considerata una persistenza dell’infezione da Bb 60.
Le metodiche dirette si rivelano particolarmente utili nel-
le fasi iniziali della malattia di Lyme, quando il movimento
anticorpale non è ancora rilevante, o nei casi in cui la siero-
logia è negativa. L’isolamento colturale resta il gold stan-
dard e, in alcuni casi, la Bb si può coltivare da lesioni cuta-
nee, siero, liquido cerebrospinale e sinoviale 15.
La PCR è una metodica diretta che ricerca la presenza
del DNA di Bb nei tessuti e nei liquidi biologici, ed è un
esame di grande importanza per porre una diagnosi diffe-
renziale tra le varie sindromi associate al Lyme 15.
Terapia
Per ciò che riguarda il trattamento delle forme precoci di
Lyme c’è un accordo generale tra i vari studi pubblicati; la
terapia è essenzialmente di tipo antibiotico e altri farmaci
sintomatici possono essere di volta in volta usati a seconda
dei quadri clinici associati. La scelta deve essere effettuata
in base a quadro clinico, stadio, sintomatologia, età del
paziente, sesso e fattori concomitanti, quali una gravidanza.
La BL è un’infezione multisistemica e tutti i distretti del-
l’organismo sono interessati, perciò è necessario tenere con-
to della capacità del farmaco di raggiungere le spirochete
per svolgere l’effetto terapeutico. Inoltre, il farmaco deve
essere in grado di diffondersi nei tessuti e nei liquidi biolo-
gici, di attraversare la barriera ematoencefalica, di penetra-
re all’interno delle cellule, di legarsi in maniera stabile alle
strutture vitali del germe ed esplicare in questa sede l’attività
antimicrobica 15.
L’amoxicillina e la doxiciclina costituiscono i farmaci di
prima scelta nella malattia di Lyme3 recente, mentre non
432 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
sono di utilità nelle forme croniche il cui trattamento è anco-
ra dibattuto.
Alcuni individui non rispondono alla terapia, cioè, dopo
la terapia, hanno una fase di remissione completa dei sinto-
mi generali, che riappaiono dopo un periodo più o meno
lungo, talora con maggiore intensità.
Altri, nonostante la terapia antibiotica, continuano a mani-
festare la sintomatologia; perciò si preferisce, in tal caso, soli-
tamente somministrare terapie antibiotiche prolungate 36.
In generale, l’artrite di Lyme viene trattata con terapia
antibiotica sistemica orale o parenterale 4, 61.
Doxiciclina o amoxicillina, entrambe somministrate anche
fino a 28 giorni di trattamento, hanno mostrato buoni risul-
tati e sono raccomandate se non vi sono segni clinici di un
coinvolgimento di tipo neurologico, eventualità in cui sarà
necessaria una terapia per via parenterale con cefotaxime o
penicillina G.
Per i pazienti pediatrici più piccoli si utilizza l’amoxicil-
lina, per quelli di età superiore agli 8 anni la doxiciclina 3, 4.
Per pazienti che manifestano disturbi articolari persistenti
dopo un ciclo completo di terapia antibiotica, si preferisce
ripetere il trattamento per altre 4 settimane con antibiotici per
via orale o, alternativamente, con 2 settimane di cefotaxime
endovena. Alcuni Autori sono concordi nel ritenere di atten-
dere qualche mese prima di iniziare il nuovo ciclo di terapia
antibiotica, essendo il processo di risoluzione dell’infiam-
mazione a livello articolare estremamente lento. Se, nono-
stante i ripetuti cicli di terapia antibiotica non si evidenzia
alcun beneficio, si può ricorrere alla sinoviectomia in artro-
scopia che riduce il periodo di infiammazione dell’articola-
zione 4.
Nei pazienti che presentano disturbi di tipo neurologico con
interessamento del sistema nervoso centrale o periferico,
dopo un’attenta valutazione neurologica e l’esecuzione di
una puntura lombare, la raccomandazione è somministrare
il cefotaxime per via parenterale (2 g/die per 2-4 settimane)
o la penicillina G 3, 4.
Il trattamento della forma cronica non è a tutt’oggi ben
definito.
L’obiettivo principale sarebbe utilizzare antibiotici capa-
ci di penetrare all’interno delle cellule, quali i macrolidi e le
tetracicline 6, 39.
In realtà, macrolidi non sono di norma utilizzati nel trat-
tamento della malattia di Lyme, ma alcuni studi conferma-
no la loro efficacia nel trattamento delle forme croniche se
associati a un agente lisosomotropico che alcalinizza l’aci-
dità intracellulare dei lisosomi, l’idrossiclorochina, appun-
to 62. L’attività della clorochina incrementerebbe di molto l’at-
tività dei macrolidi che sono inattivati normalmente dal pH
acido 6, 39.
Il razionale di tale terapia si basa sul fatto che i sintomi del-
la malattia di Lyme cronica sarebbero provocati da una per-
sistenza intracellulare della Bb e, in questo caso, i macroli-
di agirebbero nel distretto intracellulare 39, dove penicilline
e cefalosporine non arrivano.
Si è concordi nel ritenere che il trattamento di tali forme
deve essere piuttosto prolungato, da 12 a 18 mesi. Una pic-
Agosto 2005
LATE LYME DISEASE, CRONIC LYME DISEASE AND POST LYME DISEASE SYNDROME
cola percentuale di pazienti risulta, comunque, non rispon-
dere nemmeno dopo terapie prolungate.
Sebbene in parte definite e confermate, le varie sindro-
mi correlate alla malattia di Lyme richiedono ancora un
maggiore inquadramento nell’ambito di uno spettro clini-
co associato al Lyme e una diagnosi di certezza, che, rela-
tivamente semplice nelle forme precoci, spesso è difficile
se non improbabile; sta al clinico, quindi, cercare di coglie-
re quegli elementi di diagnosi differenziale, sia clinici sia
laboratoristici, tuttora oggetto di discussione tra vari Auto-
ri, per riconoscere e diagnosticare le varie forme cliniche,
escludere sindromi simili e poter scegliere efficienti strategie
terapeutiche.
Riassunto
La borreliosi di Lyme è un’infezione multisistemica
causata da Borrelia burgdorferi che interessa principal-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TREVISAN
mente la cute, le articolazioni, il sistema nervoso, il cuo-
re e l’occhio.
La valutazione clinica, supportata da dati epidemiologici,
indagini sierologiche, istologiche e colturali, costituisce il fon-
damento per porre una diagnosi certa della malattia di Lyme.
Le manifestazioni cutanee, nervose, osteoarticolari si possono
manifestare, in alcuni casi, anche a distanza di tempo dalla
puntura della zecca o, addirittura, in seguito a un’apparente
eradicazione della stessa.
In questo lavoro vengono descritte alcune sindromi clini-
che correlate alla malattia di Lyme che presentano varie
similitudini e sulla cui patogenesi sono state formulate e
vagliate varie ipotesi; l’individuazione dei principali ele-
menti di diagnosi differenziale, clinici e laboratoristici, costi-
tuisce il primum movens per riconoscere e diagnosticare le
varie forme cliniche e scegliere un opportuno trattamento.
PAROLE CHIAVE: Borreliosi di Lyme - Malattia di Lyme tar-
diva - Malattia di Lyme cronica - Artrite di Lyme resistente
al trattamento - Post-malattia di Lyme.
433

Chronic urticaria remains a major problem in terms of patho-
genesis, diagnostic work-up and management. During the last
years several new concepts regarding the disease have been
developed. Some of these aspects resulted in a different man-
agement of patients with chronic urticaria whereas others still
need research activities for confirmation or clarification of details.
Symptoms are the result of the degranulation of mast cells and
basophils. Possible mechanisms include autoimmune mecha-
nisms, infectious diseases, pseudoallergic mechanisms and oth-
ers such as internal diseases/malignancies. A detailed history
plays a main role in the diagnostic program. Further diagnostic
procedures depend on the urticaria subtype. Whereas in acute
urticaria routine diagnostic is not recommended, in chronic
urticaria a diagnostic programm considering associated infections
(particularly with Helicobacter pylori, staphylococci, streptococ-
ci, yersinia), autoreactivity and non-allergic hypersensitivity
reactions is reliable and successful. Special considerations are
indicated in the case of recurrent angioedema without whealing
and in childhood urticaria. In most cases a targeted diagnostic
program leads to the identification of potential triggering factors
and after their adequate treatment long-lasting and life quality
impairing urticaria disappears or improves within several weeks.
With regard to treatment non-sedating H1 antihistamines should
be given regularly and daily, most often increased dosage is need-
ed. Data on alternatives are insufficient but in selected cases
cyclosporin A, leukotriene receptor antagonists, or hydroxy-
chloroquine may be useful. Several questions have to be addressed
in the future and there is hope that during the next years new
therapeutic strategies will be developed to facilitate the man-
agement of this long-lasting and life-quality restricting disease.
KEY WORDS: Chronic urticaria - Autoreactivity - Infections - Heli-
cobacter pylori.
Address reprint requests to: B. Wedi, MD, Associate Professor, Depart-
ment of Dermatology and Allergology, Hannover Medical University,
Ricklinger Str. 5, D-30449 Hannover, Germany.
E-mail: Bettina_Wedi@t-online.de.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:435-43
Chronic urticaria
A review
B. WEDI, A. KAPP
Department of Dermatology and Allergology
Hannover Medical University, Hannover Germany
C hronic urticaria remains a major problem in terms
of pathogenesis, diagnostic work-up and man-
agement. During the last years several new concepts
regarding the disease have been developed. Some of
these aspects resulted in a different management of
patients with chronic urticaria whereas others still
need research activities for confirmation or clarifica-
tion of details. This review summarizes the current
concepts of classification and definition, pathogenesis
and management of chronic urticaria.
Classification and definition
The 4 main subtypes of urticaria should be clearly
differentiated: acute, chronic, physical urticaria and
a heterogeneous group of other types that do not fit
in these scheme such as urticaria pigmentosa/masto-
cytosis, urticaria vasculitis, familiar cold urticaria and
hereditary or acquired angioedema with C1-INH defi-
ciency.1-3
The cardinal clinical feature of urticaria is the occur-
rence of itchy wheals anywhere on the skin (Figure 1).
Wheals are short-lived elevated erythematous lesions
ranging from a few millimeters to several centimeters in
diameter and can become confluent. Sometimes the
435
WEDI
Figure 1.—Typical wheal and flare reaction in chronic urticaria.
wheals are paler than the surrounding skin because of
the compressing effect of the edema on the postcapillary
venules. The itching can be pricking or burning and is
usually worse in the evening or nighttime. Typically
the lesions are rubbed and not scratched, therefore,
excoriated skin is usually not a consequence of urticaria.
Acute urticaria generally disappears within 2 to 4
weeks. By definition urticaria is chronic if wheal and
flare reactions persist daily or nearly daily for at least
6 weeks. At least half of the patients suffer from con-
comitant and sometimes life threatening angioedema
most typically involving the face, lips, tongue, pharynx,
genitalia, and extremities.1 Systemic symptoms such
as fatigue, respiratory, gastrointestinal and arthralgic
symptoms are rare.
Quality of life is significantly impaired like in
patients with severe atopic dermatitis, psoriasis or
coronary artery disease.2-4
Chronic urticaria usually persists for long time since
not even 50% of patients that consulted a university der-
matology department were symptomfree within 10
years.5
It is important to know that 2 or rarely more subtypes
of urticaria can occur in the same patient such as chron-
ic urticaria and dermographism or delayed pressure
urticaria.1 In these cases urticaria is more worst.6
Pathogenesis
In chronic urticaria symptoms are the result of the
degranulation of mast cells and basophils. IgE-mediated
hypersensitivity due to exogenous allergens is generally
very rarely the cause of symptoms in chronic urticaria
436 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHRONIC URTICARIA
and there is no increased frequency of atopy. Never-
theless, a lot of direct and indirect releasing factors
may be involved. Possible mechanisms include autoim-
mune mechanisms, infectious diseases (viral, bacteri-
al, fungal, parasites), pseudoallergic mechanisms and
others such as internal diseases/malignancies.
Laboratory findings
In the last years some interesting laboratory findings
showed differences between patients with chronic
urticaria and healthy controls although a specific mark-
er is still not available.
It has been shown that the number of basophil
granulocytes is significantly decreased in chronic
urticaria 7, 8 and is negatively correlated with disease
activity.9 This led to the hypothesis that basophils may
be actively recruited, for example by chemokine
induced adhesion molecules on endothelial cells, from
the circulation to the wheals. Interestingly, not only
basophils, but also eosinophils and lymphocytes are
decreased in chronic urticaria.9 Moreover, in autoim-
mune urticaria basophil counts are significantly low-
er compare to non-autoimmune urticaria.7 Additional
data demonstrated that basophils and mast cells appear
to be activated. In this aspect, Ferrer et al. have shown
that serum IL-4 levels are significantly increased in
chronic urticaria compared to healthy controls.10 In
addition, intracellular Il-4 and IFN-γ were signifi-
cantly increased in peripheral CD4+ lymphocytes, but
not in CD8+-lymphocytes. Moreover, if urticaria was
associated with angioedema, leukotriene levels
appeared to be increased.10 Others showed that serum
IL-2 receptor levels and tryptase are significantly high-
er in chronic urticaria pointing to T-cell and mast cell
activation.11 In the subgroup with positive autologous
serum skin test (ASST) and FcεRIα autoantibodies
tryptase levels were highest. Moreover, CD40L expres-
sion on activated T-cells and bcl2 expression in acti-
vated T- and B-lymphocytes were increased in severe
chronic urticaria.12
In contrast, ASST positive patients with chronic
urticaria did not demonstrate increased stem cell fac-
tor (SCF) serum levels compared to controls.13
Diagnosis
The diagnosis of chronic urticaria is based upon a
detailed history considering potential triggering factors,
Agosto 2005
CHRONIC URTICARIA
a physical examination including a test for dermo-
graphism, laboratory investigations and, if needed,
additional specific procedures.
According to an international consensus 14 patient
history should include at least the following items
to define the subtype of urticaria and to identify
potential related causes: 1) first manifestation, fre-
quency, duration and daily variability; 2) border, size
and distribution of wheals; 3) association with
angioedema; 4) subjective symptoms like pruritus,
pain; 5) family history for urticaria, atopy; 6) prior or
current allergies, infections, internal diseases or oth-
er related causes; 7) exacerbation by physical fac-
tors or exercise; 8) drug intake (particularly NSAID
and angiotensin converting enzyme inhibitors); 9)
relation to food/food additives; 10) nicotine; 11)
occupation, hobbies; 12) relation to weekends, hol-
idays, holidays abroad; 13) surgical implants; 14)
reactions to insect stings; 15) association with men-
struation; 16) response to prior treatment; 17) stress
and 18) quality of life impairment (Zuberbier T, Bind-
slev-Jensen C, Canonica W, Grattan CEH, Greaves
MW, Henz BM et al. EAACI/GA_LEN guideline:
definition, classification, and diagnosis of urticaria.
Submitted to Allergy).
Patient diaries are very helpful to become aware of
the fluctuating intensity of the disease.
Physical examination should determine number and
size of wheals, angioedema and should include der-
mographism.
If physical triggering is suspected, appropriate and
standardized physical tests should be performed such
as pressure test with defined weights,15-18 but physical
urticaria will not be discussed in this review.
Activity of the disease should be evaluated using a
standardized score (Table I).19 This is of particular
importance for treatment trials. In addition, impair-
ment of daily life, occupational and psychosocial
aspects should be considered. Standardized instru-
ments for dermatologic diseases have been used,3, 4
however, instruments adapted for chronic urticaria
may be more useful.2
If single wheals persist for longer than 24 h biopsies
should be taken to exclude vasculitis by histology and
immunofluorescence that may indicate systemic dis-
ease like lupus erythematodes.20
Routinely only differential count and general
inflammatory parameters such as C-reactive protein
(CRP) should be determined. An international con-
sensus established by hand voting in the year 2001
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
WEDI
TABLE I.—Urticaria-Score according to Zuberbier T, Bindslev-Jen-
sen C, Canonica W, Grattan CEH, Greaves MW, Henz BM et al.
EAACI/GA_LEN guideline: definition, classification, and dia-
gnosis of urticaria. Submitted to Allergy and Zuberbier et al.19
Score Wheals Pruritus
0 None None
1 Mild Mild
(<20 wheals/24 h)
2 Moderate Moderate
(21- 50 wheals/24 h)
3 Severe Severe
(>50 wheals/24 h
or large confluent
areas of wheals)
Sum of score: (0-6).
numbered the following parameters as useful in
chronic urticaria: serology for helicobacter, gas-
troscopy, anti-streptolysin titre, serology for hepati-
tis, pseudoallergen-low diet for 3 weeks, ASST, anti-
nucleare antibodies, thyroid antibodies, oral provo-
cation tests, specific IgE, ova/parasites and other
investigations.19, 21
However, it should be noted, that this consens was
established by simple hand voting and regionally, the
optimal diagnostic schedule may differ, e.g. in Ger-
many investigation for hepatitis and parasites is not
very useful.
Our research results during the last years revealed a
reliable and successful diagnostic program based on the
following 3 parameters:1, 22-27
1) infections;
2) autoreactivity;
3) non-allergic hypersensivity reactions (pseudoal-
lergies).
Therefore, the determination of the parameters pre-
sented is used routinely in our department (Figure 2).
Infections
Recently, we reviewed the published literature
regarding “chronic urticaria and infections.” 23 It was
found that studies investigating this topic were difficult
to compare, evidence-based criteria were not applic-
able and a meta-analysis was impossible. Most stud-
ies did not consider the multiplicity of triggering fac-
tors. Further complicating factors are for example geo-
graphic variations in the frequency of infections (e.g.
for parasitosis).
437
WEDI CHRONIC URTICARIA

Non-allergic
Infections Autoreactivity hypersensitivity

Investigation of: - ASST5 - Avoidance of

- Helicobacter 1 - cellular activation aspirin and other
- Streptococci 2 (BHR, BasoAT, CAST)6 NSAID
- Staphylococci 3 - thyroid autoAbs7 - pseudoallergen-low
- Yersinia4 - (antinuclear Abs) diet8

Detailed urticaria history, dermographism

(perhaps physical tests)
differential
differentialcount,
count,CRP9,
CRP9, C1-INH function
Cl-INH function

Figure 2.—Recommended diagnostic schedule in chronic urticaria. In children, serology for Epstein-Barr-virus and cytomegalovirus should be includ-
ed; 1) urea breath test or gastroscopy with biopsies; 2) antiDNAseB-, antistreptolysin-titre; 3) antistaphylolysin-titre; 4) Yersinia-IgA, -IgG and
immunoblot; 5) autologous serum skin test; 6) BHR = basophil histamine release, BasoAT = basophil activation test (CD63 expression), CAST = cel-
lular antigen stimulation test (production of sulfidoleukotrienes); 7) basal TSH, microsomal Abs, thyroglobulin Abs, TSH receptor Abs; 8) in select-
ed patients for at least 3 weeks; 9) general parameter for inflammation/malignancy.

From the literature it can be seen that the preva-
lence of infections, either bacterial, viral, parasitic or
fungal appears not to differ compared to the general
population (for yersinia this has to be demonstrated).
However, there is a very large amount of reports
demonstrating benefit after eradication of infectious
processes and it is hardly to believe that all these are
due to spontaneous remissions. Nevertheless, avail-
able studies have common flaws and cannot be
reviewed as a meta-analysis or according to evidence-
based medicine rules. Best evidence exists for Heli-
cobacter pylori infection.23
Systematically reviewing existing studies address-
ing the effect of antibiotic therapy for chronic urticaria
patients infected with Helicobacter pylori revealed
that resolution of urticaria was more likely when
antibiotic therapy was successful in eradication of
Helicobacter.28 Ten studies met the inclusion criteria
and when data from these studies were combined,
eradication of Helicobacter pylori was both quanti-
tatively and statistically associated with remission
of urticaria, with an odds ratio of 2.9 (95% confi-
dence interval 1.4-6.8; P= 0.005). The authors con-
cluded that clinicians, after considering other caus-
es of urticaria, should constitute 1) testing for Heli-
cobacter pylori; 2) treating with appropriate antibi-
otics if Helicobacter pylori is present; and 3) con-
firming successful eradication of infection.28 We are
recommending this approach for several years.22-24, 26,
27, 29, 30
However, Helicobacter is not the only triggering
factor. Other persistent, chronic, in most cases sub-
clinical infections, for example, with streptococci,
staphylococci and yersinia can also be found.23
In the case of an identified infection, targeted erad-
ication should be performed and it should be careful-
ly looked whether eradication was successful.
It should be borne in mind that often urticaria does
not disappear until all triggering factors have been
carefully addressed, e.g. Helicobacter pylori asso-
ciated gastritis, persistent yersiniosis, elevated anti-
streptococcal titres with chronic sinusitis plus posi-
tive ASST plus exacerbation through regular aspirin
intake.

438 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

CHRONIC URTICARIA WEDI

Autoreactivity
Details of the autoimmune pathogenesis of chron-
ic urticaria have been recently reviewed.31, 32
Evidence for autoreactive mechanisms in chronic
urticaria is provided by a positive ASST (Figure 3).
However, the clinical relevance is far from being clear
because the presence of functional autoantibodies
against FcεRI and/or IgE itself does not correlate with
a positive ASST. Moreover, ASST can be still positive
although urticarial symptomatology disappeared.
Therefore, several authors recommed a functional
cellular activity assay such as basophil histamine
release for confirmation. In addition, the determina-
tion of sulfidoleukotriene de novo production in leuko-
cyte suspensions (cellular antigen stimulation test,
CAST) and flow cytometric CD63 surface expres-
sion on basophils are possible 33-35 although hista-
mine release appears to be more specific. Nevertheless, Figure 3.—Positive autologous serum skin test in chronic urticaria, read-
ing after 30 min.
these functional assays are difficult to perform and
to interpret due to their dependence on the releasabil-
ity of the donor cells used. Up to the present, the direct chronic urticaria.37 After exclusion of other causes, in
measurement of the autoantibodies is not available single cases a standardized pseudoallergen-low diet
for routine purposes since a satisfactory ELISA has not might be indicated for at least 3 weeks.38 However,
been developed. controlled provocation test often are not able to iden-
Patients with positive ASST more often have thyroid tify causal substances.39
autoantibodies. Thus, their determination is recom-
mended at least in women with chronic urticaria. It is
debated whether therapeutic implications are given SPECIAL DIAGNOSTIC ISSUE: CHILDHOOD CHRONIC
when—as it is most often the case—thyroid function URTICARIA
is normal. In selected cases assessment of antinuclear Recently, it has been shown that 20-30% of chil-
antibodies may be advisable to exclude systemic lupus dren with acute urticaria progressed into chronic
erythematosus. urticaria. In almost all (91%) acute urticaria was con-
sidered to be induced by acute infection.40 Similarly,
Non-allergic hypersensitivity (pseudoallergic reac- assocation with infections also plays a major role in
tions) chronic urticaria.23 Persistent chronic often bacterial
infections (e.g. with streptococci, staphylococci, but
NSAID, particularly aspirin, are a common exac-
also with Helicobacter pylori and yersinia) are most
erbating factor in chronic urticaria and should be avoid-
common in chronic urticaria.40, 41 Furthermore, posi-
ed. However, as single causal factor they are rare. If sus-
tive ASST indicating autoreactivity can also be found
pected a placebo-controlled provocation test should
in about 30% of children with chronic urticaria.41, 42 In
be envisaged.
children and young adults serology for Epstein-Barr
In this aspect, recently an interesting publication
virus and cytomegalovirus should be included in the
demonstrated that 33% of patients with acute urticaria
diagnostic program.
due to NSAID intake developed chronic urticaria 1 to
10 years later in contrast to 1% of an atopic control pop-
RECURRENT ANGIOEDEMA
ulation.36 The authors concluded that NSAID intoler-
ance may predispose to later development of chronic Recurrent angioedema without wheals may be
urticaria. caused by hereditary or acquired C1-esterase inhibitor
In addition, food additives are suggested to trigger deficiency or dysfunction, drugs (particularly

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 439

WEDI
angiotensin converting enzyme inhibitiors, angiotensin
II receptor antagonists),43, 44 and perhaps also persis-
tent infections (e.g. Helicobacter pylori-associated
gastritis, yersiniosis).45-47
If angioedema are described principally at least the
functional activity of the C1-esterase inhibitor should
be assessed to exclude deficiency or dysfunction.48
ACE-inhibitors and perhaps also AT II receptor antag-
onists should be avoided. It has to be considered that
angioedema which often are located in the oropharynx
occur within several weeks after initiation of treat-
ment but can also occur first after several years of
intake.49
Treatment
Henderson et al.50 addressed the question what
agents are commonly used to treat urticaria in 9.2 mil-
lion visits in office-based practices between 1990 and
1997 in the USA. They found that H1-antihistamines
were prescribed in 56% of consultations, systemic
corticosteroids in 14% and other drugs in 12%. Inter-
estingly, allergists were the least likely to prescribe
corticosteroids whereas internists were the most like-
ly.51 Looking at alternatives, H2-antihistamines, β-
agonists, doxepin, nifedipine and, interestingly enough
methotrexate were prescribed. These drugs are not
prescribed in our clinic.
An italian study described prescription of antihist-
amines in 351 patients with chronic urticaria, with a
good response in only 40%.51 In this study 221 patients
were treated with other drugs such as steroids,
cyclosporine A, cromolyn, LT receptor antagonists
and adrenaline.
It is undeniable that we do not have a standard treat-
ment of chronic urticaria.
To make preparations for the recent international
urticaria consensus conference in Berlin 2004 we per-
formed a systematic review of randomised controlled
trials (RCTs) in chronic urticaria (Zuberbier T, Bind-
slev-Jensen C, Canonica W, Grattan CEH, Greaves
MW, Henz BM et al. EAACI/GA_LEN guideline:
definition, classification, and diagnosis of urticaria.
Submitted to Allergy). A literature search using MED-
LINE and EMBASE, in part also by hand-searching
was done. In chronic urticaria RCTs demonstrated
ineffective treatment with sedating H1-antihistamines
plus cimetidine or terbutaline and also with tranex-
amic acid or cromolyn. However, these were single
440 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHRONIC URTICARIA
studies with several flaws, so that their level of evidence
is very low.
In MEDLINE and EMBASE there is no study
addressing the efficacy of corticosteroids in chronic
urticaria although they are widely used. Sometimes
corticosteroids are needed for example to achieve rapid
control to cover social or occupational events but it is
consent that prolonged daily treatment should be clear-
ly avoided due to severe side effects.
In the treatment of chronic urticaria best evidence
exists for non-sedating H1-antihistamines such as aze-
lastine, cetirizine, desloratadine, ebastine, fexofenadine,
levocetirizine, loratadine and mizolastine (alphabeti-
cal order). The quality of these RCTs is high and tak-
en together non-sedating H1 antihistamines can be
recommended with highest Grade A. However, we
should bear in mind that most of these studies includ-
ed only patients with mild to moderate urticaria. High
quality RCTs that compared non-sedating H1-anti-
histamines are not available and from the existing evi-
dence it appears that differences are rather small if
ever existing. However, from clinical experience it is
well known that non-responders with one non-sedat-
ing antihistamines may respond favourably to anoth-
er. Urticaria experts often use increased dosage and
they are clearly needed in most patients with chronic
urticaria. However, RCTs addressing this point in
chronic urticaria are missing.
Several drugs, such as cyclosporin A, LT receptor
antagonists and stanazolol have been combined with
non-sedating H1 antihistamines. However, from an
evidence based view the grade of recommendation is
low. The situation is even more bad looking at other
alternatives such as doxepin, oxatomide, nifedipin,
montelukast, and warfarin. For several drugs such as
corticosteroids, dapsone, sulfasalazine, methotrexate,
interferon, for plasmapheresis and intravenous
immunoglobulins (IVIGs) we do not have RCTs. Evi-
dence is based on case series, uncontrolled trials or
expert opinion.
Nevertheless, cyclosporin A may be a valuable alter-
native in severely affected patients although potential
side effects have to be considered. Grattan et al. 52
investigated a dose of 4 mg/kg per day in combination
with a doubled dose of cetirizine in ASST+ patients.
An unblinded study by Baskan et al. 53 also included
ASST+ patients and used the same dose but their aim
was to compare 4 weeks versus 12 weeks treatment.
The unblinded and uncontrolled study by Toubi 54
Agosto 2005
CHRONIC URTICARIA
investigated the effect of low-dose cyclosporine. They
stated that efficacy of cyclosporine was independent of
ASST reactivity. However, all these studies included
very few patients.
Available data for leukotriene receptor antagonists
are inconsistent and difficult to compare.55-60 These
drugs may be of benefit in subgroups of patients with
chronic urticaria that have a positive ASST and/or suf-
fer from ASA and/or food intolerance. It may be also
worthwhile to compare patients with and without
angioedema.
In our hands hydroxychloroquine is effective in
patients with autoimmune urticaria. Several years ago
in vitro we found significant inhibitory effects of
chloroquine on serum activity of ASST positive
patients.35 Chloroquine inhibited histamine release,
basophil CD63 expression and leukotriene de novo
production that was induced by ASST positive chron-
ic urticaria sera.
This benefit may be also confirmed by an australian
open label study 61 and a recent randomised placebo-
controlled controlled trial 62 although both studies
have several flaws.
Taken together non-sedating H1 antihistamines
remain the key treatment of chronic urticaria although
these drugs are insufficient in several patients even in
increased dosage. Thus, more and well designed RCTs
are clearly needed to recommend or refuse potential
alternative drugs. Although difficult to perform these
RCTs should include best characterised patients and
should try to minimise bias by carefully addressing
critical points of internal validity.
Several questions have to be addressed in the future
and there is hope that during the next years new ther-
apeutic strategies will be developed to facilitate the
management of this long-lasting and life-quality
restricting disease.
Riassunto
Orticaria cronica: una review
L’orticaria cronica rimane una delle patologie maggior-
mente problematiche in tema di eziopatogenesi, work-up dia-
gnostico e trattamento. Negli ultimi anni sono stati portati
avanti nuovi concetti riguardanti questa patologia. Alcuni di
questi aspetti hanno portato a modificare il management del-
la malattia, mentre i risultati di altri studi di ricerca devono
ancora essere confermati e validati. I sintomi dell’orticaria deri-
vano, da un punto di vista fisiopatologico, dalla degranulazione
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
WEDI
delle mast cell e dei basofili attraverso possibili meccanismi
di tipo autoimmune, infettivo, pseudoallergico, e altri (con-
nessi a malattie internistiche e a neoplasie).
Un’accurata anamnesi della malattia gioca un ruolo fon-
damentale nel programma diagnostico. Ulteriori indagini
sono indicate in base al sottotipo di patologia. Nell’orticaria
acuta, il classico iter diagnostico non viene generalmente
raccomandato, mentre nella forma cronica si consiglia viva-
mente di eseguire un programma mirato alla valutazione di
infezioni (in particolar modo da Helicobacter pilori, stafi-
lococchi, streptococchi, Yersinia), test di autoreattività e test
di ipersensibilità non allergologici. Particolare considera-
zione deve essere posta all’angioedema e all’orticaria del-
l’infanzia. Nella maggior parte dei casi l’iter diagnostico
porta alla identificazione dei fattori trigger della malattia; è
stato visto che il trattamento a lungo termine dei fattori trig-
ger migliora la qualità di vita dei pazienti e le manifestazio-
ni cliniche della malattia possono scomparire o manifestar-
si nuovamente, ma dopo parecchi anni.
Vengono consigliati gli antistaminici H1 (pochi effetti
collaterali) che devono essere somministrati quotidiana-
mente e regolarmente; a volte è necessario aumentare la
posologia. Altri trattamenti sono risultati non soddisfacenti
anche se la ciclosprina A, gli antagonisti leucotrienici e l’i-
droclorochina possono essere efficaci in casi selezionati.
Ci sono ancora molte domande a cui rispondere in futuro
con la speranza che nei prossimi anni le strategie terapeuti-
che possano facilitare il management e la qualità di vita di
questa patologia cronica e limitante le attività quotidiane.
Parole chiave: Orticaria cronica - Autoreattività - Infezioni
- Helicobacter pylori.
References
1. Wedi B. Acute, chronic or physical urticaria. What causes the hives?
MMW Fortschr Med 2002;144:28-32. German.
2. Baiardini I, Giardini A, Pasquali M, Dignetti P, Guerra L, Specchia C
et al. Quality of life and patients’ satisfaction in chronic urticaria and
respiratory allergy. Allergy 2003;58:621-3.
3. Poon E, Seed PT, Greaves MW, Kobza-Black A. The extent and nature
of disability in different urticarial conditions. Br J Dermatol
1999;140:667-71.
4. O’Donnell BF, Lawlor F, Simpson J, Morgan M, Greaves MW. The
impact of chronic urticaria on the quality of life. Br J Dermatol
1997;136:197-201.
5. Van der Valk PG, Moret G, Kiemeney LA. The natural history of
chronic urticaria and angioedema in patients visiting a tertiary refer-
ral centre. Br J Dermatol 2002;146:110-3.
6. Kozel MMA, Mekkes JR, Bossuyt PM, Bos JD. Natural course of
physical and chronic urticaria and angioedema in 220 patients. J Am
Acad Dermatol 2001;45:387-91.
7. Bakos N, Hillander M. Comparison of chronic autoimmune urticaria
with chronic idiopathic urticaria. Int J Dermatol 2003;42:613-5.
8. Grattan CE, Walpole D, Francis DM, Niimi N, Dootson G, Edler S et
al. Flow cytometric analysis of basophil numbers in chronic urticaria:
basopenia is related to serum histamine releasing activity. Clin Exp
Allergy 1997;27:1417-24.
9. Grattan CE, Dawn G, Gibbs S, Francis DM. Blood basophil num-
bers in chronic ordinary urticaria and healthy controls: diurnal vari-
441
WEDI
ation, influence of loratadine and prednisolone and relationship to
disease activity. Clin Exp Allergy 2003;33:337-41.
10. Ferrer M, Luquin E, Sanchez-Ibarrola A, Moreno C, Sanz ML, Kaplan
AP. Secretion of cytokines, histamine and leukotrienes in chronic
urticaria. Int Arch Allergy Immunol 2002;129:254-60.
11. Hidvegi B, Nagy E, Szabo T, Temesvari E, Marschalko M, Karpati S
et al. Correlation between T-cell and mast cell activity in patients
with chronic urticaria. Int Arch Allergy Immunol 2003;132:177-82.
12. Toubi E, Adir-Shani A, Kessel A, Shmuel Z, Sabo E, Hacham H.
Immune abberations in B and T lymphocytes derived from chronic
urticaria patients. J Clin Immunol 2000;20:371-8.
13. Asero R, Tedeschi A, Lorini M, Gerosa M, Meroni P, Riboldi P. Cir-
culating stem cell factor in patients with chronic idiopathic urticaria.
Ann Allergy Asthma Immunol 2003;91:79-81.
14. Zuberbier T, Greaves MW, Juhlin L, Merk H, Stingl G, Henz BM. Man-
agement of urticaria: a consensus report. J Investig Dermatol Symp
Proc 2001;6:128-31.
15. Fleischer M, Grabbe J. Physical urticaria. Hautarzt 2004;55:344-9.
16. Muller BA. A comprehensive review of physical urticaria. Compr
Ther 2002;28:214-21.
17. Kontou-Fili K, Borici-Mazi R, Kapp A, Matjevic LJ, Mitchel FB.
Physical urticaria: classification and diagnostic guidelines. An EAACI
position paper. Allergy 1997;52:504-13.
18. Black AK, Lawlor F, Greaves MW. Consensus meeting on the defin-
ition of physical urticarias and urticarial vasculitis. Clin Exp Derma-
tol 1996;21:424-6.
19. Zuberbier T, Aberer W, Grabbe J, Hartmann K, Merk H, Ollert M et
al. Diagnostik und Therapie der Urtikaria. JDDG 2003;1:655-64.
20. Mehregan DR, Hail MJ, Gibson LE. Urticarial vasculitis: a histopatho-
logic and clinical review of 72 cases. J Am Acad Dermatol
1992;26:441-8.
21. Zuberbier T. Urticaria. Allergy 2003;58:1224-34.
22. Wedi B, Kapp A. Helicobacter pylori infection in skin diseases: a
critical appraisal. Am J Clin Dermatol 2002;3:273-82.
23. Wedi B, Raap U, Kapp A. Chronic urticaria and infections. Curr Opin
Allergy Clin Immunol 2004;4:387-96.
24. Wedi B, Liekenbröcker T, Kapp A. Persistent bacterial infection and
serum activity in chronic urticaria - role of molecular mimikry? Aller-
gologie 2001;24:480-90. German.
25. Wedi B, Novacovic V, Koerner M, Kapp A. Chronic urticaria is fre-
quently triggered by focal, particularly gastrointestinal infection:
analysis of 325 cases. J Investig Dermatol Symp Proc 2001;6:163.
26. Wedi B, Wagner S, Werfel T, Körner M, Manns MP, Kapp A.
Rationelles diagnostisches Vorgehen bei der chronischen Urtikaria. Z
Hautkr H + G 2000;75:78-84.
27. Wedi B, Wagner S, Werfel T, Manns MP, Kapp A. Prevalence of Heli-
cobacter pylori associated gastritis in chronic urticaria. Int Arch Aller-
gy Immunol 1998;116:288-94.
28. Federman DG, Kirsner RS, Moriarty JP, Concato J. The effect of
antibiotic therapy for patients infected with Helicobacter pylori who
have chronic urticaria. J Am Acad Dermatol 2003;49:861-4.
29. Raap U, Liekenbrocker T, Wieczorek D, Kapp A, Wedi B. New ther-
apeutic strategies for the different subtypes of urticaria. Hautarzt
2004;55:361-6. German.
30. Wedi B, Kapp A. Helicobacter pylori infection and skin diseases. J
Physiol Pharmacol 1999;50:753-76.
31. Kaplan AP. Chronic urticaria: pathogenesis and treatment. J Allergy
Clin Immunol 2004;114:465-74.
32. Grattan CE, Sabroe RA, Greaves MW. Chronic urticaria. J Am Acad
Dermatol 2002;46:645-57.
33. Kikuchi Y, Kaplan AP. Mechanisms of autoimmune activation of basophils
in chronic urticaria. J Allergy Clin Immunol 2001;107:1056-62.
34. Sabroe RA, Francis DM, Barr RM, Black AK, Greaves MW. Anti-
Fc(epsilon)RI auto antibodies and basophil histamine releasability
in chronic idiopathic urticaria. J Allergy Clin Immunol 1998;102 (4
Pt 1):651-8.
35. Wedi B, Novacovic V, Koerner M, Kapp A. Chronic urticaria serum
442 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
CHRONIC URTICARIA
induces histamine release, leukotriene production and basophil CD63
surface expression – inhibitory effects of anti-inflammatory drugs. J
Allergy Clin Immunol 2000;105:552-60.
36. Asero R. Intolerance to nonsteroidal anti-inflammatory drugs might
precede by years the onset of chronic urticaria. J Allergy Clin Immunol
2003;111:1095-18.
37. Zuberbier T. Pseudoallergens and chronic urticaria. Allergologie
2001;24:457-62. German.
38. Zuberbier T, Chantraine-Hess S, Hartmann K, Czarnetzki BM.
Pseudoallergen-free diet in the treatment of chronic urticaria. A
prospective study. Acta Derm Venereol 1995;75:484-7.
39. Werfel T, Wedi B, Kleine-Tebbe J, Niggemann B, Saloga J, Sen-
nekamp J et al. Vorgehen bei Verdacht auf eine pseudo-allergische
Reaktion durch Nahrungsmittelinhaltsstoffe. Allergo J 1999;8:
135-141.
40. Sackesen C, Sekerel BE, Orhan F, Kocabas CN, Tuncer A, Adalioglu
G. The etiology of different forms of urticaria in childhood. Pediatr Der-
matol 2004;21:102-8.
41. Wieczorek D, Raap U, Liekenbrocker T, Kapp A, Wedi B. Chronic
urticaria in childhood. Hautarzt 2004;55:357-60. German.
42. Brunetti L, Francavilla R, Miniello VL, Platzer MH, Rizzi D, Lospal-
luti ML et al. High prevalence of autoimmune urticaria in children with
chronic urticaria. J Allergy Clin Immunol 2004;114:922-7.
43. Bowen T, Cicardi M, Farkas H, Bork K, Kreuz W, Zingale L et al.
Canadian 2003 International Consensus Algorithm for the Diagnosis,
Therapy, and Management of Hereditary Angioedema. J Allergy Clin
Immunol 2004;114:629-37.
44. Bork K, Hardt J, Schicketanz KH, Ressel N. Clinical studies of sud-
den upper airway obstruction in patients with hereditary angioedema
due to C1 esterase inhibitor deficiency. Arch Intern Med 2003;
163:1229-35.
45. Heymann WR. Acquired angioedema. J Am Acad Dermatol
1997;36:611-5.
46. Sabroe RA, Kobza Black A. Angiotensin-converting enzyme (ACE)
inhibitors and angio-oedema. Br J Dermatol 1997;136:153-8.
47. Varvarovska J, Sykora J, Stozicky F, Chytra I. Acquired angioedema
and Helicobacter pylori infection in a child. Eur J Pediatr 2003;162:
707-9.
48. Charlesworth EN. Differential diagnosis of angioedema. Allergy
Asthma Proc 2002;23:337-9.
49. Wedi B, Raap U, Wieczorek D, Kapp A. Urticaria - an update. Aller-
gologie 2004;27:435-43. German.
50. Henderson RL, Fleischer AB, Feldman SR. Allergists and dermatol-
ogists have far more expertise in caring for patients with urticaria
than other specialists. J Am Acad Dermatol 2000;43:1084-91.
51. Nettis E, Pannofino A, D’Aprile C, Ferrannini A, Tursi A. Clinical and
aetiological aspects in urticaria and angio-oedema. Br J Dermatol
2003;148:501-6.
52. Grattan CE, O’Donnell BF, Francis DM, Niimi N, Barlow RJ, Seed
PT et al. Randomized double-blind study of cyclosporin in chronic
‘idiopathic’ urticaria. Br J Dermatol 2000;143:365-72.
53. Baskan EB, Tunali S, Turker T, Saricaoglu H. Comparison of short-
and long-term cyclosporine A therapy in chronic idiopathic urticaria.
J Dermatolog Treat 2004;15:164-8.
54. Toubi E, Blant A, Kessel A, Golan TD. Low-dose cyclosporin A in the
treatment of severe chronic idiopathic urticaria. Allergy 1997;52:
312-6.
55. Bagenstose SE, Levin L, Bernstein JA. The addition of zafirlukast to
cetirizine improves the treatment of chronic urticaria in patients with
positive autologous serum skin test results. J Allergy Clin Immunol
2004;113:134-40.
56. Di Lorenzo G, Pacor ML, Mansueto P, Pellitteri ME, Lo Bianco C, Dit-
ta V et al. Randomized placebo-controlled trial comparing deslo-
ratadine and montelukast in monotherapy and desloratadine plus
montelukast in combined therapy for chronic idiopathic urticaria. J
Allergy Clin Immunol 2004;114:619-25.
57. Nettis E, Colanardi MC, Paradiso MT, Ferrannini A. Desloratadine in
Agosto 2005
CHRONIC URTICARIA
combination with montelukast in the treatment of chronic urticaria: a
randomized, double-blind, placebo-controlled study. Clin Exp Aller-
gy 2004;34:1401-7.
58. Erbagci Z. The leukotriene receptor antagonist montelukast in the
treatment of chronic idiopathic urticaria: a single-blind, placebo-con-
trolled, crossover clinical study. J Allergy Clin Immunol 2002;110:
484-8.
59. Reimers A, Pichler C, Helbling A, Pichler WJ, Yawalkar N. Zafir-
lukast has no beneficial effects in the treatment of chronic urticaria.
Clin Exp Allergy 2002;32:1763-8.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
WEDI
60. Pacor ML, Di Lorenzo G, Corrocher R. Efficacy of leukotriene recep-
tor antagonist in chronic urticaria. A double-blind, placebo-controlled
comparison of treatment with montelukast and cetirizine in patients
with chronic urticaria with intolerance to food additive and/or acetyl-
salicylic acid. Clin Exp Allergy 2001;31:1607-14.
61. Baumgart KW, Mullins R. Use of hydroxychloroquine in refractory
urticaria. J Allergy Clin Immunol 2000;105:795-6.
62. Reeves GEM, Boyle MJ, Bonfield J, Dobson P. Impact of hydroxy-
chloroquine therapy on chronic urticaria: chronic autoimmune urticaria
study and evaluation. Int Med J 2004;34:182-6.
443
 CLINICAL CASES
Granular parakeratosis is an acquired disorder of keratiniza-
tion of unknown origin affecting mainly flexures of adults. A 48-
year-old woman presented a 1-year history of keratotic papules
on the infra- and submammary regions. A skin biopsy revealed
granular parakeratosis confined to epidermal foci. The patient's
eruption resolved completely with topical steroid therapy.
Etiopathogenesis and differential diagnosis are discussed.
KEY WORDS: Granular parakeratosis - Keratinization disorder -
Skin.
G ranular parakeratosis (GP) is an histologic phe-
nomenon with distinct clinical characteristics.
It is normally localized in intertriginous areas and it
was first described in 1991 by Northcutt et al. as
axillary GP. 1 Since then, the disorder has been
described in other intertriginous areas such as the
inguinal region, inter- and submammary region, the
vulva and perianal region, the trunk and the knee.2-
7 Clinically, the GP presents with unilateral or bilat-
eral erythematous hyperkeratotic papules and plaques
that are often pruritic. Women beyond the 5th decade
are mainly affected,3 although the disorder may also
occur in children.8-11 The course is chronic with poor
response to therapy and tendency to spontaneous
resolution.1, 4, 6
Received: September 29, 2004.
Accepted for publication: July 1, 2005.
Address reprint requests to: Dott. C. Tomasini, Clinica Dermatologica
II, Via Cherasco 23, 10126 Torino.
E-mail: ctomasini@molinette.piemonte.it
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
G ITAL DERMATOL VENEREOL 2005;140:445-9
Granular parakeratosis
C. TOMASINI, Z. SEIA
Division of Dermatology, University of Turin, Turin, Italy
Herein, we report and discuss a case of GP in inter-
and submammary regions.
Case report
We present a case of a 48-year-old healthy woman
with a 1-year history of dermatosis in the infra- and
submammary regions. Symptoms included itching
and burning sensation. On clinical examination there
were numerous, small, erythematous, keratotic
papules disseminated in the submammary region
and in the inframammary folds (Figure 1). The
lesions were friable and many of them could be
removed by scraping. The patient referred that sim-
ilar lesions had happened in the axillae and inguinal
folds, and subsequently had spontaneously regressed.
Family history was negative for blistering or kera-
tinization diseases, psoriasis, or other cutaneous dis-
orders. She was using no new personal hygiene prod-
ucts. A punch biopsy specimen was obtained from
the right submammary fold. The histological exam-
ination revealed discrete foci of psoriasiform epi-
dermal hyperplasia with minimal spongiosis, preser-
vation of granular layer, and a thick parakeratotic
cornified layer with retention of keratohyaline gran-
ules (Figures 2, 3). A focal collection of neutrophils
445
TOMASINI GRANULAR PARAKERATOSIS

Figure 1.—Reddish hyperkeratotic papules in the inter-and submamma- Figure 2.—The horny layer is parakeratotic with granular appearance.
ry regions.

in the cornified layer was detected. In the superficial

dermis there were ectatic vessels and a superficial
perivascular infiltrate of lymphocytes. Periodic acid-
Schiff stain was negative for fungi. A diagnosis of GP
was established. A swab was also taken for micro-
biologic examination: cultures were negative for
bacteria, dermatophytes and Candida. Treatment
with topical fluconazole cream was ineffective. Top-
ical betamethasone dipropionate ointment was then
introduced twice daily with resolution of the rash
within 30 days.

Discussion and conclusions
GP is an acquired disorder of keratinization that
affects flexures, initially described in 1991 in 4 patients
who had an erythematous eruption in the axillae.1
Since then, to the best of our knowledge, 42 addition-
al cases - including the present case - have been report-
ed. Thirteen of these cases occurred in sites other than
the axillae, including the groin, infra- e submammary
areas, perianal area, trunk and the knee.2-7 Eight cas-
es occurred in infants with ages at presentation rang-
ing from 9 to 22 months.8-11 Interestingly, in this small
series of cases, 2 clinical patterns could be identified:
bilateral linear plaques in the inguinal folds and ery-
thematous, geometric plaques underlying pressure
points from the diaper.8
The etiology of GP was initially thought to be an
unusual contact reaction to a deodorant or antiper-
Figure 3.—The corneocytes are nucleated and replete with keratohyaline
granules.
spirant,1, 4, 12, 13 but this could not explain the fre-
quent unilateral involvement, 4 the inconsistent
response to discontinuation of the suspected irritant,
the negative patch tests, and the localization in non-
axillary intertriginous folds.1 The absence of spon-
giosis on histology also militates against contact
dermatitis as an etiologic factor. Conversely, the ten-
dency to localize to intertriginous or occluded regions
would implicate that heat, moisture, friction, and
obesity are contributing factors.1, 4, 12, 13 Furthermore,
the well-demarcated, geometric distribution under-
lying pressure points of disposable diapers would
also suggest that diapers play a role in GP of the
infancy.8-11

446 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005

GRANULAR PARAKERATOSIS
A possible role of fungal/yeast infections in the
development of GP has been suggested by finding of
positive cultures for Candida albicans from lesions
of GP and histologic observation of hyphal elements
within granular parakeratotic corneocytes.12-14
In adults, the putative pathophysiological mecha-
nism of GP involves a defective pathway of kera-
tinization in which profilaggrin is unable to form
filaggrin, the protein component of keratohyaline
granule. 12 The function of filaggrin has not been
completely elucidated, but in the cornified cells it
is thought to serve as the matrix protein that embeds
and promotes the aggregation and disulphide bind-
ing of keratin filaments producing the keratin pattern
structure of the lower cornified cells. The conver-
sion of profilaggrin into its monomeric subunits must
be carefully controlled to prevent premature aggre-
gation. Patients with GP exhibit a lack of degradation
of keratohyaline granules and aggregation of ker-
atin filaments.2
The clinical differential diagnosis of GP involving
the flexures includes a wide spectrum of disorders
characterized by erythemato-keratotic papules, such
as Hailey-Hailey disease, Darier disease, pemphi-
gus vegetans, plane warts, acanthosis nigricans, pso-
riasis, tinea infection, seborrheic dermatitis, inter-
trigo, candidiasis, Langerhans cell histiocytosis,
nummular eczema and seborrheic keratosis in early
stage. The diagnosis is confirmed by examination
of a biopsy specimen showing characteristic find-
ings of parakeratotic corneocytes containing kera-
tohyaline granules situated usually over a hyper-
plastic epidermis, although occasionally granular
parakeratotic changes are confined only to follicular
infundibulum.7
Interestingly, in newborns GP shows striking clin-
ical and pathological resemblance to a disorder first-
ly described in German literature in 1975 by Gart-
mann et al. as pomaden Kruste 15 and later reported in
English literature under the name "pigmented and
hyperkeratotic napkin dermatitis",16 attributed to
overtreatment of the groins of babies with ointments
and oils. It is likely that these diseases actually repre-
sent examples of GP.
In conclusion, GP represents a pattern of altered
keratinization in which multiple etiologies may be
operative, one of them is microbic, and this might also
explain why GA has a propensity for intertriginous
sites.
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TOMASINI
Approaches to treatment of GP have been largely
empirical. Although many cases have a self resolu-
tion,4-12, 17 a variable response has been observed
with oral and topical corticosteroid, antimicrobials,
antifungals, and keratolytics. 18-21 Whilst topical
retinoids do not appear to be effective, the use of vit-
amin D analogues was noted to be effective in some
cases, supporting the theory of abnormal differenti-
ation.1, 6, 17-19
References
1. Northcutt AD, Nelson DM, Tschen JA. Axillary granular paraker-
atosis. J Am Acad Dermatol 1991;24:541-4.
2. Wallace CA, Pichardo RO, Yosipovitch G, Hancox J, Sangueza OP.
Granular parakeratosis: a case report and literature review. J Cutan
Pathol 2003;30:332-5.
3. English JC, Derdeyn AS, Wilson WM, Patterson JW. Axillary gran-
ular parakeratosis. J Cutan Med Surg 2003;7:330-2.
4. Meheregan DA, Thomas JE, Meheregan DR. Intertriginous granular
parakeratosis. J Am Acad Dermatol 1998;29:495-6.
5. Meheregan DA, Vandersteen P, Sikorski L, Meheregan DR. Axillary
granular parakeratosis. J Am Acad Dermatol 1995;33:373-5.
6. Wohlrab J, Juftul M, Wolter M, Marsch WC. Submammary granular
parakeratosis: an acquired punctate hyperkeratosis of exogenic origin.
J Am Acad Dermatol 1999;40:813-4.
7. Resnik KS, DiLeonardo M. Follicular granular parakeratosis. Am J
Dermatopathol 2003;25: 428-9.
8. Chang MW, Kaufmann JM, Orlow SJ, Cohen DE, Mobini N, Kamino
H. Infantile granular parakeratosis: recognition of two clinical patterns.
J Am Acad Dermatol 2004;50 (5 Suppl):S93-6.
9. Trowers AB, Assaf R, Jaworsky C. Granular parakeratosis in a child.
Pediatr Dermatol 2002;19:146-7.
10. Patrizi A, Neri I, Misciali C, Fanti PA. Granular parakeratosis: four
pediatric cases. Br J Dermatol 2002;147:1003-6.
11. Pimentel DR, Michalany N, Morgado de Abreu MA, Petlik B, Mota
de Avelar Alchorne M. Granular parakeratosis in children: case report
and review of the literature. Pediatr Dermatol 2003;20:215-20.
12. Metze D, Rutten A. Granular parakeratosis:a unique acquired disor-
der of keratinization. J Cutan Pathol 1999;26:339-52.
13. Barnes CJ, Lesher JL, Sangueza OP. Axillary granular parakerato-
sis. Int J Dermatol 2001;40:439-41.
14. Resnik KS, Kantor GR, DiLeonardo M. Dermatophyte-related gran-
ular parakeratosis. Am J Dermatopathol 2004;26:70-1.
15. Gartmann H, Steigleder GK. [Inguinal "pomade" crust of infants] Z
Hautkr 1975;50:667-9. German.
16. Patrizi A, Neri I, Marzaduri S, Fiorillo L. Pigmented and hyperkera-
totic napkin dermatitis: a liquid detergent irritant dermatitis. Derma-
tology 1996;193:36-40.
17. Sceppa J, Mowad C, Elenitsas R. Crusted plaques in the axillae. Arch
Dermatol 2001;137:1241-6.
18. Brown SK, Heilman ER. Granular parakeratosis: resolution with top-
ical tretinoin. J Am Acad Dermatol 2002;47(5 Suppl):S279-80
19. Webster CG, Resnik KS, Webster GF. Axillary granular paraker-
atosis: response to isotretinoin. J Am Acad Dermatol 1997;37:
789-90.
20. Chamberlain AJ, Tam MM. Intertriginous parakeratosis respon-
sive to potent topical corticosteroids. Clin Exper Dermatol 2003;
28:50-2.
21. Contreras ME, Gottfried LC, Bang RH, Palmer CH. Axillary intert-
riginous granular parakeratosis responsive to topical calcipotriene
and ammonium lactate. Int J Dermatol 2003;42:382-3.
447
TOMASINI
Paracheratosi granulare
L a paracheratosi granulare (granular parakeratosis, GP),
descritta per la prima volta nel 1991 da Northcutt et al.
con il nome di GP ascellare 1, è un pattern istologico a cui
corrisponde un quadro clinico peculiare che si osserva alle
pieghe corporee. Da allora sono stati riportati casi di GP a
interessamento inguinale, intra e sottomammario, genita-
le, perianale, al tronco e, persino alle ginocchia 2-7. Dal
punto di vista clinico si apprezzano papule e placche erite-
matose ipercheratosiche uni o bilaterali, spesso prurigino-
se. La dermatosi colpisce prevalentemente le pazienti di
sesso femminile di età inferiore ai 50 anni 3, sebbene siano
stati descritti anche casi pediatrici 8-11. Il decorso è, in gene-
re, cronico con scarsa risposta alla terapia e tendenza alla
risoluzione spontanea 1, 4, 6.
In questo lavoro viene segnalato un caso di GP della regio-
ne intra e sottomammaria con particolare attenzione ai pro-
blemi interpretativi che esso ha suscitato.
Caso clinico
Una paziente di 48 anni, in buone condizioni di salute,
giungeva all’ osservazione degli Autori per una dermatosi
localizzata nella regione infra e sottomammaria da un anno.
La paziente lamentava intenso prurito e sensazione di bru-
ciore. All’ anamnesi non risultava l’ uso di nuovi prodotti
per l’ igiene personale. Dal punto di vista clinico si osser-
vavano numerose piccole papule eritematose e cheratosiche
disseminate in zona sottomammaria e tra le pieghe mam-
marie (Figura 1). La componente cheratosica appariva fria-
bile e facilmente asportabile con il grattamento. La pazien-
te riferiva che lesioni simili erano apparse dapprima ai cavi
ascellari e alle pieghe inguinali e successivamente erano
regredite spontaneamente. Negativa risultava la familiarità per
malattie bollose, disordini della cheratinizzazione, psoriasi
o altre dermatosi. Veniva effettuato un prelievo bioptico di una
lesione localizzata alla piega sottomammaria destra. L’ esa-
me istologico rivelava la presenza di aree di iperplasia epi-
dermica psoriasiforme con minima spongiosi. Lo strato gra-
nuloso appariva conservato, mentre lo strato corneo era ispes-
sito e paracheratosico con ritenzione di granuli di chera-
toialina e accumuli di neutrofili (Figure 2, 3). Nel derma
superficiale si potevano osservare vasi ectasici e un infiltra-
to linfocitario superficiale perivascolare. La colorazione PAS
risultava negativa. Sulla base di queste caratteristiche isto-
logiche veniva formulata la diagnosi di GP. Un tampone per
esame microbiologico non evidenziava presenza di batteri,
dermatofiti o Candida. Veniva inizialmente intrapresa una
terapia con fluconazolo topico con scarsi risultati e, succes-
sivamente, veniva utilizzato un composto a base di betame-
tasone dipropionato in 2 applicazioni al dì con remissione del-
la dermatosi a distanza di 4 settimane.
448 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
GRANULAR PARAKERATOSIS
Discussione e conclusioni
La GP è un disordine acquisito della cheratinizzazione
che colpisce le pieghe, descritto per la prima volta nel 1991
in 4 pazienti affetti da un’ eruzione eritemato-cheratosica
localizzata alle ascelle 1. Da allora, per quanto ne sappia-
mo, sono stati riportati 42 casi: di questi 13 coinvolgevano
le ascelle, le pieghe inguinali, le aree infra e sottomamma-
rie, perianali e le ginocchia 2-7 e 8 riguardavano bambini di
età compresa tra i 9 e i 22 mesi 8-11. In questa serie limitata
venivano individuati 2 diversi pattern clinici: placche linea-
ri, bilaterali in sede inguinale e placche eritematose geome-
triche localizzate nei punti di pressione delle zone del pan-
nolino 8.
L’ eziologia della GP non è al momento nota. Inizialmente
si era ipotizzato che la GP fosse dovuta a un’ anomala rea-
zione da contatto a un deodorante o a un prodotto antiper-
spirante 1, 4, 12, 13, ma quest’ ipotesi non rendeva ragione del
frequente coinvolgimento monolaterale 4, la mancanza di
regressione con la sospensione del prodotto topico, la nega-
tività dei patch-test e la localizzazione in altre aree intertri-
ginose 1. Inoltre, l’ assenza di spongiosi nei preparati istologici
non supportava tale ipotesi. Secondo alcuni Autori la parti-
colare disposizione delle lesioni nelle aree intertriginose
suggeriva che il calore, l’ umidità la frizione e/o l’ obesità fos-
sero fattori predisponenti di tale dermatosi 1, 4, 12, 13. La net-
ta delimitazione nei punti di pressione al di sotto dei pannolini
potrebbe implicare un ruolo favorente degli assorbenti nel-
la GP dei neonati 8-11. La possibile eziologia micotica è sta-
ta suggerita dalla positività per Candida Albicans e dal riscon-
tro istologico di ife tra i corneociti in alcuni casi di GP 12-14.
Negli adulti, il meccanismo patofisiologico della GP
potrebbe consistere in un difetto di cheratinizzazione con
mancata sintesi di profilaggrina in filaggrina, una componente
proteica dei granuli di cheratoialina 12. La funzione della
filaggrina non è ancora del tutto chiara, ma si presume che,
all’ interno delle cellule cornee, funga da matrice di adesio-
ne essenziale per la produzione di ponti disolfuro tra i fila-
menti di cheratina, producendo, così, il pattern strutturale
della cheratina tipico delle cellule cornee. La conversione del-
la profillaggrina nelle due subunità monometriche deve esse-
re attentamente controllata per prevenire un’ aggregazione
prematura. I pazienti affetti da GP presentano un difetto di
degradazione dei granuli di cheratoialina e di aggregazione
dei filamenti di cheratina 2.
La diagnosi differenziale clinica di una dermatosi papu-
lo-cheratosica localizzata alle pieghe comprende un ampio
spettro di disordini quali la malattia di Hailey-Hailey, la
malattia di Darier, il pemfigo vegetante, le verruche piane, l’
acanthosis nigricans, la psoriasi, la tinea, la dermatite sebor-
roica, l’ intertrigo, la candidosi, l’ istiocitosi a cellule di Lan-
gerhans, l’ eczema nummulare e le cheratosi seborroiche in
fase iniziale 1, 4-6, 9, 12. L’ esame istologico è dirimente per la
Agosto 2005
GRANULAR PARAKERATOSIS
corretta diagnosi evidenziando la caratteristica parachera-
tosi granulare a livello dell’ epidermide, sebbene in qualche
caso tale fenomeno sia limitato all’ infundibulo follicolare 7.
Nei neonati, la GP presenta notevoli similarità a una par-
ticolare dermatosi descritta per la prima volta da Gartmann
et al. nel 1975 con il nome di “Pomaden Kruste” 15 nella
letteratura tedesca e poi riportata da quella anglosassone
come “pigmented and hyperkeratotic napkin dermatitis” 16,
attribuita a un trattamento eccessivo della cute genitale dei
neonati con creme e olii. È verosimile che questi quadri, in
realtà, rappresentino esempi di GP.
In conclusione, la GP rappresenta un disordine della che-
ratinizzazione a eziologia verosimilmente multifattoriale.
Ciò potrebbe spiegare la particolare predisposizione per le
aree intertriginose.
Il trattamento della GP non è codificato. Sebbene molti casi
mostrino una risoluzione spontanea 4,12, 17, si è notata una
risposta variabile agli steroidi topici e per os, agli antimi-
crobici, agli antifungini e ai cheratolitici 18-21. Sebbene i
retinoidi topici non appaiano efficaci, l’ uso di retinoidi ana-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
TOMASINI
loghi della vitamina D ha sortito risultati incoraggianti, avva-
lonando l’ ipotesi di un’ anomalia della differenziazione che-
ratinocitaria alla base di questa dermatosi 1, 6, 17-19.
Riassunto
La paracheratosi granulare (GP) è un disordine acquisito
della cheratinizzazione a eziologia sconosciuta che colpi-
sce prevalentemente le pieghe di soggetti adulti. In questo
lavoro è riportato il caso di una paziente di 48 anni che pre-
sentava da circa un anno papule cheratosiche alle pieghe
infra e sottomammarie. Una biopsia cutanea evidenziava il
quadro della paracheratosi granulare. L’ eruzione regrediva
rapidamente con terapia steroidea topica. Vengono, inoltre,
discusse le problematiche eziopatogenetiche e di diagnosi dif-
ferenziale.
PAROLE CHIAVE: Paracheratosi granulare - Disordini della
cheratinizzazione - Cute.
449
 LETTERS TO THE EDITOR
G ITAL DERMATOL VENEREOL 2005;140:451-8

Non recurrent eosinophilic cellulitis:

report of an atypical case
M. L. BERNARDINI 1, G. BRANDOZZI 1,
A. CAMPANATI 1, L. GIORNETTA 1,
M. GIANGIACOMI 2, A. OFFIDANI 1
1Division of Dermatology
Università Politecnica delle Marche
Ospedali Riuniti Hospital, Ancona, Italy
2Division of Anatomy and Istology
Università Politecnica delle Marche
Ospedali Riuniti Hospital, Ancona, Italy

Dear Sir,
Figure 1.—Large oedematous and erythematous lesion on the right knee.

E osinophilic cellulitis (EC) was described for the first

time by Wells in 1971 1 as a distinctive pathologic enti-
ty of unknown etiology, although it was given prominence by
many authors because of its association with skin viral and
bacterial infections, arthropod bites, cutaneous parasitic
infestations (onchocerciasis, ascariasis and toxocariasis),
hypereosinophilic syndromes, myeloproliferative disorders,
leukaemias, anal carcinoma, drug intake or surgical inter-
vention. Further associations were a history of Reynaud pos-
itive phenomenon and urticaria, genetic inheritance.2 Patients
affected by Wells’ syndrome (WS) usually present with one
or more cutaneous plaques resembling an acute bacteric cel-
lulitis while cultural examination for bacteria is always neg-
ative and oral antibiotics never seemed to improve clinical or
histologic lesions, except for a single case report in which
there was a good response to minocycline.3 Lesions resolve
spontaneously after weeks or months without residual scar-
ring. We report the case of a 23 year-old female presenting
with a 2 weeks history of an itchy annular-erythematous and
oedematous plaque, surrounded by a wide violaceous and Figure 2.—Smaller lesion involving the left knee.
swelling border first localized on the anterior surface of the
right knee and then also on the anterior surface of the left knee
(Figures 1, 2). Our patient had already been administered
an antihistaminic drug per os (cetirizine 10 mg/die for 10
acute bacteric cellulitis she was prescribed antibiotics per os
days) which barely influenced her clinical manifestations.
(amoxicillin/clavulanic acid, 2 g/die) for 10 days but her
Fever and general malaise were absent. At her first derma-
clinical manifestations did not improve in any way. Labora-
tologic visit plaques appeared hot on palpation and the exam-
ination revealed the presence of a slight regression of the tory analysis disclosed: the absence of peripheral eosinophil-
erythema in the central area and of an enlarging red-viola- ia; erythrocyte sedimentation rate: 40 mm in the first hour;
ceous and swelling border. According to a first diagnosis of normal urinalysis, immunoglobulin (IgE included) levels,
immunologic serology (C3 and C4; anti-nuclear antibod-
ies); cultural examination for bacteria and fungi was nega-
The case has been presented as a communication at the 79th SIDEV
National Congress of the Società Italiana di Dermatologia e Venereologia tive; stool examinations were negative for parasites and ova.
2004 and at the 1st Congresso Marchigiano di Dermatologia e Venereologia Ultrasonography showed an absence of joint involvement.
(25 September, 2004). Haematoxylin-eosine stained histologic section, on bilater-

Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA 451

LETTERS TO THE EDITOR
Figure 3.—EE-10×. Acanthosis and focal hyperkeratosis. Presence of der-
mal perivascular lymphocytic infiltration.
Figure 4.—EE-20×. Magnification of the previous image. Infiltration of the
dermis with eosinophilic granulocytes, accompanied by slight perivascu-
lar lymphocytic infiltration. Oedema is absent in the upper dermis.
al biopsy specimens, showed oedematous dissociation of
dermal collagen fibers and moderate perivascular lympho-
plasmacellular infiltrate associated with the presence of
eosinophilic infiltration. Both diffuse infiltration of dermal
fibers or small clusters of eosinophilic granulocytes (E) were
present. Vasculitis was absent. Histiocytes palisade sur-
rounding focal deposition of eosinophilic material and nuclear
debris (Figures 3-5), the so called “flame figure” (FF) was
detected. Because of both clinical and histologic findings
452 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
Figure 5.—EE-40×. Flame figure. Histiocytes surrounding focal deposi-
tion of eosinophilic material.
we concluded that our patient fulfilled the criteria for the
diagnosis of WS. We explained the poor presence of flame
figures in our patient’s histologic sections as a probable con-
sequence of the first antihistaminic treatment she was admin-
istered. We prescribed betametasone per os, 2 m/die, pro-
gressively tapered over 10 days and antihistamines per os (cet-
irizine 10 m/die) for 15 days. Skin lesions resolved within 2
weeks and she has experienced no manifestation recurrence
since, after 2 years follow-up. In WS a predominant
eosinophilic infiltrate was observed early in the upper and
deep dermis. Distinctive FF become apparent in the subacute
phase when degranulating E coat basophilic collagen bundles
with eosinophilic major basic protein (MBP). The last phase
showed a histiocyte palisade around FF. Typical FF were
detected but scantely in the patient’s histologic sections: this
was probably due to the antihistaminic treatment she was
already administered before she was referred to our Derma-
tologic Clinic. The excessive dermal infiltration of E with FF
appears to be a peculiar response to a variety of potential
triggering events. Clonally expanded type-2 helper T cells
overproducing interleukin-5 (IL-5), the main cytokine in the
development and differentiation of E, could be the underly-
ing mechanism in eosinophilic diseases. Yagi et al.4 have
observed a high proportion of circulating T CD4+CD7- cells
in patients with WS before treatment and their findings sug-
gest that such cells could play a pivotal role in the patho-
genesis of EC by producing IL-5. IL-5 not only mobilizes
eosinophils from bone marrow to blood but also IL-5 seems
to promote skin homing by presumably altering the expres-
sion of adhesion molecules. Espana et al.5 found a good cor-
relation between the clinical activity of the disease, the lev-
els of E in the blood and bone marrow and the IL-5 and
eosinophilic cationic protein levels in the peripheral blood.
Agosto 2005

The FF associated with EC are not pathognomonic, as they
can be seen in other unrelated dermatoses. FF may be seen
incidentally in bullous pemphigoid, pemphigoid gestation-
is, tinea, spider and insect bite reactions and other inflam-
matory conditions in which numerous eosinophils are pre-
sent.6 Diagnosis is, therefore, the result of the matching of all
anamnestic, clinical and histologic findings.
References
1. Wells GC. Recurrent granulomatous dermatitis with eosinophilia.
Trans St Johns Hosp Dermatol Soc 1971;57:46-56.
2. Weiss G, Shemer A, Confino Y, Kaplan B, Trau H. Wells’ syndrome:
report of a case and review of the literature. Int J Dermatol 2001;40:
148-52.
3. Stam-Westerveld EB, Daenen S, Van der Meer Jb, Jonkman MF.
Eosinophilic cellulitis (Wells’ syndrome): treatment with minocy-
cline. Acta Derm Venereol 1998;78:157.
4. Yagi H, Tokura Y, Matsushita K, Hanaoka K, Furukawa F, Takigawa
M. Wells’ syndrome: a pathogenic role for circulating CD4+ CD7- cells
expressing interleukin-5 mRNA. Br J Dermatol 1997;136:918-23.
5. Espana A, Sanz ML, Sola J, Gil P. Well’s sindrome (eosinophilic cel-
lulitis): correlation between clinical activity, eosinophil levels,
eosinophil cation protein and interleukin-5. Br J Dermatol
1999;140:127-30.
6. Moosavi M, Mehregan DR. Wells’ syndrome: a clinical and histopatho-
logic review of seven cases. Int J Dermatol 2003;42:62-7.
Address reprint requests to: Dr. M. L. Bernardini, Via V. Veneto 24,
60122 Ancona. E-mail: mbernardini@email.it
Un caso di cellulite eosinofilica
(Sindrome di Wells) non recidivante
con quadro istologico atipico
Egregio Direttore,
L a cellulite eosinofilica (CE) è stata descritta per la prima
volta da Wells nel 1971 come un’entità patologica distin-
ta a eziologia sconosciuta 1. Viene, comunque, segnalata in
letteratura una concomitanza della malattia con infezioni
cutanee batteriche o virali, morsi di artropodi, infestazioni
(oncocerchiasi, ascariasi e toxocariasi), sindromi ipereosi-
nofile, disordini mieloproliferativi, leucemie, carcinoma
anale, assunzione di farmaci o interventi chirurgici; più rara-
mente è stata descritta un’anamnesi positiva per il fenome-
no di Reynaud, orticaria ed ereditarietà genetica 2. I pazien-
ti con sindrome di Wells (SW) in genere presentano una o
poche placche cutanee che ricordano una cellulite acuta, tut-
tavia l’esame colturale per batteri non risulta mai positivo e
gli antimicrobici somministrati per via sistemica non sem-
brano migliorare il quadro clinico o istologico delle lesioni,
a eccezione di un singolo caso di risposta alla minociclina
riportato in letteratura 3. Le lesioni iniziali evolvono velo-
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
LETTERS TO THE EDITOR
cemente in placche che tendono alla risoluzione spontanea,
dopo settimane o mesi, senza residuare cicatrici.
Segnaliamo il caso di una paziente di 23 anni che si pre-
sentava alla nostra attenzione per la comparsa, da circa 15
giorni, di una placca anulare eritemato-edematosa, con ampio
bordo periferico edematoso e violaceo, localizzata dapprima
sulla superficie estensoria del ginocchio destro e, in seguito,
anche alla superficie estensoria del ginocchio sinistro (Figu-
re 1, 2). La paziente veniva inizialmente trattata con un anti-
istaminico per os (cetirizina 10 mg/die per 10 giorni) che
sortiva solo modesti effetti clinici. Al momento della prima
visita dermatologica la paziente lamentava una sintomatologia
locale fortemente pruriginosa. Risultavano assenti febbre e
malessere generale. Le placche, calde al tatto, apparivano
di grandi dimensioni, di colore rosso-rosa, con area centra-
le color camoscio e bordo periferico rosso-violaceo ed ede-
matoso in espansione. Nell’ipotesi di una forma di cellulite
batterica acuta veniva prescritto un farmaco antibiotico
(amoxicillina/acido clavulanico 1 g, 2 compresse/die) per
10 giorni. Il quadro sintomatologico e obiettivo, tuttavia,
non traeva beneficio dalla terapia e la paziente mostrava una
lieve progressione delle manifestazioni. Venivano, pertan-
to, programmati specifici esami di laboratorio e una biopsia
incisionale bilaterale per la diagnosi istologica delle lesioni.
Le indagini di laboratorio mostravano: emocromo con formula
leucocitaria nella norma e assenza di eosinofilia periferica.
Gli indici di flogosi erano positivi: VES=40 mm 1° ora. Nel-
la norma risultavano anche le analisi urinarie, il dosaggio
delle immunoglobuline, incluse le IgE, la sierologia immu-
nologica (complemento, anticorpi anti-nucleo); l’esame col-
turale per batteri e funghi era negativo; gli esami delle feci
risultavano negativi per uova e parassiti. All’esame ecogra-
fico risultava assente il coinvolgimento delle articolazioni
del ginocchio sottostanti. La sezione istologica, colorata in
ematossilina-eosina, mostrava una discreta dissociazione
edematosa delle fibre collagene del derma e un moderato
infiltrato linfoplasmacellulare perivascolare, associato alla pre-
senza di granulociti eosinofili (E). Questi ultimi apparivano
sia dispersi tra le fibre che in piccoli aggregati. Non vi era
alcuna evidenza di vasculite. Era, inoltre, presente un depo-
sito focale di materiale eosinofilico con detriti nucleari, cir-
condato da alcuni istiociti (Figure 3-5). Tale reperto era inter-
pretabile come una classica «figura a fiamma» (FF), carat-
teristica, sia pure non patognomonica, delle CE. L’insieme dei
reperti unitamente ai dati clinico-anamnestici orientavano
verso una diagnosi di SW. Veniva prescritto, pertanto, corti-
sone per via generale (betametasone per os, 2 mg/die a sca-
lare) per un totale di 10 giorni di terapia e, in associazione,
un anti-istaminico per os (cetirizina compresse 10 mg/die) per
15 giorni. Il quadro clinico tendeva a mostrare netti segni di
miglioramento già dopo la prima settimana di terapia e si
aveva una completa risoluzione entro 2 settimane. La pazien-
te non ha mostrato a tutt’oggi recidive, a un follow-up di cir-
ca 2 anni. Nella SW i reperti istologici sono generalmente
caratterizzati da un iniziale infiltrato di E nel derma super-
ficiale e profondo. Le FF si formano durante la fase subacu-
ta quando gli E, degranulanti, rivestono i fasci di fibre col-
453
LETTERS TO THE EDITOR

lagene di proteina eosinofilica basica maggiore. Nella fase

di risoluzione gli istiociti fagocitici si dispongono a palizzata
circondando le FF. Le classiche FF, nel nostro caso, appaio-
no poco evidenti: con tutta probabilità ciò è dovuto al fatto
che una terapia anti-istaminica era già stata instaurata prima
che la paziente si rivolgesse alla nostra attenzione. La mas-
siva infiltrazione dermica di E, con la comparsa di FF, sem-
bra essere una risposta peculiare a una varietà di potenziali
eventi trigger. L’iperproduzione di interleuchina 5 (IL-5), la
citochina più importante per lo sviluppo e la differenziazio-
ne degli E, da parte di cellule T helper 2 in espansione clo-
nale, potrebbe rappresentare il meccanismo che sottende le
patologie caratterizzate dalla presenza di infiltrato di E. Yagi
et al. 4 hanno osservato che i linfociti T CD4+CD7- risulta-
no aumentati nel siero di pazienti con SW non in trattamen-
to e suggeriscono che queste cellule svolgano un ruolo fon-
damentale nella patogenesi della CE grazie alla produzione
di IL-5. IL-5 non solo mobilita gli E dal midollo osseo al
sangue periferico, ma sembra promuovere anche il proces- Figura 1. — At the time of our observation: a large bulla with a serous con-
so di homing cutaneo degli stessi, presumibilmente alteran- tent at the level of the internal malleolus and some erosions partially cove-
do l’espressione delle specifiche molecole di adesione. Espa- red by squamous scabs on the medial surface of the left leg.
na et al. 5 hanno mostrato l’esistenza di una buona correla-
zione tra attività clinica di malattia, livelli di E nel sangue peri-
ferico e midollo osseo e livelli di IL-5 e proteina cationica ve in reducing the symptoms and severity of the disease in
eosinofilica nel sangue periferico. Una CE con FF non è both adults and children.
patognomonica di questa sindrome poiché rappresenta una However, although limited to individual cases, an increa-
reazione istologicamente ben definita, osservabile anche in sing number of reports have described its efficacy in treating
altre patologie. Può essere, infatti, occasionalmente presen- other immunomediated dermatoses, including lichen ruber
te in corso di pemfigoide bolloso, herpes gestationis, tigna, planus, lichen sclerosus et atrophicus, gangrenous pyoder-
reazione da puntura di insetti o di aracnidi e altre condizio- ma and alopecia areata.2
ni infiammatorie in cui risultino presenti numerosi eosino- The case of a woman with localized bullous pemphigoid
fili 6. La diagnosi, pertanto, deve essere il risultato della that rapidly resolved after topical treatment with 0.1% tacro-
combinazione di reperti anamnestici, clinici e istologici. limus ointment is reported.
A 45-year-old woman in good general conditions attended
our outpatient clinic about 1 month after the appearance of
erythematovesicular-bullous lesions associated with intense
Topical tacrolimus pruritus on the lower third of her left leg (Figure 1).
These lesions arose in the area of a large scar that was the
in the tretment sequela of 2 previous surgical operations undergone becau-
of localized bullous pemphigoid se of a torn Achilles tendon (the first in 1997 and the second
in 1999). No lesions were found in other skin areas nor any
mucosal involvement.
E. FRIGERIO, C. FRANCHI
Histological and direct immunofluorescence examina-
G. CAINELLI,G. F. ALTOMARE
tions of the perilesional skin confirmed the clinical diagno-
Department of Dermatology sis of localized bullous pemphigoid. The results of routine
University of Milan, Galeazzi Hospital, Milan, Italy blood chemistry tests were normal, and indirect immuno-
fluorescence was negative.
Given the marked atrophy of the cicatricial tissue, we
Dear Sir, decided to avoid the use of highly potent topical steroids
and administer twice-daily a local therapy with 0.1% tacro-

T acrolimus ointment is a new topical immunomodulator

that by specifically inhibiting the activity of calcineu-
rin blocks the early phases of T cell activation and the pro-
duction of various cytokines (IL2, IL3, IL4, G-CSF, TNF…).1
A number of clinical studies have so far been carried out
in order to investigate the use of topical tacrolimus in the
treatment of atopic dermatitis, since it proved to be effecti-
limus ointment with an occlusive bandage and systemic anti-
histamine treatment to control the pruritic symptoms (levo-
cetirizine 5 mg/day).
At the control examination after 2 weeks, the manifesta-
tion had improved with the re-epithelialisation of the pre-
vious bulla and the absence of new lesions (Figure 2); mode-
rate pruritus persisted. One month therapy led to complete

454 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA Agosto 2005


Figura 2. — At the two-week control examination: re-epithelialisation of
the previous bulla and the absence of new lesions.
remission of the dermatosis, and so maintenance treatment
with a single daily administration of 0.1% tacrolimus oint-
ment was recommended. No relapse occurred after 2 months.
In the localized bullous pemphigoid the same desmosomal
antigens as those of generalized bullous pemphigoid have
been recognized, and T lymphocytes play a determining role
in the pathogenesis of the lesions.
It has been demonstrated that T lymphocytes are impor-
tant for the induction and regulation of both cell- and anti-
body-mediated immune responses in various autoimmune
diseases.
It has been found that the serum of patients with bullous
pemphigoid contains self-reactive T lymphocytes that pro-
duce both Th2 (IL4 and IL13) and Th1 cytokines (gamma-
IFN), which respectively regulate the secretion of IgG4 and
IgG1 by B lymphocytes. In particular, self-reactive T cells that
recognized the bullous pemphigoid antigen (BP 180) and
give rise to a Th2-mediated response have only been found
in the serum of bullous pemphigoid patients.3
Given the limited extension of the lesions, the majority of
cases of localized bullous pemphigoid do not require sytemic
immunosuppressive treatment, and the therapy of choice is
based on the use of highly potent topical corticosteroids.
The authors of 2 case reports published in 2003 (1 of
dyshidrosis-form localized bullous pemphigoid and 1 of
generalized bullous pemphigoid) described the good results
obtained using this new topical immunomodulator inhibiting
T lymphocytes.4, 5
In our case, we preferred using this new immunomodulator
rather than conventional topical corticosteroid therapy in
order to avoid worsening the already considerably atrophied
cicatricial tissue at the site of the bullous lesions because,
unlike corticosteroids, tacrolimus does not give rise to local
side effects even after long term treatment; in particular, it
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
LETTERS TO THE EDITOR
does not show the atrophising effect which is typical of topi-
cal corticosteroids.
Its only side effect is a sensation of burning and transient
pruritus at the application site, which our patient did not
experience despite the use of an occlusive bandage.
Although large-scale studies are necessary in order to
demonstrate the real efficacy of this new immunomodulator,
our experience suggests that tacrolimus is a possible alter-
native to topical corticosteroids also in the treatment of
autoimmune bullous dermatoses.
Bibliografia
1. Dumont FJ. FK 506, an immunosuppressant targeting calcineurin
function. Curr Med Chem 2000;7:731-48.
2. Gupta AK, Adamiak A, Chow M. Tacrolimus: a review of its use for
the management of dermatoses. J Eur Acad Dermatol Venereol
2002;16:100-14.
3. Büdinger L, Borradori L, Yee C, Eming R, Ferencik S, Grosse-Wil-
de H et al. Identification and characterization of autoreactive T cell
responses to bullous pemphigoid antigen 2 in patients and healthy
controls. J Clin Invest 1998;102:2082-9.
4. Ko M-J, Chu C-Y. Topical tacrolimus therapy for localized bullous
pemphigoid. Br J Dermatol 2003;149:1079-80.
5. Chu J, Bradley M, Marinkovich MP. Topical tacrolimus is a useful
adjunctive therapy for bullous pemphigoid. Arch Dermatol
2003;139:813-5.
Address reprint requests to: Prof. G. Altomare, Via Riccardo Galeazzi
4, 20161 Milano. E-mail: gianfranco.altomare@unimi.it
Tacrolimus topico nella terapia
del pemfigoide bolloso localizzato
Egregio Direttore,
I l tacrolimus è un nuovo immunomodulatore che inibendo
in modo specifico l’attività della calcineurina blocca le
fasi precoci di attivazione delle cellule T e la produzione di
svariate citochine (IL2, IL3, IL4, G-CSF, TNF …)1.
A tutt’oggi sono stati effettuati vari studi clinici riguardanti
l’utilizzo del tacrolimus topico nella terapia della dermatite
atopica dove si è rivelato efficace nel ridurre i sintomi e la gra-
vità della malattia sia nell’adulto che in età pediatrica.
Sono, comunque, in continuo aumento le segnalazioni,
seppur spesso limitate a singoli casi, della sua efficacia tera-
peutica anche in altre dermatosi immunomediate, tra cui,
per citarne alcune, il lichen ruber planus, il lichen sclero-
atrofico, il pioderma gangrenoso e l’alopecia areata 2.
Riportiamo il caso di una paziente affetta da pemfigoide
bolloso localizzato in cui il trattamento topico con tacrolimus
0,1% unguento ha determinato la rapida risoluzione della
dermatosi.
455
LETTERS TO THE EDITOR
La paziente, di 45 anni, in buone condizioni generali, si è
presentata presso il nostro ambulatorio per la comparsa da cir-
ca 1 mese al III inferiore della gamba sinistra di alcuni ele-
menti eritemato-vescico-bollosi associati a intenso prurito
(Figura 1).
Tali lesioni insorgevano nel contesto di un’ampia cicatri-
ce esito di 2 pregressi interventi chirurgici per la rottura del
tendine d’Achille (il primo nel 1997 e il secondo nel 1999).
Non si evidenziavano lesioni in altre zone cutanee né inte-
ressamento delle mucose.
L’esame istologico e l’immunofluorescenza diretta effet-
tuati su cute perilesionale confermavano la diagnosi clinica
di pemfigoide bolloso localizzato. Gli esami ematochimici
di routine risultavano nella norma e l’immunofluorescenza
indiretta era negativa.
Considerata la spiccata atrofia del tessuto cicatriziale,
abbiamo ritenuto opportuno evitare l’utilizzo di topici ste-
roidei a elevata potenza e abbiamo deciso di impostare una
terapia locale con tacrolimus 0,1% unguento 2 volte/die con
bendaggio occlusivo e anti-istaminici sistemici per control-
lare la sintomatologia pruriginosa (levocetirizina 5 mg/die).
Al controllo dopo 2 settimane la manifestazione era miglio-
rata con riepitelizzazione della pregressa bolla e assenza di
nuove lesioni (Figura 2), permaneva modesto prurito. A 1
mese dall’inizio della terapia la dermatosi andava in completa
remissione per cui, come mantenimento, veniva consigliata
un’unica applicazione quotidiana di tacrolimus 0,1% unguen-
to. Non si sono evidenziate recidive a distanza di 2 mesi.
Nel pemfigoide bolloso localizzato, vengono riconosciu-
ti gli stessi antigeni desmosomiali del pemfigoide bolloso
generalizzato e nella patogenesi delle lesioni rivestono un ruo-
lo determinante i linfociti T.
È stato dimostrato che i linfociti T sono importanti sia
nell’induzione sia nella regolazione di entrambe le risposte
immuni, cellulo-mediate e anticorpo-mediate, in diverse
patologie autoimmuni.
Nel siero di pazienti affetti da pemfigoide bolloso sono sta-
ti riscontrati linfociti T autoreattivi producenti sia citochine
Th2 (IL4 e IL13) che Th1 (IFN gamma) le quali regolano
rispettivamente la secrezione di IgG4 e IgG1 da parte dei
linfociti B. In particolare, solo nel siero di questi pazienti
sono state ritrovate cellule T autoreattive che riconoscono
l’antigene del pemfigoide bolloso (BP180) dando origine a
una risposta autoimmune Th2 mediata 3.
Il pemfigoide bolloso localizzato, nella maggior parte dei
casi, data la limitata estensione delle lesioni, non necessita
del ricorso a un trattamento immunosoppressivo sistemico,
infatti la terapia d’elezione si basa sull’utilizzo di cortico-
steroidi topici a elevata potenza.
Nel 2003 sono stati riportati 2 casi, un pemfigoide bollo-
so localizzato disidrosiforme e un pemfigoide bolloso gene-
ralizzato, in cui gli Autori si sono avvalsi, con buoni risultati,
di questo nuovo immunomodulatore topico che inibisce i
linfociti T 4, 5.
Nel nostro caso abbiamo preferito ricorrere a tacrolimus
unguento piuttosto che alla terapia convenzionale con cor-
ticosteroidi topici per evitare il peggioramento della già
456 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
importante atrofia del tessuto cicatriziale sede delle lesioni
bollose.
Il tacrolimus, infatti, a differenza dei corticosteroidi non
determina effetti collaterali locali anche nell’utilizzo per
lunghi periodi, in particolare non presenta quell’effetto atrofo-
genico proprio dei corticosteroidi topici. L’unico effetto col-
laterale consiste nell’insorgenza di bruciore e prurito tran-
sitorio nella sede di applicazione, effetto tra l’altro non veri-
ficatosi nella nostra paziente nonostante il ricorso alla tera-
pia in occlusiva.
Sulla base della nostra esperienza, in attesa di studi su
ampie casistiche che permettano di dimostrare l’effettiva
efficacia di questo nuovo immunomodulatore, riteniamo il
tacrolimus una possibile alternativa ai corticosteroidi topici
anche nelle dermatosi bollose autoimmuni.
The safety profile of topical pimecrolimus
in the treatment of atopic dermatitis
G. GIROLOMONI 1, C. GELMETTI 2, A. VIERUCCI 3
1Department
of Biomedical and Surgical Sciences
Section of Dermatology, University of Verona, Verona, Italy
2Institute of Dermatological Sciences
IRCCS Ospedale Maggiore, University of Milan, Milan, Italy
3Department of Pediatrics, University of Florence
A. Meyer Pediatric Hospital, Florence, Italy
Dear Editor
T he United States Food & Drug Administration (FDA)
has recently issued a Public Health Advisory about the
safety profile of topical calcineurin inhibitors (better
known as topical immunomodulators or TIMs: pime-
crolimus and tacrolimus). The FDA also announced of its
intention to add a “black-box” warning (a special warning
on the drug package) to the labeling for the 2 drugs, and
that current indications for which these drugs have been
granted approval might be subject to change. This action
was based on a recommendation from the FDA Pediatric
Advisory Committee because of concerns of potential
safety risks (especially skin cancer and lymphoma) in
pediatric patients affected with atopic dermatitis (AD)
and receiving therapy with a TIM. We feel deeply troubled
by the FDA’s actions, because there is no evidence that top-
ical use of pimecroliums and tacrolimus is harmful. The
American Academy of Dermatology, the American Acad-
emy of Allergy, Asthma and Immunology, the Society for
Pediatric Dermatology, The British Association of Der-
Agosto 2005

matologists, the European Society for Pediatric Derma-
tology, the German Society of Dermatology, the Austrian
Society of Dermatology, the European Dermatology
Forum, the European Academy of Dermatology and
Venereology, and other scientific societies as well as
patient associations (National Society for Atopic Eczema)
have unanimously expressed similar concern for the FDA’s
actions.
Over the last decade, both agents have been extensive-
ly studied in clinical trials and their efficacy and safety in
the treatment of AD have been demonstrated.1-5 In partic-
ular, pimecrolimus has been investigated in more than
19 000 patients (of which 2 600 were infants and 7 300
were children). In addition, more than 5 million patients,
over half of them children, have been treated with pime-
crolimus since its approval, and post marketing surveil-
lance studies (PMS) show that there is no clinical evi-
dence suggesting an increased risk of malignancies in
patients treated with pimecrolimus. In particular: a) the
incidence of malignancies in patients treated with pime-
crolimus in the course of clinical studies was lower than that
in patients in the control group treated with placebo or
topical corticosteroids (2 cases out of 19 000 versus 5 cas-
es out of 4 000); b) PMS data from clinical use in over 5
million patients treated with pimecrolimus, as of March
2005, show very few cases of malignancies. The number
is well below the expected background incidence in the
population treated and, in addition, is lower than the rate
expected in the general population. Moreover, no causal
relationship with pimecrolimus was established for any
of the cases reported; c) those lymphomas identified by
spontaneous adverse event reporting systems do not have
the clinical presentation and histology that characterize
lymphomas occurring in the setting of immunosuppres-
sive therapy.
Systemic absorption of both drugs is very limited,6 and
even though in some patients blood concentrations have
been detected, the values are usually very low and insufficient
to cause the sustained systemic immunosuppression that
would be responsible for lymphomas. This is true even in
young children with moderate-to-severe dermatitis affect-
ing a large body surface area.
There is no evidence of photocarcinogenic or mutagenic
potential in animals treated with pimecrolimus. Lymphomas
that have the characteristics of diseases related to systemic
immunosuppression have only been observed in animals
exposed to high systemic levels of calcineurin inhibitors.
These animals experienced prolonged systemic exposure
that is much greater than that achieved with topical applica-
tion in humans.
There is no increased incidence of systemic or cutaneous
infections in patients treated with topical pimecrolimus.
There is no evidence of systemic immunosuppression due
to topical pimecrolimus as showed by antibody response to
vaccination and by tests on delayed-type hypersensitivity.7
The European Medicines Agency (EMEA) did not take any
immediate action but decided to start a referral process. This
Vol. 140 - N. 4 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
LETTERS TO THE EDITOR
is a process consisting of a review of the full body of the
clinical data that will allow a full evidence-based evaluation
of the risk /benefit profile of these drugs.8 The EMEA feels
that such a comprehensive review of the clinical data is the
most effective way to achieve an objective assessment and
provide guidance for patients and physicians regarding the
use of TIMs. In the meantime, drug labels remain unchanged.
AD is a chronic, recurring and frustrating condition.
Many patients suffer from AD on the face and sensitive skin
sites where long-term application of topical corticosteroids
is not indicated. Patients need alternatives to topical steroids
due to side effects. The health and safety of our patients are
of paramount importance to physicians. We are concerned
that the aforementioned warnings confuse and unneces-
sarily worry our patients and their families as well as health
care providers. It is the responsibility of health authorities
to present a balanced and fair review of the evidence. Cur-
rent labeling sufficiently describes the appropriate use and
safety of these medications. We strongly believe that the
recent recommendations of the Pediatric Advisory Com-
mittee and the FDA Health Alert are not justified on the
basis of scientific evidence and should be revised. In order
to provide further evidence confirming the safety of pime-
crolimus, long-term clinical studies have been started,
with a registry including 4 000 pediatric patients who will
be followed for a period of 10 years.
References
1. Ashcroft DM, Dimmock P, Garside R, Stein K, Williams HC. Efficacy
and tolerability of topical pimecrolimus and tacrolimus in the treatment
of atopic dermatitis: meta-analysis of randomised controlled trials. Br
Med J 2005;330:516-24.
2. Meurer M, Fartasch M, Albrecht G, Vogt T, Worm M, Ruzicka T et al.
Long-term efficacy and safety of pimecrolimus cream 1% in adults with
moderate atopic dermatitis. Dermatology 2004;208:365-72.
3. Kempers S, Boguniewicz M, Carter E, Jarratt M, Parisier D, Stewart
D et al. A randomized investigator-blinded study comparing pime-
crolimus cream 1% with tacrolimus ointment 0.03% in the treatment
of pediatric patients with moderate atopic dermatitis. J Am Acad Der-
matol 2004;51:515-25.
4. Papp KA, Werfel T, Folster-Holst R, Ortonne JP, Potter PC, de Prost
Y et al. Long-term control of atopic dermatitis with pimecrolimus
cream 1% in infants and young children: a two-year study. J Am Acad
Dermatol 2005;52:240-6.
5. Luger TA, Lahfa M, Folster-Holst R, Gulliver WP, Allen R, Molloy S
et al. Long-term safety and tolerability of pimecrolimus cream 1% and
topical corticosteroids in adults with moderate to severe atopic der-
matitis. J Dermatol Treat 2004;15:169-78.
6. Billich A, Aschauer H, Aszòdi A, Stuetz A. Percutaneous absorption
of drugs used in atopic eczema: pimecrolimus permeates less through
skin than corticosteroids and tacrolimus. Int J Pharmacol 2004;269:
29-35.
7. Papp KA, Breuer K, Meurer M, Ortonne JP, Potter PC, de Prost Y
et al. Long-term treatment of atopic dermatitis with pimecrolimus
cream 1% in infants does not interfere with the development of
protective antibodies after vaccination. J Am Acad Dermatol
2005;52:247-53.
8. European Medicines Agency: Committee for Medical Products for
Human Use. Press release. April 18-21, 2005.
457
LETTERS TO THE EDITOR
Il profilo di sicurezza del
pimecrolimus topico
nella terapia della dermatite atopica
Egregio Direttore,
L a Food & Drug Administration (FDA) americana ha
recentemente richiamato l’attenzione sul profilo di sicu-
rezza degli inibitori della calcineurina a uso topico, meglio
noti come immunomodulatori topici (o topical immunomo-
dulators, TIM): pimecrolimus e tacrolimus. In questo avvi-
so, inoltre, la FDA informa la classe medica e il pubblico del-
la propria intenzione di apporre una speciale avvertenza
(«black-box» warning) sulla confezione dei farmaci riguar-
do il potenziale rischio di sviluppo di neoplasie (special-
mente linfomi e tumori cutanei) nei pazienti pediatrici affet-
ti da dermatite atopica (DA) sottoposti a terapia con TIM, e
l’eventualità di apportare una modifica alle attuali indica-
zioni per cui tali farmaci sono registrati. Questa eventuale
azione dell’FDA ci preoccupa, in quanto non esiste alcuna
evidenza che l’uso topico di questi farmaci sia pericoloso.
Preoccupazione riguardo a quest’azione dell’FDA è stata
manifestata dalla American Academy of Dermatology, Ame-
rican Academy of Allergy, Asthma and Immunology, Society
for Pediatric Dermatology, British Association of Derma-
tologists, European Society for Pediatric Dermatology, Ger-
man Society of Dermatology, Austrian Society of Dermato-
logy, European Dermatology Forum, European Academy
of Dermatology and Venereology, e altre società scientifiche,
nonché da associazioni di pazienti (National Society for Ato-
pic Eczema).
Durante gli ultimi 10 anni, entrambi i farmaci sono stati
studiati in maniera approfondita in numerosi trials clinici, e
la loro efficacia e sicurezza nella terapia della DA è stata
ampiamente dimostrata 1-5. In particolare, gli studi clinici
con pimecrolimus hanno coinvolto più di 19 000 pazienti
(di cui 7 600 bambini e 2 600 con meno di 2 anni di età). Inol-
tre, più di 5 milioni di pazienti, più della metà dei quali bam-
bini, sono stati trattati con pimecrolimus dopo la sua appro-
vazione, e studi di farmacovigilanza postmarketing hanno
dimostrato che non esiste alcuna evidenza clinica che sug-
gerisca un rischio aumentato di tumori nei pazienti trattati con
pimecrolimus. In particolare si è evidenziato che: a) l’inci-
denza di neoplasie nei pazienti trattati con pimecrolimus nel
corso degli studi clinici è inferiore a quella dei pazienti del
gruppo di controllo trattati con placebo o corticosteroidi ( 2
casi su 19 000 contro 5 casi su 4 000); b) i dati di sorve-
glianza postmarketing, aggiornati a Marzo 2005, riportano
un numero esiguo di neoplasie, inferiore a quello atteso nel-
la popolazione generale. Inoltre, in nessuno dei casi riportati
458 GIORNALE ITALIANO DI DERMATOLOGIA E VENEREOLOGIA
è stata stabilita una relazione causale con pimecrolimus; c)
i rari casi di linfoma osservati presentano un pattern istolo-
gico differente da quello che tipicamente caratterizza i linfo-
mi insorgenti nei pazienti sottoposti a terapie immunosop-
pressive.
L’assorbimento sistemico di entrambi i farmaci è assai
limitato 6, e anche in quei rari casi in cui le concentrazioni
ematiche dei farmaci siano rilevabili, sono molto basse e
assolutamente insufficienti a causare un’immunosoppres-
sione protratta. Questo si verifica anche nei bambini con
dermatite moderata/severa estesa a vaste aree corporee.
Non esiste alcuna evidenza di un potenziale fotocarcino-
genico o mutagenico del pimecrolimus negli animali. I linfo-
mi che tipicamente si associano alla immunosoppressione
sono stati descritti solo in animali esposti per tempi prolun-
gati a elevati livelli sistemici di inibitori della calcineurina,
livelli decine di volte più alti di quelli che sono mai stati
misurati nell’uomo dopo applicazione topica.
Non esiste alcuna evidenza che i pazienti trattati con pime-
crolimus abbiano un’incidenza più alta di infezioni sistemi-
che o cutanee, e non c’è alcuna evidenza che terapie anche
prolungate con pimecrolimus causino immunosoppressione
sistemica, come dimostrato dalle risposte anticorpali alle
vaccinazioni anti-infettive o dalle riposte ai test che misurano
l’ipersensibilità ritardata 7.
L’agenzia europea per i farmaci (European Medicines
Agency, EMEA) non ha preso alcun provvedimento imme-
diato, ma ha deciso di aprire una procedura di «referral» che
consiste in una revisione dettagliata dei dati clinici che con-
sentirà una valutazione accurata, basata su prove scientifiche
convincenti, del rapporto rischio/beneficio di questi farma-
ci 8. L’EMEA ritiene che tale procedura rappresenti il mez-
zo migliore e più efficace per consentire una valutazione
obiettiva in grado di fornire a medici e pazienti le regole per
un corretto utilizzo di questi farmaci. Nel frattempo le indi-
cazioni resteranno invariate.
La DA è una patologia cronica, ricorrente e a suo modo fru-
strante. In molti pazienti la malattia colpisce il volto, il col-
lo e altre sedi «sensibili» dove l’applicazione prolungata di
corticosteroidi non è appropriata. Questi pazienti necessi-
tano, quindi, di alternative agli steroidi topici. Noi siamo
preoccupati che l’allarme menzionato sia causa di confu-
sione e di preoccupazione impropria nei pazienti, nelle loro
famiglie e nel personale sanitario. Dal momento che è respon-
sabilità delle autorità sanitarie presentare le evidenze in
maniera giusta e corretta, riteniamo che le informazioni
attualmente presenti nei farmaci descrivono in maniera appro-
priata le loro modalità d’impiego e il profilo di sicurezza. Per-
tanto, reputiamo che le recenti raccomandazioni dell’FDA
non siano giustificate dall’evidenza scientifica e dovrebbe-
ro essere riconsiderate. Al fine di fornire ulteriore conferma
della sicurezza di pimecrolimus, sono stati iniziati studi a
lungo termine e un registro che include 4 000 pazienti pedia-
trici che saranno seguiti per un periodo di 10 anni.
Agosto 2005