CME

The Pathogenesis of Hemangiomas: A Review

Constantijn G. Bauland, M.D.
Learning Objectives: After reading this article, the participant should be able
Maurice A. M. van Steensel, to: 1. Differentiate between hemangiomas and vascular malformations. 2. De-
M.D., Ph.D. scribe arguments for the trophoblast origin of hemangiomas. 3. Give arguments
Peter M. Steijlen, M.D., Ph.D. for the angioblast theory for the origin of hemangiomas. 4. Identify key genes
Paul N. M. A. Rieu, M.D., involved in the origin of hemangiomas.
Ph.D. Background: Hemangiomas of infancy are common endothelial tumors. They
Paul H. M. Spauwen, differ from vascular malformations in their tissue architecture and biological
M.D., Ph.D. properties. To date, there is no universally accepted theory that explains the
Nijmegen and Maastricht, pathogenesis and pathophysiology of hemangiomas.
The Netherlands Methods: Theories from the medical literature from 1981 to 2004 were gath-
ered, categorized, and reviewed.
Results: Current research is mostly on the cellular and genetic levels. The most
authoritative theories focus on angioblast origins, trophoblast origins, mutations
in cytokine regulatory pathways, and field defects as the cause of the deranged
angiogenesis of hemangiomas.
Conclusions: To date, no single theory can easily explain all the characteristics
of hemangiomas, such as predilection for the female sex, usual occurrence after
birth, spontaneous involution, abnormal tissue architecture, and distribution
within a developmental field. Hemangiomas are probably the final common
expression of several pathophysiological mechanisms taking effect alone or in
combination. (Plast. Reconstr. Surg. 117: 29e, 2006.)

H
emangiomas are endothelial neoplasms,
distinct from vascular malformations.
They affect approximately one in 10 to 20
Caucasian children, with a 3:1 predilection for
the female sex. 1– 4 The incidence in non-
Caucasian infants is lower.5 At birth, a precursor
lesion may be present as an erythematous patch
or a telangiectasia. Within weeks the lesion ex-
pands rapidly. A hemangioma usually occurs as a
solitary, superficial lesion in the head and neck
area, but it can be present anywhere in or on the
body. Size can vary greatly, from nodular lesions
several millimeters in diameter to plaque-like
tumors covering the entire face or a quadrant of
the body. Typically, there is fast growth during
the first trimesters of life and a gradual involu-
tion in the following years.
Vascular malformations differ from hemangi-
omas in several crucial ways. Most distinctively,
vascular malformations are present at birth,
From the Departments of Plastic Surgery and Pediatric Sur-
gery, Radboud University Nijmegen Medical Center, and the
Department of Dermatology, University Hospital Maastricht.
Received for publication November 2, 2004; revised May 17,
2005.
Copyright ©2006 by the American Society of Plastic Surgeons
DOI: 10.1097/01.prs.0000197134.72984.cb
grow commensurately with the body, and never
show signs of spontaneous involution.6
The dramatic appearance and fast growth of a
hemangioma may seem alarming, but usually no
treatment is indicated. In selected cases, how-
ever, prompt intervention is needed. These cases
involve hemangiomas that ulcerate, obstruct the
airway, or cause visual deficits or cardiovascular
symptoms. Current treatment modalities include
intralesional or systemic corticosteroid treat-
ment, systemic interferon treatment, local bleo-
mycin treatment, and surgery.7,8
The earliest theories stated that maternal behavior
or thoughts during pregnancy were the cause of
hemangiomas and vascular malformations. The word
“birthmark” itself implies a connection between birth
and a vascular lesion and the term “strawberry nevus”
indicates that maternal intake of red fruits was once
thought to cause birthmarks.9 Viral origins have been
considered and rejected.10,11 Current theories focus
on progenitor cells, developmental field defects, pla-
cental involvement, derangement of angiogenesis,
and mutations in the cytokine regulatory pathway.
There is no universally accepted theory that explains
the cause and biological behavior of hemangiomas.
PATHOPHYSIOLOGY
The histology of an early lesion is character-
ized by cellular masses of endothelium. At first, the

www.plasreconsurg.org 29e
 Plastic and Reconstructive Surgery • February 2006
endothelium contains no lumina, but gradually a
vascular architecture develops.6,12–14 Mast cells are
numerous.15,16
After 6 to 10 months, the growth curve flattens.
At this stage, histological examination reveals a
diminishing endothelial cellularity and an in-
crease in fibrofatty tissue. Eosinophil granulocytes
invade the lesion. The vascular lumina increase in
size, and the endothelial lining becomes thinner.
Involution coincides with increased apoptosis of
endothelial and stromal cells.17–19 Proliferation
and involution are not mutually exclusive; they
can both be present in the same hemangioma at
the same time.12,19,20
Hemangiomas have intrinsic hormonal as-
pects. For example, the serum level of estradiol-
17␤ is significantly higher in patients with hem-
angiomas compared with patients with other
vascular tumors. The cytosol estrogen receptor
level in hemangiomas is also markedly higher than
it is in normal tissue.21 The clinical response to
treatment with steroids is paralleled by a decrease
in the serum level of estradiol-17␤.21
At the age of 5 years, 50 percent of the hem-
angiomas have involuted, and at the age of 9 years
90 percent have.5 Particularly in superficial lo-
cated hemangiomas, a fibrofatty residue, soft-tis-
sue surplus, erythema, teleangiectasias, or pig-
mentation imperfection can remain.
CURRENT THEORIES ON THE ORIGIN
OF HEMANGIOMAS
To date, there is no universally accepted
model for the etiology of hemangiomas. There is
general consensus on the importance of endothe-
lial cells, but the source of the endothelial cells is
speculative and there is no consensus on the pos-
sible mechanisms by which the hemangioma en-
dothelium interacts with surrounding cells. There
is a scientific debate on whether they are of mu-
tational or placental origin or fit in a developmen-
tal field disruption.13,22–29
NONENDOTHELIAL CELLS IN
HEMANGIOMA FORMATION
Hemangiomas contain a large population of
nonendothelial cells. Stromal cells extracted from
hemangiomas can release vascular endothelial
growth factor (VEGF). When grafted to nude
mice, these stromal cells elicit an angiogenic
response.22 A defined population of dendritic cells
with an intimate relationship with hemangioma
vessels has been identified. Immunochemically,
these cells are closely related to the monocyte/
macrophage or dendritic cell systems.30 These
30e
dendritic cells may possibly have a role in hem-
angioma formation through cytokine secretion.30
In proliferating hemangiomas, increased lev-
els of mast cells are seen aligned along hemangi-
oma vessels.13 In involuting hemangiomas, in-
creased levels of mast cells are also reported.14,15 In
the proliferative phase, mast cells express fibro-
blast growth factor 2 (FGF2), a potent angiogenic
polypeptide.15 In addition, mast cells can rapidly
release VEGF and can sustain this release.16 Mast
cells probably play a complex role involving stim-
ulation of angiogenesis in the proliferative phase
and inhibition of angiogenesis in the involuting
phase.15,16 Their exact role has not been eluci-
dated.
ANGIOBLAST THEORY
A classic hypothesis is that hemangiomas de-
velop from embryonic sequestrations of omnipo-
tent angioblastic cells. According to this theory,
angioblastic cells form an anlage that fails to con-
nect to the normal vascular system and develops
into a hemangioma.31 Alternatively, a hemangi-
oma could be a tumor of a primitive cell type
capable of differentiating into different cell types.
This could account for the heterogeneity of hem-
angioma cellular elements with diverse antigenic
profiles.32
LYVE-1 is a specific marker for lymphatic ves-
sels. CD34 is a marker for vascular endothelium.
Given the coexpression of these markers in the
endothelium of proliferating hemangiomasm it
has been proposed that proliferating hemangio-
mas are arrested in an early developmental stage
of vascular differentiation.33
The AC133/CD133 antigen is a marker for
progenitor cells. The KDR VEGF receptor is a
marker for endothelial cells. Coexpression of
these markers in endothelial cells of growing hem-
angiomas suggests that endothelial progenitor
cells may play a role in the pathogenesis of
hemangiomas.28
The CD14 antigen is a marker for monocytes.
The CD83 antigen is a marker for dendritic cells,
differentiated from monocytes. Coexpression of
these antigens in endothelial cells of a growing
hemangioma suggests that hemangioma endothe-
lium might be of monocytic origin.27
MANIFESTATION OF A
DEVELOPMENTAL FIELD DEFECT
A developmental field is a region or part of an
embryo that responds as a coordinated unit to
embryonic induction and results in complex or
multiple anatomic structures.34 A field defect is the
Volume 117, Number 2 • Pathogenesis of Hemangiomas
result of a disturbance of the development of a
morphogenetic field. One of the essential at-
tributes of the developmental field for this theory
is heterogeneity. This means that different insults
during development can result in identical mal-
formations, the variability of which depends on
the timing of the insult.
The human embryo as it exists before orga-
nogenesis starts is considered the primary field.
The organs and patterns superimposed on the
primary developmental field are secondary fields.
Primary field defects affect a range of tissues, de-
pending on the morphogenetic process that is
affected. If only one organ system is affected, the
defect is considered monotopic; if two or more
organ systems are affected, the defect is polytopic.
Defects that affect organogenesis are secondary
and can be either monotopic or polytopic.
The clinical finding that hemangiomas are
sometimes grouped within a developmental field
and can be associated with various developmental
defects supports the hypothesis that a hemangi-
oma is a manifestation of a field defect.25 For in-
stance, the distribution of facial hemangiomas is
nonrandom. There are striking segmental pat-
terns for facial hemangiomas, and there is a pre-
dilection for regions of embryological fusion.31,35
Cervicofacial hemangiomas in the beard area are
associated with another hemangioma located in
the upper airway or subglottis.35,36
Sacral hemangiomas are associated with an
array of multiple sacral and genitourinary anom-
alies, including imperforate anus, renal anoma-
lies, genital anomalies, lipomenigomyelocele,
tethered spinal cords, and bony deformity of the
pelvis.25,37 A syndrome has not yet been claimed.
Lumbar hemangiomas are associated with a teth-
ered spinal cord.38
Facial hemangiomas have been associated
with sternal defects, a supraumbilical raphe, and
malformations of the aorta, the carotid, and ver-
tebral arteries.39 – 42
A clear example of hemangiomas embedded
in a field disruption is the PHACE syndrome (pos-
terior fossa brain malformation, hemangioma, ar-
terial anomalies, coarctation of the aorta and car-
diac defects, and eye abnormalities).43 It is
characterized by a plaque-like, flat facial heman-
gioma, sternal defects, malformation of cranial
arteries, and a posterior cerebral fossa defect.43– 45
This syndrome may be understood as a primary
polytopic field defect. It is conceivable that a dis-
turbance of morphogenesis, which occurs during
the development of the heart, may also affect
structures that are adjacent in embryogenesis. The
facial vasculature can then be influenced by what-
ever process disturbs midline closure, cardiac mor-
phogenesis, and neural tube patterning.43 The
phenotype shows considerable variation, but that
is consistent with this theory.
PLACENTAL ORIGIN
The expression of the erythrocyte-type glucose
transporter molecule GLUT1 is limited to only a
few tissue types: the microvascular endothelia of
blood tissue barriers of the central nervous system
and the placental trophoblast. It is not present in
the vasculature of normal skin and subcutis, but it
is highly expressed in the capillary endothelium of
hemangiomas. No other benign vascular tumors,
reactive vascular proliferations, or vascular mal-
formations show GLUT1 expression.46 The high
expression of GLUT1 in hemangiomas is a uni-
versal trait of hemangiomas, irrespective of mitotic
activity, phase, or anatomical location. As such, it
is probably an intrinsic feature of hemangioma
endothelium. The expression of other placenta-
associated vascular antigens, such as the Lewis Y
antigen, has also been noticed in hemangiomas.24
Several other key placental antigens could not be
demonstrated in hemangioma tissue,47 but the
sharing of several immunological markers has
prompted the suggestion that hemangiomas may
originate from embolization of placental cells.
In favor of a placental origin for hemangiomas
is the finding that chorionic villus sampling is as-
sociated with a three- to fourfold increased inci-
dence of hemangiomas in children born to
women who underwent this procedure.48,49 In cho-
rionic villus sampling, a fetal trophoblast sample
is surgically removed from the placenta. During
this procedure, placental embolization and ma-
ternofetal transfusion occurs, replacing up to 40
percent of the fetal plasma.50,51 Amniocentesis
does not involve a placental disruption and does
not influence the risk for hemangiomas.48,49
It is open to debate whether placental embo-
lization or an aberrant differentiation through a
placental developmental pathway causes the bio-
chemical similarities between hemangioma and
placenta.
DERANGEMENT OF ANGIOGENESIS
Folkman52 proposed that hemangiomas are
caused by unregulated angiogenesis or an imbal-
ance between angiogenic and angiostatic factors.
It is without dispute that angiogenesis is an im-
portant factor in hemangiogenesis, and much in-
vestigational effort is being put into research on
the mechanism of angiogenesis in hemangiomas.
31e
 Plastic and Reconstructive Surgery • February 2006
Key growth factors in angiogenesis include
VEGF and basic fibroblast growth factor (bFGF).
VEGF and bFGF play pivotal roles in physiological
angiogenesis, such as embryonic development,
growth, and differentiation, wound healing, and
placentation.53 Messenger RNA for FGF and VEGF
are upregulated in proliferative hemangiomas.54
E-selection is an endothelial-cell-specific leuko-
cyte adhesion molecule with a role in angiogenesis.
The endothelium of proliferating hemangiomas
shows a significantly increased expression of E-se-
lectin compared with involuting hemangiomas.55
Tissue inhibitor of metalloproteinase 1 is an
inhibitor of angiogenesis. It is expressed in invo-
luting hemangiomas but not in proliferating
hemangiomas.14 It has been suggested that angio-
genic factors are sequestered onto the extracellu-
lar matrix and are released by the action of en-
dothelial-derived proteases. These growth factors
could then initiate a cascade leading to a heman-
gioma. Tissue inhibitor of metalloproteinase 1
would stop this cascade by inhibiting the
proteases.56
TIE2 is an endothelium-specific receptor ty-
rosine kinase for angiopoietin 1, stimulating an-
giogenesis. Angiopoietin 2 also targets TIE2, but
its effects are complex. It has an antagonistic effect
to angiopoietin 1, but it can exercise an angio-
genic effect in the presence of VEGF. In heman-
gioma-derived endothelial cells, TIE2 expression
is increased and the response to angiopoietin-1 is
enhanced. The expression of angiopoietin 2 has
been shown to be dysregulated.57
Insulin-like growth factor 2 is highly expressed
in proliferating hemangiomas. It is a known mi-
togen and suppressor of apoptosis. Its exact role in
hemangioma formation is unclear, but it may play
a role in the expression of VEGF and GLUT1.58
Cytokine level alterations are also found in the
normal tissues surrounding a hemangioma. The
epidermis overlying a growing hemangioma ex-
presses increased levels of the angiogenetic cyto-
kines bFGF and VEGF. The skin over an involuting
hemangioma expresses normal levels of bFGF and
VEGF and increased levels of the angiostatic cy-
tokine interferon ␤.26
MUTATION IN CYTOKINE
REGULATORY PATHWAY
A twin study showed no indications for hered-
ity of the common hemangioma.59 However, rare
familial occurrence of hemangiomas does exist
and is suggestive of an autosomal dominant in-
heritance with moderate to high penetration.60 In
three out of five cases of familial hemangiomas,
32e
linkage to a fragment of the 5q chromosome could
be established. The area concerned contains three
genes involved in vessel growth: PDGFR␤, which
codes for platelet-derived growth factor receptor
␤, FGFR4, which codes for fibroblast growth factor
receptor 4, and FLT4 (VEGFR3), which codes for
VEGF receptor 3.61
Loss of heterozygosity is usually the conse-
quence of postzygotic mutational events. In some
sporadic hemangiomas, a significant loss of het-
erozygosity occurs on a fragment of the 5q
chromosome.23,29,62
Several other genes and their protein products
are known to be involved in vascular proliferation,
including inducers, such as FGFs, VEGF, and an-
giopoetin 1, and inhibitors, such as angiostatin,
endostatin, and thrombospondin. It has been sug-
gested that hemangioma proliferation is caused by
aberrant upregulation of VEGF and bFGF
pathways.54 Mutations in these genes could con-
ceivably produce the abnormal vessel growth seen
in hemangioma. The presence of a tumor sup-
pression gene for hemangiomas on chromosome
5q was suggested, but a candidate gene was not
proposed.62
Alternatively, mutations in upstream or down-
stream targets of these genes could produce the
same effect. Involvement of the VEGF pathway is
also suggested by the finding of mutations in the
genes for VEGF receptors VEGFR2 and VEGFR3
in two out 15 hemangiomas studied.23
DISCUSSION
Molecular data suggest that genes with a pivotal
role in normal angiogenesis are also involved in the
pathogenesis of hemangiomas. Molecular genetic
research is starting to offer insight into the possible
causes of isolated hemangiomas and may even ex-
plain primary field defects. Increasingly, associations
of malformations that were once thought to be field
defects turn out to be single gene defects. In these
cases, abnormal tissue architecture can often be ob-
served. Examples include the Denys-Drash syn-
drome, a disorder characterized by male pseudoher-
maphroditism, congenital nephrotic syndrome,
early renal failure, caused by mutations in the Wilms’
tumor gene WT1,63 and holoprosencephaly types 2
and 3, caused by mutations in the SIX3 gene, which
plays a role in eye and forebrain development, and
sonic hedgehog gene SHH, a crucial gene for brain
architecture, respectively.64,65 Currently, there is no
definitive proof that hemangiomas are in fact the
result of clonal expansion of one cell bearing a single
gene defect, although data exist suggesting clonality
by HUMARA (human androgen receptor) assay.
Volume 117, Number 2 • Pathogenesis of Hemangiomas
Monoclonality could not be shown to be probable in
all hemangiomas, and monoclonality was also dem-
onstrated in endothelial control tissue samples.23,29
Future studies will need to address this issue.
If hemangiomas were truly caused by single
gene mutations, however, one would expect them
to represent somatic mosaicism. In that case, mul-
tiple hemangiomas should be observed following
established mosaic or nevoid patterns. Some au-
thors believe that they do, but so far, formal proof
is lacking.66
While attractive, the developmental field the-
ory does not explain why histogenesis is abnormal
in hemangiomas. In a polytopic field defect, tis-
sues are malformed but their cell populations and
microarchitecture are normal.34
Isolated hemangiomas are, by definition, sec-
ondary monotopic field defects. Again, the problem
with the field defect hypothesis is that it would pre-
dict solitary hemangiomas to be true malformations.
Instead, they are tumors with abnormal tissue archi-
tecture. Vascular malformations would be obvious
and better candidates for secondary field defects,
and isolated hemangiomas probably do not repre-
sent field defects.67 The hemangiomas associated
with multiple malformations may be caused by sec-
ondary effects of the field defect: the abnormal sig-
naling within the developmental field may lead to
disturbances of angiogenesis. Angiogenetic and an-
giostatic cytokine levels change not only within the
hemangioma itself as it goes through its life cycle, but
also in the normal tissue surrounding the heman-
gioma. This implies that a hemangioma might in-
duce angiogenesis in susceptible surrounding tis-
sues, or vice versa: angiogenic active or dysfunctional
surroundings may provoke a hemangioma. It is also
conceivable that it is not the field disruption itself
that causes the hemangioma, but the field disrup-
tion that makes the field more susceptible to other
factors that cause a hemangioma, for example, the
seeding of trophoblast cells.
Unfortunately, it cannot be excluded that the
association of hemangiomas and field defects or
other anomalies is not causal. The high incidence
of hemangiomas and the low incidence of the
various anomalies make it difficult to validate such
an association.
The placental seeding theory is attractive
because it explains a finding that no other the-
ory plausibly explains: the correlation between
chorionic villus sampling and the increased in-
cidence of hemangiomas. Because hemangio-
mas are only seen in humans, ultimate experi-
ments to test this theory would involve human
fetal and neonatal tissues.
Angiogenesis is an incompletely understood and
highly complex process. Because hemangiomas are
endothelial, vascular tumors, hemangioma forma-
tion ipso facto involves deranged angiogenesis. A de-
rangement of angiogenic checks and balances could
therefore be a cause of hemangiomas, but it could
also be the effect of an underlying disturbance. Of
particular interest are the findings that hemangio-
mas may produce not only growth factors but also
the receptors for these growth factors.
Angioblast origin and mutational origin are
hard to distinguish in molecular research. Mono-
clonality is likely in both cases. The susceptibility
to angiogenic factors of the specific environment
of the angioblast or the mutated cell, the timing
of the postzygotic mutation, and the downstream
effects of the mutated gene can make it difficult to
differentiate between the two routes of hemangi-
oma formation. It is even possible that both mech-
anisms result in the same pathophysiological path-
way.
Autosomal dominant hereditary hemangioma
formation, in some cases linked to chromosome
5q, indicates that genetic causes exist in at least
some cases. The most probable candidate genes
are PDGFR␤, FGFR4, and FLT4 (VEGFR3).
A theory on the origin of hemangiomas should
aspire to explain the biological hallmarks of hem-
angiomas: occurrence after birth, autonomous
growth, spontaneous involution, and a marked
predilection for the female sex. Most research,
however, is focused on what starts a hemangioma
and not on its biological behavior. Future research
should also aim to elucidate why they occur so
soon after birth and why hemangiomas involute.
Possible causes could be a shift in cytokine ex-
pression after birth or a down-regulation within
the hemangioma.68 It is also possible that the pla-
centa produces angiostatic factors that inhibit
hemangioma growth in utero. A possible candi-
date might be a placenta-produced, soluble VEGF
antagonist, sFLT1, that is secreted into the mater-
nal circulation.69 Its levels in the circulation of
normal fetuses and neonates or those with hem-
angiomas have never been established.
All hemangiomas might not be the same. An
isolated nodular hemangioma or a hemangioma
that occurs after choriovillus sampling could very
well be a different entity from a familial hemangi-
oma or a plaque-like hemangioma embedded in a
field disruption. Hemangiomas may be the result of
primary polytopic field defects insofar as they are
part of a malformation complex, but isolated hem-
angiomas may very well be caused by congenital or
acquired genetic defects and as such could represent
33e
 Plastic and Reconstructive Surgery • February 2006
an entirely different disorder. Because some single
gene defects may lead to disorders formerly inter-
preted as secondary polytopic or monotopic field
defects with abnormal histogenesis, the interpreta-
tion of associated hemangiomas as manifestations of
primary polytopic field defects is becoming increas-
ingly doubtful.
Research on hemangioma origins has led to a
sketchy picture. Genetic causes and trophoblast or
angioblast origin are the most promising research
fields for the etiology of hemangiomas of infancy,
but they still lack the power to account for many
hemangioma properties.
Constantijn G. Bauland, M.D.
Department of Plastic Surgery
Radboud University Nijmegen Medical Center
Reinier Postlaan 4
P. O. Box 9101
6500 HB Nijmegen, The Netherlands
bauland@nvpc.nl
REFERENCES
1. Rieu, P. N. M. A., Verheyen, N., and Verbeek, A. L. M.
Prevalence of hemangioma. In Proceedings of the 12th Meeting
of the International Society for the Study of Vascular Anomalies,
Berlin, Germany, June 27–28, 1998.
2. Bruckner, A. L., and Frieden, I. J. Hemangiomas of infancy.
J. Am. Acad. Dermatol. 48: 477, 2003.
3. Holmdahl, K. Cutaneous hemangiomas in premature and
mature infants. Acta Pediatr. 44: 370, 1955.
4. Jacobs, A. H. Strawberry hemangiomas: The natural history
of the untreated lesion. Calif. Med. 86: 8, 1957.
5. Wahrman, J. E., and Honig, P. J. Hemangiomas. Pediatr. Rev.
15: 266, 1994.
6. Mulliken, J. B. Diagnosis and natural history of hemangio-
mas. In J. B. Mulliken and M. A. Young (Eds.), Vascular
Birthmarks: Hemangiomas and Malformations. Philadelphia:
Saunders, 1988. Pp. 41– 62.
7. Mulliken, J. B. Treatment of hemangiomas. In J. B. Mulliken
and M. A. Young (Eds.), Vascular Birthmarks: Hemangiomas
and Malformations. Philadelphia: Saunders, 1988. Pp. 77–103.
8. Muir, T., Kirsten, M., Fourie, P., Dippenaar, N., and Ionescu,
G. O. Intralesional bleomycin injection (IBI) treatment for
haemangiomas and congenital vascular malformations. Pe-
diatr. Surg. Int. 19: 766, 2004.
9. Mulliken, J. B., and Young, M. A. Vascular birthmarks in
folkore, history, art and literature. In J. B. Mulliken and M. A.
Young (Eds.), Vascular Birthmarks: Hemangiomas and Malfor-
mations. Philadelphia: Saunders, 1988. Pp. 3–23.
10. Frank, K. Überlegungen zur möglichen Ursache der Hä-
mangiomatose unter Berücksichtigung ihres klinischen
Bildes. Pädiatr. Grenzgeb. 20: 231, 1981.
11. Williams, R. L., Risau, W., Zerwes, H. G., Drexler, H., Aguzzi,
A., and Wagner, E. F. Endothelioma cells expressing the
polyoma middleT oncogene induce hemangiomas by host
cell recruitment. Cell 57: 1053, 1989.
12. Mulliken, J. B., and Glowacki, J. Hemangiomas and vascular
malformations in infants and children: A classification based
on endothelial characteristics. Plast. Reconstr. Surg. 69: 412,
1982.
13. Mulliken, J. B. Pathogenesis of hemangiomas. In J. B. Mul-
liken and M. A. Young (Eds.), Vascular Birthmarks: Hemangi-
34e
omas and Malformations. Philadelphia: Saunders, 1988. Pp.
63-76.
14. Takahasi, K., Mulliken, J. B., Kozakewich, H. P. W., Rogers,
R. A., Folkman, J., and Ezekowitch, A. B. Cellular markers
that distinguish the phases of hemangioma during infancy
and childhood. J. Clin. Invest. 93: 2357, 1994.
15. Tan, S. T., Velickovic, M., Ruger, M., and Davis, P. F. Cellular
and extracellular markers of hemangioma. Plast. Reconstr.
Surg. 106: 529, 2000.
16. Tan, S. T., Wallis, R. A., He, Y., and Davis, P. F. Mast cells and
hemangioma. Plast. Reconstr. Surg. 113: 999, 2004.
17. Iwata, J., Sonobe, H., Furihata, M., Ido, E., and Ohtsuki, Y.
High frequency of apoptosis in infantile capillary hemangi-
oma. J. Pathol. 179: 403, 1996.
18. Mancini, A. J., and Smoller, B. R. Proliferation and apoptosis
in hemangiomas. Am. J. Dermapathol. 18: 505, 1996.
19. Razon, M. J., Kräling, B. M., Mulliken, J. B., and Bisschoff,
J. Increased apoptosis coincides with onset of involution in
infantile hemangioma. Microcirculation 5: 189, 1998.
20. Frischer, J. S., Huang, J., Serur, A., Kadenhe, A., Yamashiro,
D. J., and Kandel, J. J. Biomolecular markers and involution
of hemangiomas. J. Pediatr. Surg. 39: 400, 2004.
21. Sasaki, G. H., Pang, C. Y., and Witliff, J. L. Pathogenesis and
treatment of infant skin strawberry hemangioma: Clinical
and in vitro studies of hormonal effects. Plast. Reconstr. Surg.
73: 359, 1984.
22. Berard, M., Sordello, S., Ortega, N., et al. Vascular endothe-
lial growth factor confers a growth advantage in vitro and in
vivo to stromal cells cultured from neonatal hemangiomas.
Am. J. Pathol. 150: 1315, 1997.
23. Walter, J. W., North, P. E., Waner, M., et al. Somatic mutation
of vascular endothelial growth factor receptors in juvenile
hemangioma. Genes Chromosomes Cancer 33: 295, 2002.
24. North, P. E., Waner, M., Mizeracki, A., et al. A unique mi-
crovascular phenotype shared by juvenile hemangiomas and
human placenta. Arch. Dermatol. 137: 559, 2001.
25. Burns, A. J., Kaplan, L. C., and Mulliken, J. B. Is there an
association between hemangioma and syndromes with dys-
morphic features? Pediatrics 88: 1257, 1991.
26. Bielenberg, D. R., Bucana, C. D., Sanchez, R., Mulliken, J. B.,
Folkman, J., and Fidler, I. J. Progressive growth of infantile
cutaneous hemangioma is directly correlated with hyperpla-
sia and angiogenesis of adjacent epidermis and inversely
correlated with expression of the angiogenesis inhibitor IFN-
beta. Int. J. Oncol. 14: 401, 1999.
27. Ritter, M. R., Friedlander, D. F., Copeland, D., Reinisch, J.,
Friedlander, S. F., and Friedlander, M. Haemangioma en-
dothelium shows evidence of differentiation from mono-
cytes. In Proceedings of the 15th Meeting of the International Society
for the Study of Vascular Anomalies, Wellington, New Zealand,
February 22-25, 2004.
28. Yu, Y., Flint, A. F., Mulliken, J. B., Wu, J. K., and Bischoff,
J. Endothelial progenitor cells in infantile hemangioma.
Blood 103: 1373, 2004.
29. Boye, E., Yu, Y., Paranya, G., Mulliken, J. B., Olsen, B. R., and
Bisschof, J. Clonality and altered behavior of endothelial cells
from hemangiomas. J. Clin. Invest. 107: 745, 2001.
30. Nguyen, V. A., Fürhapter, C., Romani, N., Weber, F., and
Sepp, N. Infantile hemagioma is a proliferation of ␤4-neg-
ative endothelial cells adjacent to HLA-DR-positive cells with
dendritic cell morphology. Hum. Pathol. 35: 739, 2004.
31. Pack, G. T., and Miller, T. R. Hemangiomas: Classification,
diagnosis and treatment. Angiology 1: 405, 1950.
32. Smoller, B. R., and Apfelberg, D. S. Infantile (juvenile) capillary
hemangioma: A tumor of heterogenous cellular elements.
J. Cutan. Pathol. 20: 330, 1993.
Volume 117, Number 2 • Pathogenesis of Hemangiomas
33. Dadras, S. S., North, P. E., Bertoncini, J., Mihm, M. C., and
Detmar, M. Infantile hemangiomas are arrested in an early
developmental vascular differentiation state. Mod. Pathol. 17:
1068, 2004.
34. Martinez-Frias, M. L., Frias, J. L., and Opitz, J. M. Errors of
morphogenesis and developmental field theory. Am. J. Med.
Genet. 76: 291, 1998.
35. Waner, M., North, P. E., Scherer, K. A., Frieden, I. J., Waner,
A., and Mihm, M. C. The nonrandom distribution of facial
hemangiomas. Arch. Dermatol. 139: 869, 2003.
36. Orlow, S. J., Isakoff, M. S., and Blei, F. Increased risk of
symptomatic hemangioma of the airway in association with
cutaneous hemangiomas in a “beard” distribution. J. Pediatr.
131: 643, 1997.
37. Goldberg, N. S., Hebert, A. A., and Esterly, N. B. Sacral
hemangiomas and multiple congenital abnormalities. Arch.
Dermatol. 122: 684, 1986.
38. Albright, A. L., Gartner, J. C., and Wiener, A. S. Lumbar
cutaneous hemangiomas as indicators of tethered spinal
cords. Pediatrics 83: 977, 1989.
39. Gorlin, R. J., Kantanaputra, P., Aughton, D. J., and Mulliken,
J. B. Marked female predilection in some syndromes asso-
ciated with facial hemangiomas. Am. J. Med. Genet. 52: 130,
1994.
40. Hersh, J. H., Waterfill, D., Rutledge, J., et al. Sternal mal-
formation/vascular dysplasia syndrome. Am. J. Med. Genet. 21:
177, 1985.
41. Kishnani, P., Iafolla, A. K., McConkie-Rosell, A., van Hove,
J. L. K., Kanter, R. J., and Kahler, S. G. Hemangioma, su-
praumbilical midline raphé, and coarctation of the aorta
with a right aortic arch: Single causal entity? Am. J. Med. Genet.
58: 44, 1995.
42. Crisponi, G., Marras, A. R., Corrias, A., Memo, L., Gorinati,
M., and Flora, P. G. Two patients with varying combinations
of sternal cleft, haemangiomas, midline abdominal raphe,
coarctation of the aorta with a right aortic arch. Clin. Dys-
morphol. 9: 103, 2000.
43. Frieden, I. J., Reese, V., and Cohen, D. PHACE syndrome:
The association of posterior brain fossa malformations, hem-
angiomas, arterial anomalies, coarctation of the aorta and
cardiac defects, and eye abnormalities. Arch. Dermatol. 132:
307, 1996.
44. Metry, D. W., Dowd, C. F., Batkovich, A. J., and Frieden, I. J.
The many phaces of PHACE syndrome. J. Pediatr. 139: 117,
2001.
45. Poetke, M., Frommeld, T., and Berlien, H. P. PHACE syn-
drome: New views on diagnostic criteria. Eur. J. Pediatr. Surg.
12: 366, 2002.
46. North, P. E., Waner, M., Mizeracki, A., and Mihm, M. C.
GLUT-1: A newly discovered immunohistochemical marker
for juvenile hemangiomas. Hum. Pathol. 31: 11, 2000.
47. Bree, A. F., Siegfried, E., Sotelo-Avila, C., and Nahass, G.
Infantile hemagiomas: Speculation on placental trophoblas-
tic origin. Arch. Dermatol. 137: 573, 2001.
48. Kaplan, P., Normandin, J., Wilson, G. N., Plauchu, H., Lip-
pman, A., and Vekemans, M. Malformations and minor
anomalies in children whose mothers had prenatal diagno-
sis: Comparison between CVS and amniocentesis. Am. J. Med.
Genet. 37: 366, 1990.
49. Burton, B. K., Schulz, C. J., Angle, B., and Burd, L. I. An
increased incidence of hemangiomas in infants born follow-
ing chorionic villous sampling (CVS). Prenat. Diagn. 15: 209,
1995.
50. Los, F. J., Noomen, P., Vermeij-Keers, C., et al. Chorionic
villus sampling and materno-fetal transfusions: An immuno-
logical pathogenesis of vascular disruptive syndromes? Pre-
nat. Diagn. 16: 193, 1996.
51. Jansen, M. W., Brandenburg, H., Wildschut, H., et al. The
effect of chorionic villous sampling on maternal serum al-
pha-fetoprotein levels. Prenat. Diagn. 17: 953, 1997.
52. Folkman, J. Clinical application of research on angiogenesis.
N. Engl. J. Med. 333: 1757, 1995.
53. Ferrara, N., and Keyt, B. Vascular endothelial growth factor:
Basic biology and clinical implications. In I. D. Goldberg and
E. M. Rosen (Eds.), Regulation of Angiogenesis. Basel, Switzer-
land: Birkhäuser Verlag, 1997. Pp. 209-232.
54. Chang, J., Most, D., Bresnick, S., et al. Proliferative heman-
giomas: Analysis of cytokine gene expression and angiogen-
esis. Plast. Reconstr. Surg. 103: 1, 1999.
55. Kräling, B. M., Razon, M. J., Boon, L., et al. E-selectin is
present in proliferating endothelial cells in human heman-
giomas. Am. J. Pathol. 148: 1181, 1996.
56. Ezekowitz, R. A. B. The relationship between facial end air-
way hemangiomas: Does seeing red bode ill? J. Pediatr. 131:
514, 1997.
57. Yu, Y., Varughese, J., Brown, L. F., Mulliken, J. B., and
Bischoff, J. Increased Tie2 expression, enhanced response to
angiopoietin-1, and dysregulated angioppoietin-2 expres-
sion in hemangioma-derived endothelial cells. Am. J. Pathol.
159: 2271, 2001.
58. Ritter, M. R., Dorell, M. I., Edmonds, J., Friedlander, S. F.,
and Friedlander, M. Insulin-like growth factor 2 and poten-
tial regulators of hemangioma growth and involution iden-
tified by large-scale expression analysis. Proc. Natl. Acad. Sci.
U.S.A. 99: 7455, 2002.
59. Cheung, D. S. M., Warman, M. L., and Mulliken, J. B. Hem-
angioma in twins. Ann. Plast. Surg. 38: 269, 1997.
60. Blei, F., Walter, J., Orlow, S. J., and Marchuk, D. A. Familial
segregation of hemangiomas and vascular malformations as
an autosomal dominant trait. Arch. Dermatol. 134: 718, 1998.
61. Walter, J. W., Blei, F., Anderson, J. L., Orlow, S. J., Speer,
M. C., and Marchuk, D. A. Genetic mapping of a novel form
of infantile hemangioma. Am. J. Med. Genet. 82: 77, 1999.
62. Berg, J. N., Walter, J. W., Thisanagayam, U., et al. Evidence
for loss of heterozygosity of 5q in sporadic haemangiomas:
Are somatic mutations involved in haemangioma formation?
J. Clin. Pathol. 54: 249, 2001.
63. Schneider, S., Wildhardt, G., Ludwig, R., and Royer-Pokora,
B. Exon skipping due to a mutation in a donor splice site in
the WT-1 gene is associated with Wilms’ tumor and severe
genital malformations. Hum. Genet. 91: 599, 1993.
64. Roessler, E., Belloni, E., Gaudenz, K., et al. Mutations in the
human sonic hedgehog gene causes holoprosencephaly. Nat.
Genet. 14: 357, 1996.
65. Wallis, D. E., Roessler, E., Hehr, U., et al. Mutations in the
homeodomain of the human SIX3 gene cause holoprosen-
cephaly. Nat. Genet. 22: 196, 1999.
66. Happle, R. Personal communication, 2003.
67. Brouillard, P., and Vikkula, M. Vascular malformations: Lo--calized
defects in vascular morphogenesis. Clin. Genet. 63: 340, 2003.
68. Malamitsi-Puchner, A., Tziotis, J., Protonotariou, E.,
Sarandakou, A., and Creatsas, G. Angiogenic factors in the
perinatal period: Diversity in biological functions reflected in
their serum concentrations soon after birth. Ann. N. Y. Acad.
Sci. 900: 169, 2000.
69. Clark, D. E., Smith, S. K., He, Y., et al. A vascular endothelial
growth factor antagonist is produced by the human placenta
and released into the maternal circulation. Biol. Reprod. 59:
1540, 1998.
35e