VDE_275.

fm Page 45 Saturday, February 9, 2002 2:17 PM

Veterinary Dermatology 2002, 13, 45 49

sBlackwell Science Ltd

Intracorneal vacuoles in skin diseases with parakeratotic

hyperkeratosis in the dog: a retrospective light-microscopy
study of 111 cases (1973 2000)
DAVID A. SENTER,* DANNY W. SCOTT,* WILLIAM H. MILLER JR*
and HOLLIS N. ERB
Departments of *Clinical Sciences, Biomedical Sciences and Population Medicine and Diagnostic Sciences,
College of Veterinary Medicine, Cornell University, Ithaca, NY 14853-6401, USA
(Received 20 December 2000; accepted 25 March 2001)

Abstract Two recent case reports described a congenital keratinization defect (congenital follicular parakeratosis; CFP) in Rottweiler and Siberian Husky dogs. Skin biopsy specimens revealed marked parakeratosis
targeting the hair follicle and numerous intracorneal vacuoles. A retrospective histopathological study was
conducted on skin biopsy specimens from 111 dogs with diseases associated with parakeratotic hyperkeratosis
to determine whether intracorneal vacuoles were present. Additional criteria evaluated were the size and location
of the vacuoles and the degree of parakeratosis. Cases examined included dogs with primary idiopathic seborrhoea, necrolytic migratory erythema (NME), Malassezia dermatitis, zinc-responsive dermatosis, hereditary
nasal hyperkeratosis of Labrador Retriever dogs, thallotoxicosis and CFP. Thirty-seven cases (37/111, 33%) had
intracorneal vacuoles, including nine cases of primary idiopathic seborrhoea (9/29, 31%), 10 cases of NME (10/
18, 56%), five cases of Malassezia dermatitis (5/19, 26%), five cases of zinc-responsive dermatosis (5/36, 14%),
five cases of hereditary nasal hyperkeratosis (5/5, 100%) and three cases of CFP (3/3, 100%). If present, intracorneal vacuoles were found throughout all layers of the parakeratin. The sizes of intracorneal vacuoles varied
among diseases, but large (> 5 m) vacuoles only were present in CFP. Biopsies with a larger degree of parakeratosis were significantly more likely to have intracorneal vacuoles (P = < 0.001). Based on this study, intracorneal vacuoles are a common finding in many parakeratotic skin diseases of the dog, but large (> 5 m) vacuoles
are found only in CFP.

Keywords: dog, histopathology, intracorneal vacuoles, parakeratosis, skin.

INTRODUCTION

MATERIALS AND METHODS

Parakeratotic hyperkeratosis is a common, nonspecific

histopathological finding in many skin diseases of the
dog, yet there is a variety of skin disorders in which
parakeratosis is a consistent feature and an important
component of the diagnostic histopathological findings. Surface or diffuse (surface and follicular) parakeratosis are the most common presentations.1 Recently,
a hereditary disorder of cornification called congenital
follicular parakeratosis (CFP) was described in Rottweiler and Siberian Husky dogs, in which the parakeratosis was often confined to the hair follicle.2,3 An
interesting histopathological finding in this disorder
was the presence of numerous vacuoles within the parakeratotic layer.2,3 The purpose of this retrospective
light-microscopy study was to determine whether the
vacuoles are found exclusively in CFP or if they are
also a feature of other parakeratotic skin diseases.

A retrospective study of canine skin-biopsy specimens

submitted to the Dermatology Service at the Cornell
University College of Veterinary Medicine from 1973
to 2000 was performed. All skin biopsy specimens with
a final diagnosis of necrolytic migratory erythema
(NME), primary idiopathic seborrhoea, Malassezia
dermatitis, zinc-responsive dermatosis, thallotoxicosis,
hereditary nasal hyperkeratosis of Labrador Retriever
dogs or CFP were reviewed. A definitive diagnosis
was based on the history, clinical signs, histopathological findings and, where appropriate, other laboratory
test results and response to treatment.1 Slides were
included in the study only if there were at least two
layers of parakeratotic hyperkeratosis on the surface or
follicular epithelium. A total of 111 individual cases
was included in the study.
The biopsy samples had been taken with 46 mm
biopsy punches or wedge sections, fixed in 10% neutralbuffered formalin, routinely processed and stained
with haematoxylin and eosin (H&E). If there was more
than one section on the slide, the section with the most
parakeratosis was selected. All slides were reviewed by

Correspondence: Danny W. Scott, Department of Clinical Sciences,

College of Veterinary Medicine, Cornell University, Ithaca,
NY 14853-6401 USA. Tel.: +1 607 253 3029; Fax: +1 607 253 3534;
E-mail: shb3@cornell.edu
2002 Blackwell Science Ltd

45

VDE_275.fm Page 46 Saturday, February 9, 2002 2:17 PM

46

D. A. Senter et al.

one of the authors (DAS) and several observations

were made: presence of intracorneal vacuoles, location
and relative sizes of the vacuoles, and degree of parakeratosis. Location was assessed by determining if
the vacuoles were present throughout all layers of the
parakeratin. Also, in slides with parakeratosis of the
surface and follicular epithelium, it was noted whether
the vacuoles were located in both the follicular and the
surface parakeratin. The largest vacuoles in each section
were measured with a micrometer, and the size was
recorded as small (< 2.5 m diameter), medium (2.5
5.0 m diameter) and large (> 5.0 m diameter). The
degree of parakeratosis was assessed by counting the
individual layers, and recorded as mild (25 layers),
moderate (610 layers), marked (1120 layers) and
severe (> 20 layers). In addition, biopsy specimens
from three dogs with CFP were stained with Sudan
black IV4 to document the presence of lipid within the
vacuoles. Biopsy specimens from two of the three dogs
had been formalin-fixed, routinely processed, and
paraffin embedded. Biopsy specimens from one dog
were fixed in 10% neutral-buffered formalin, but not
routinely processed prior to cryostat sectioning and
Sudan black staining.
Statistical evaluation
Exact binomial 95% confidence intervals (CI) were calculated ( 6 v 6.04b, 1997, Centers for Disease
Control and Prevention) for the percentage of slides
with vacuoles.5 Because the size of the vacuoles and the
degree of parakeratosis were ordinal data, the median
was calculated for each disease and for all diseases,
together.6 The Wilcoxon rank-sum test7 was used to
test the relationship between the degree of parakeratosis and the presence of vacuoles. This was interpreted
one-sided, and a one-tailed P-value of < 0.05 was considered statistically significant. The calculation of
medians and the Wilcoxon rank-sum tests were performed using 7 (Analytical Software, Tallahassee, FL, USA).

RESULTS
Of the 111 slides reviewed, 37 (37/111, 33%, CI = 25,
43) had intracorneal vacuoles (Table 1). Excluding the
three specimens with the diagnosis of CFP, 34 (34/108,
32%, CI = 23, 41) had vacuoles. Vacuoles were present
in nine cases of primary seborrhoea (9/29, 31%,
CI = 15, 51), 10 cases of NME (10/18, 56%, CI = 31,
78), five cases of Malassezia dermatitis (5/19, 26%,
CI = 9, 51) and five cases of zinc-responsive dermatosis
(5/36, 14%, CI = 5, 29). All cases of hereditary nasal
hyperkeratosis of Labrador Retriever dogs (5/5, 100%,
CI = 48, 100) and CFP (3/3, 100%, CI = 29, 100) had
vacuoles, but none were present in the one case of
thallotoxicosis (0/1, 0%, CI = 0, 98).
If intracorneal vacuoles were present, they were
found throughout all layers of the parakeratin. In all
but two biopsies from diseases characterized by
2002 Blackwell Science Ltd, Veterinary Dermatology, 13, 45 49

Table 1. The presence of intracorneal vacuoles in parakeratotic skin

diseases of dogs (1973 2000, Cornell University)
Number of cases
With vacuoles
Diagnosis
Primary idiopathic seborrhoea
Necrolytic migratory erythema
Malassezia dermatitis
Zinc-responsive dermatosis
Hereditary nasal hyperkeratosis
of the Labrador Retriever
Thallotoxicosis
Congenital follicular parakeratosis
Total

Total

No.

29
18
19
36
5

9
10
5
5
5

31
56
26
14
100

1
3
111

0
3
37

0
100
33

Figure 1. Skin biopsy specimen from the hock of a dog with

necrolytic migratory erythema. Note the multiple small vacuoles
(arrow) within the parakeratotic layer (H&E, 900).

parakeratosis of the surface and follicular epithelium,

the vacuoles were present in both the surface and
follicular parakeratin (5/5 cases of zinc-responsive dermatosis, 8/10 cases of NME). The two cases of NME
could not be evaluated fully because there were no hair
follicles present in the sections reviewed.
When sections were examined at 400, it was noted
that individual slides had a variety of different sized
intracorneal vacuoles. In all instances, if medium or large
vacuoles were observed, then vacuoles of smaller sizes also
were present. Each section examined was categorized
on the basis of the largest intracorneal vacuoles found.
The intracorneal vacuoles were small (< 2.5 m) in
all biopsies from Malassezia dermatitis and zincresponsive dermatosis (table available upon request,
Fig. 1). The median size of vacuoles in primary seborrhoea, NME, hereditary nasal hyperkeratosis and all
diseases combined was medium (2.55.0 m, Fig. 2).
Large vacuoles were observed only in the three cases of
CFP (Fig. 2).

VDE_275.fm Page 47 Saturday, February 9, 2002 2:17 PM

Intracorneal vacuoles in parakeratosis

47

DISCUSSION

Figure 2. Skin biopsy specimen from the lateral thorax of a dog with
congenital follicular parakeratosis. The large (long arrow), medium
(short arrow), and small vacuoles are present within the parakeratin
(H&E, 900).

Figure 3. Skin biopsy specimen from the lateral thorax of a dog with
congenital follicular parakeratosis. Note the red-stained intracorneal lipid vacuoles (arrow) (Sudan black IV, 360).

The median degree of parakeratosis (table available

upon request) with or without intracorneal vacuoles
was moderate for primary seborrhoea and Malassezia
dermatitis, marked for zinc-responsive dermatosis and
thallotoxicosis, severe for NME, hereditary nasal
hyperkeratosis and CFP, and marked for all diseases
combined. The degree of parakeratosis was more
prominent in all biopsies with vacuoles than in all
biopsies without vacuoles (P < 0.001). This trend was
significant in primary seborrhoea (P < 0.001), NME
(P = 0.04) and Malassezia dermatitis (P = 0.02), but
not significant in zinc-responsive dermatosis
(P = 0.10).
Of the three biopsy specimens that had been stained
with Sudan black IV, the only specimen that revealed
sudanophilic material (lipid) within the intracorneal
vacuoles was the one that had not been routinely processed prior to cryostat sectioning and staining (Fig. 3).

Parakeratotic hyperkeratosis implies altered epidermopoiesis secondary to inflammatory, hormonal,

developmental or neoplastic causes.1 Although parakeratosis is only one histopathological finding, the location of the parakeratosis can be helpful in establishing
a diagnosis.1 Of the diseases reviewed in this study,
parakeratosis of the surface and follicular epithelium
is suggestive of zinc-responsive dermatosis,8 NME9 and
thallotoxicosis.10 Surface and infundibular parakeratosis is a characteristic of Malassezia dermatitis,11 but
only surface parakeratosis is seen in hereditary nasal
hyperkeratosis of Labrador Retriever dogs,12 as hair
follicles are not found in the planum nasale. Findings
consistent with primary idiopathic seborrhoea are a
mixed surface orthokeratotic and parakeratotic hyperkeratosis with focal areas of parakeratosis (parakeratotic caps) overlying oedematous dermal papillae.13
Prominent parakeratosis targeting the entire follicular
epithelium is a unique feature of CFP.2,3
Another striking feature of CFP is the presence of
vacuoles within the parakeratotic keratin.2,3 In a recent
case report by Lewis et al.,2 these vacuoles were evaluated using transmission electron microscopy, and were
determined to contain lipid. The lipid nature of the
vacuoles was confirmed by Sudan black staining in one
of our cases of CFP. Sudan black staining was negative in our other two cases of CFP, but the lipid was
most likely lost during the processing and paraffinembedding procedures.14 Structural studies of the epidermis in normal dogs and cattle have demonstrated the
presence of sudanophilic material in the intercellular
spaces of the outer layers of the stratum corneum that
was assumed to be sebaceous lipid.15,16 However, nonsudanophilic globules also were found within the lipid,
and might have been sweat gland secretions. Therefore,
without electron microscopy or lipid stains, we cannot
definitively determine whether the vacuoles found in
the other parakeratotic diseases in this study truly
contain lipid. It was not the aim of our retrospective
light-microscopy study to determine the nature of the
vacuoles, only to document the significance of their
occurrence and size.
The significance of the lipid vacuoles in CFP is
unknown. Intracorneal lipid vacuoles have been
reported in lamellar ichthyosis in a llama17 as well as
in various forms of ichthyosis in man;18 a defect in epidermal lipid metabolism has been implicated.
Because routine processing and embedding techniques are known to result in the loss of many layers
of the outer stratum corneum,15,16 the measurements of
the degree of parakeratosis are inexact. Nevertheless,
parakeratosis was found to be most prominent in
NME, hereditary nasal hyperkeratosis and CFP. Interestingly, these were also the diseases that were most
likely to have intracorneal vacuoles. A significant relationship between the degree of parakeratosis and the
presence of vacuoles was documented, but it is unclear
why this occurs, or why vacuoles are present in some
2002 Blackwell Science Ltd, Veterinary Dermatology, 13, 4549

VDE_275.fm Page 48 Saturday, February 9, 2002 2:17 PM

48

D. A. Senter et al.

biopsies and not in others. Perhaps, the vacuoles are a

result of increased and/or abnormal epithelial lipogenesis
secondary to the altered epidermopoiesis associated with
parakeratosis. Further studies are needed to ascertain
the contents of the vacuoles in parakeratotic diseases
other than CFP in order to discern the pathogenesis.
In conclusion, intracorneal vacuoles are a common
finding in many parakeratotic skin diseases other than
CFP in the dog, and are more likely to be found in
biopsy specimens with marked to severe parakeratosis.
The presence of intracorneal vacuoles is a nonspecific
finding, but may be used to further substantiate a
histopathologic diagnosis. Zinc-responsive dermatosis
was the least likely of the commonly seen parakeratotic
skin diseases to have intracorneal vacuoles, whereas
vacuoles were seen in all cases of CFP and hereditary
nasal hyperkeratosis. Furthermore, CFP was the only
disease in which vacuoles > 5 m in diameter were
observed.

REFERENCES
1. Scott, D.W., Miller, W.H. Jr., Griffin, C.E., eds. Diagnostic methods. In: Small Animal Dermatology, 6th edn.
Philadelphia: W.B. Saunders, 2000: 71206.
2. Lewis, D.T., Messinger, L.M., Ginn, P.E. Multiple
congenital defects in five rottweiler dogs. Veterinary
Dermatology 1998; 9: 6172.
3. Scott, D.W., Miller, W.H. Jr. Congenital follicular parakeratosis in a rottweiler and a Siberian husky. Canine
Practice 2000; 25: 16 19.
4. Sheehan, D.C., Hropchak, B.B. Theory and Practice of
Histotechnology, 2nd edn. St. Louis, MO: C.V. Mosby,
1980: 202 13.
5. Dawson-Saunders, B., Trapp, R.G. Basic and Clinical
Biostatistics, 2nd edn. Norwalk, CT: Appleton and
Lange, 1994: 144.
6. Dawson-Saunders, B., Trapp, R.G. Basic and Clinical

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

Biostatistics, 2nd edn. Norwalk, CT: Appleton and

Lange, 1994: 43.
Dawson-Saunders, B., Trapp, R.G. Basic and Clinical
Biostatistics, 2nd edn. Norwalk, CT: Appleton & Lange,
1994: 11719.
Scott, D.W., Miller, W.H. Jr., Griffin, C.E., eds. Nutritional skin diseases. In: Small Animal Dermatology, 6th edn.
Philadelphia: W.B. Saunders, 2000: 111224.
Scott, D.W., Miller, W.H. Jr., Griffin, C.E., eds. Endocrine
and metabolic diseases. In: Small Animal Dermatology,
6th edn. Philadelphia: W.B. Saunders, 2000: 780886.
Scott, D.W., Miller, W.H. Jr., Griffin, C.E., eds. Environmental skin diseases. In: Small Animal Dermatology,
6th edn. Philadelphia: W.B. Saunders, 2000: 1073111.
Mauldin, E.A., Scott, D.W., Miller, W.H. Jr. et al. Malassezia dermatitis in the dog: a retrospective histopathological and immunological study of 86 cases (199095).
Veterinary Dermatology 1997; 8: 191202.
Pag, N., Paradis, N., Lapointe, J.M., et al. Hereditary
nasal hyperkeratosis in Labrador retrievers. Proceedings
of the American Academy of Veterinary Dermatology and
American College of Veterinary Dermatology 1991; 15:
412.
Scott, D.W., Miller, W.H. Jr., Griffin, C.E., eds. Congenital
and hereditary defects. In: Small Animal Dermatology,
6th edn. Philadelphia: W.B. Saunders, 2000: 9131004.
Logan, M.E., Zaim, M.T. Histologic stains in dermatopathology. Journal of American Academy of Dermatology
1990; 22: 82030.
Lloyd, D.H., Garthwaite, G. Epidermal structure and
surface topography of canine skin. Research in Veterinary
Science 1982; 33: 99104.
Lloyd, D.H., Dick, W.D.B., Jenkinson, D.M. Structure
of the epidermis in Ayrshire bullocks. Research in Veterinary Science 1979; 26: 1729.
Belknap, E.B., Dunstan, R.W. Congenital ichthyosis in
a llama. Journal of the American Veterinary Medical
Association 1990; 197: 7647.
Busman, M.M., Goodkin, P.E., Fahrenback, W.H. et al.
Harlequin ichthyosis with epidermal lipid abnormality.
Archives in Dermatology 1979; 115: 18993.

Rsum Deux rcents cas cliniques ont dcrit un trouble congnital de la kratinisation (parakratose folliculaire congnitale, CFP) chez le Rottweiler et le Siberian Husky. Les biopsies cutanes de ces cas ont montr une
parakratose marque ciblant les follicules pileux, avec de nombreuses vacuoles intracornes. Cette tude histopathologique rtrospective a t ralise sur 111 chiens prsentant une hyperkratose parakratosique pour
dterminer si des vacuoles intracornes taient prsentes dans des dermatoses varies. La taille et la position des
vacuoles, le degr de parakratose ont galement t analyss. Les cas regroupaient des chiens atteints de
sborrhe primaire idiopathique, drythme ncrolytique migrant (NME), de dermatite Malassezia, de dermatose rpondant au zinc, dhyperkratose nasale idiopathique du Labrador Retriever, de thallotoxicose et de
CFP. Dans trente-sept cas (37/111, 33%) des vacuoles taient prsentes, parmi lesquels neuf cas de sborrhe primaire idiopathique (9/29, 31%), 10 cas de NME (10/18, 56%), cinq cas de dermatite Malassezia (5/19, 26%),
cinq cas de dermatose rpondant au zinc (5/36, 14%), cinq cas dhyperkratose nasale idiopathique (5/5, 100%)
et trois cas de CFP (3/3, 100%). Lorsquelles taient prsentes, les vacuoles intracornes taient observes dans
toutes les couches parakratosiques. La taille des vacuoles intracornes tait variable en fonction des maladies,
mais la prsence de vacuoles de grande taille (> 5 m) na t observe que pour les cas de CFP. Les biopsies
prsentant une parakratose marque avaient significativement plus de chance de prsenter des vacuoles
intracornes (P = < 0.001). Cette tude montre que les vacuoles intracornes sont frquemment rencontres dans
diverses dermatoses parakratosiques chez le chien, mais des vacuoles de grande taille (> 5 m) ne sont rencontres que dans la CFP.

2002 Blackwell Science Ltd, Veterinary Dermatology, 13, 45 49

VDE_275.fm Page 49 Saturday, February 9, 2002 2:17 PM

Intracorneal vacuoles in parakeratosis

49

Resumen Dos casos clnicos describieron un defecto congnito de la queratinizacin (paraqueratosis congnita
folicular; CFP) en perros Rottweiler y Husky Siberiano. Las biopsias cutneas mostraron paraqueratosis intensa
afectando el folculo piloso y numerosas vacuolas intracorneales. Se realiz un estudio histopatolgico retrospectivo con biopsias cutneas de 111 perros con enfermedades asociadas a la hiperqueratosis paraquerattica,
con el fin de examinar la presencia de vacuolas intracorneales. Se evaluaron otros criterios como el tamao y la
localizacin de las vacuolas, as como el grado de paraqueratosis. Los casos examinados incluan perros con
seborrea idioptica primaria, eritema necroltico migratorio (NME), dermatitis por Malassezia, dermatosis con
respuesta al zinc, hiperqueratosis nasal hereditaria del Labrador Retriever, talotoxicosis y CFP. Treinta y siete
casos (37/111, 33%) mostraron vacuolas intracorneales, incluyendo nueve casos de seborrea idioptica primaria
(9/29, 31%), 10 casos de NME (10/18, 56%), cinco casos de dermatitis por Malassezia (5/19, 26%), cinco casos
de dermatosis con respuesta al zinc (5/36, 14%), cinco casos de hiperqueratosis nasal hereditaria (5/5, 100%) y
tres casos de CFP (3/3, 100%). Cuando se encontraban presentes, las vacuolas intracorneales de encontraban en
todas las capas de paraqueratina. El tamao de las vacuolas intracorneales variaba entre enfermedades, pero las
vacuolas grandes (> 5 m) slo se hallaban en CFP. La posibilidad de tener vacuolas intracorneales era significativamente mayor en biopsias con un mayor grado de paraqueratosis (P = < 0.001). Basndonos en este estudio,
las vacuolas intracorneales son un hallazgo habitual en muchas enfermedades cutneas con paraqueratosis en
el perro, pero las vacuolas grandes (> 5 m) se encuentran slo en la CFP.
Zusammenfassung Zwei krzlich verffentlichte Fallberichte beschrieben einen kongenitalen Keratinisationsdefekt (kongenitale follikulre Parakeratose, KFP) beim Rottweiler und Sibirischem Huskie. Hautbiopsieproben zeigten deutliche Parakeratose der Haarfollikel und zahlreiche intrakorneale Vakuolen. Eine retrospective
histopathologische Studie wurde an Hautbiopsieproben von 111 Hunden mit Erkrankungen, die mit Parakeratose verbunden waren, durchgefhrt, um festzustellen, ob intranukleare Vakuolen vorhanden waren. Zustzliche
bewertete Kriterien waren die Grsse und Lokalisierung der Vakuolen und der Grad der Parakeratose. Untersuchte Flle schlossen Hunde mit primrer idiopathischen Seborrhoe, nekrolytischem migratorischem Erythem
(NME), Malassezia Dermatitis, auf Zink ansprechende Dermatose, erbliche nasale Hyperkeratose des Labrador
Retrievers, Thalliumintoxikation und KFP ein. Siebenunddreissig Flle (37/111, 33%) hatten intrakorneale
Vakuolen, neun Flle mit primrer idiopathischer Seborrhoe (9/29, 31%), 10 Flle mit NME (10/18, 56%), fnf
Flle mit Malassezia dermatitis (5/19, 26%), fnf Flle mit auf Zink ansprechende Dermatose (5/36, 14%), fnf
Flle der erblichen, nasalen Hyperkeratose (5/5, 100%) and drei Flle mit CFP (3/3, 100%). Wenn vorhanden,
wurden die intrakornealen Vakuolen in allen Schichten des Parakeratins gefunden. Die Grsse dieser intrakornealen Vakuolen variierte mit der Erkrankung, aber grosse (> 5 m) Vakuolen waren nur bei KFP vorhanden.
Biopsien mit einer ausgeprgteren Parakeratose hatten mit grsserer Wahrscheinlichkeit intrakorneale Vakuolen
(P = < 0.001). In dieser Studie waren intrakorneale Vakuolen ein hufiger Befund bei vielen parakeratotischen
Hauterkrankungen des Hundes, aber grosse (> 5 m) Vakuolen wurden nur bei KFP gefunden.

2002 Blackwell Science Ltd, Veterinary Dermatology, 13, 4549

VDE_275.fm Page 50 Saturday, February 9, 2002 2:17 PM