ARTICLE
Auteur(s) : Denise Josephina Johanna
Hermans1, Jan Bernardus Maria Boezeman1,
Peter Cornelius Maria Van de Kerkhof1, Paul Nicolas
Maria Antonius Rieu2, Carine Joanna Maria Van der
Vleuten1
1Department of Dermatology Radboud University
Medical Centre, René Descartesdreef 1, 6525 GC Nijmegen, The
Netherlands
2Department of Pediatric Surgery Radboud University
Medical Centre, Geert Grooteplein-Zuid 10, 6525 GA Nijmegen, The
Netherlands
accepté le 5 Octobre 2008
Although infantile hemangiomas (IH) occur in 10-12% of children
younger than 1 year of age, this benign tumor is not well
understood. One of these unknown areas is ulceration [1]. Although
ulceration is the most common and distressing complication of IH,
possibly affecting between 5% and 13% of the IH, little is known
about its pathogenesis [1, 2]. In the literature three factors were
found that may play a role: localization with a high risk of
trauma, local factors such as bacteria (either infection or
colonization) (figure
1) and tissue hypoxia caused by a fast growing IH that
outreaches its blood supply (figure 2) [2, 3]. In the
past, few studies were carried out that investigated ulcerating IH.
The purpose of the present study was to get a better insight into
the clinical characteristics of ulcerated IH and non-ulcerated IH
in a retrospective analysis of hemangioma patients, who visited our
multidisciplinary hemangioma-study group over the past 10 years.
For this extensive patient group, the known differences between
ulcerated and non-ulcerated IH from earlier studies were
investigated again and supplemented with other characteristics that
were also considered important. Better knowledge of the differences
between ulcerated and non-ulcerated IH will hopefully deliver
better insight into the pathogenesis and treatment of this
distressing problem in young children.
Patients and methods
Patients
A retrospective analysis was performed of all IH at the Radboud
University Medical Centre Nijmegen (UMCN), the Netherlands, from
1997 to 2007. The medical records and photo documentation of 465
patients with a hemangioma were reviewed. The following information
was obtained from each patient file:
Hemangioma: ulceration or not (1), number of IH per patient (2),
size (3), anatomic localization (4), type (5), age at time of
discovery (6).
Ulceration: phase of growth at time of ulceration (7), duration
(8).
Other: sex (9), prematurity/gestational age (10), birth weight
(11), multiple gestation (12), (transcervical) chorionic villus
sampling (13), atopic constitution/dermatitis (14).
Analysis
Data from the patient files were transported into a database and
analyzed to define the characteristics of the study group and
subsequently the differences between the ulcerated and
non-ulcerated IH. The numeric data were reported as mean ± SEM. For
statistical analysis the t-test for unpaired values was used. The
other non-numeric data were analyzed with the Fisher exact test.
The two tailed hypothesis was employed to interpret data.
A p-value ≤ 0.05 was regarded as statistically significant.
Results
After registration of data mentioned above, the files were divided
in two groups: IH with active or past ulceration versus IH which
never ulcerated. Comparison between the two groups was carried out.
In the description below, the results of all examined
characteristics are reported.
Ulcerated vs. non-ulcerated IH
A total of 465 records of patients were investigated. Of these 107
(23%) were diagnosed with an ulcerated IH versus 358 (77%) patients
with one or more non-ulcerated IH. The patients with an ulcerated
IH had a total of 235 IH, 108 ulcerated and 127 non-ulcerated. The
358 patients with only non-ulcerated IH had a total of 815 IH.
Number of IH
The average number of IH in the group of patients with an ulcerated
IH was 2.2 (SEM ± 0.5). For the patient-group with only
non-ulcerated IH this mean number was 2.3 (SEM ± 0.3). This
difference is not statistically significant (p = 0.9).
Size of IH
The surface of the IH was measured with a tape-measure in 2
perpendicular directions, using the maximum diameter in each to
calculate the final surface.
The mean length and width of the ulcerated IH were respectively
5.1 cm. (SEM 0.4) and 4.3 cm. (SEM 0.3). For the
nonulcerated IH this was 2.1 cm. (SEM 0.1), and 1.8 cm.
(SEM 0.1). This shows that the ulcerating IH had a significantly
larger surface than the non-ulcerated IH (28.6 cm2
vs. 6.0 cm2, p < 0.05). Data concerning the size
were only known from 51.6% of the ulcerated IH and 62.0% of the
non-ulcerated IH.
Anatomical localization
The IH were classified in 7 different region categories: (1)
extremities, (2) head and neck, (3) perineum, (4) groin, (5) trunk,
(6) buttock and (7) sacral region. In general, most IH were located
on the head and neck region, the extremities and trunk. The
predilection area for ulceration was the head and neck region, but
also IH in the perineal and buttock area often turned out to be
ulcerating. Ulcerated IH were significantly more often localized in
the diaper-region (p < 0.05) (table
1).
The localization of the IH was not uniformly documented. 29.5%
of the non-ulcerated IH had an unknown localization.
Table 1 Localization
|
Localization
|
|
Ulcerated (%)
|
Non-ulcerated (%)
|
|
Extremities
|
|
21.3
|
20.8
|
|
Head and neck
|
|
47.2
|
52.9
|
|
Torso
|
|
12.0
|
22.4
|
|
Perineum
|
} diaper area
|
12.0
|
1.5
|
|
Buttock
|
6.5
|
1.4
|
|
Groin
|
0.9
|
0.6
|
|
Sacral region
|
|
0
|
0.5
|
Subtype of IH
IH were divided in 3 clinical subtypes; (1) superficial, (2) deep
and (3) mixed IH with involvement of both epi-dermis and subcutis.
When both ulcerated and non-ulcerated IH were compared, ulceration
was more frequently seen in the IH with a superficial component.
78.2% of the ulcerated IH had a superficial component compared to
61.1% of the non-ulcerated IH. This was a significant difference (p
< 0.05). The deep IH ulcerated significantly less frequently (p
< 0.05) (table 2). It was known
whether they were superficial, deep or mixed for 65.0% of the
ulcerated IH and 64.9% of the non-ulcerated IH.
Added to this, IH can be divided in morphological subtypes;
localized and segmental IH can be distinguished. Localized IH are
focal tumor like lesions. The less commonly occurring segmental IH
tend to be more plaque-like and involve a region or segment of
skin. Ulcerated IH had a segmental distribution significantly more
often (29.3%) compared to 2.1% of the non-ulcerated IH (p <
0.05) (table 2). Data concerning the
distribution were available for 85.2% of ulcerated IH and 95.5% of
non-ulcerated IH.
Table 2 Subtypes
|
Clinical type
|
Ulcerated (%)
|
Non-ulcerated (%)
|
|
Superficial total
|
78.2
|
61.1
|
|
Deep
|
1.5
|
25.3
|
|
Mixed
|
20.3
|
13.6
|
|
Morphological type
|
|
|
|
Segmental
|
29.3
|
2.1
|
|
Localized
|
70.7
|
97.9
|
Phase of ulceration
The growth characteristics of IH can be divided in phases:
proliferation phase (0-9 months), a short plateau phase and
involution phase (until the age of 10-12 years). Growth
characteristics of IH in an individual infant may vary. In this
study the phase in which the ulceration occurred was investigated.
Most frequently, ulceration took place in the proliferation phase
(83.1%), but also in the involution phase (15.3%) and the plateau
phase (1.7%). It was known in which phase the ulceration had taken
place for 54.6% of the IH.
Duration of ulceration
In 20.4% of cases the duration of ulceration was well documented
and varied between 4 days and 7 months, with a mean of 8.1 weeks
(SEM 1.6).
Sex
For the patients examined the sex distribution was 67.3% girls and
33.7% boys, which makes a ratio of 1:2. For the group of patients
with an ulcerated IH the number of male and female patients was
respectively 27 (25.2%) and 80 (74.8%), a ratio of 1:3. This means
a tendency to significantly more females with an ulcerated IH
compared to female patients with just non-ulcerated IH (p = 0.08)
(figure 3).
Prematurity/gestational age
The mean gestational age in the group with an ulcerated IH was 35.0
weeks (SEM 0.6). For the group with just non-ulcerated IH the mean
gestational age was 36.7 weeks (SEM 0.47). This was significantly
different (p = 0.04).
Prematurity in this study was defined as birth before 37 weeks
of gestation. Analysis of the number of prematures in both patient
groups, resulted in 61.1% prematures in the patient group with
ulcerated IH compared to 26.5% of the children in the patient group
with only non-ulcerated IH. This was a significant difference (p
< 0.05) (figure
3). It was known whether or not they were premature for
33.6% of the patients with an ulcerated IH and 27.4% of the
children with non-ulcerated IH.
Subsequently the degree of the prematurity for both groups was
compared. For the premature children with ulcerated IH the mean
gestational age was 33.1 weeks (SEM 0.6). For the patients with
just non-ulcerated IH this mean age was 32.7 weeks (SEM 0.6). This
was not significantly different (p = 0.7). For 30.8% of the
children with an ulcerated IH, the exact gestational age was known,
compared to 20.1% of the children with no ulcerated IH.
Birth weight
The mean birth weight found for the patient group with an ulcerated
IH was 2,649 grams (SEM 140.7) and 3,022 grams (SEM
189.3) for the other patient group with only non-ulcerated IH.
These indicates a tendency to a significantly lower birth weight
for children with ulcerated IH (p = 0.06).
Birth weight was documented for 18.7% of the patients with an
ulcerated IH and for 20.1% of the patients with just nonulcerated
IH.
Multiple gestation
For both patient groups it was registered whether or not they were
part of twins. In the total study group 5.6% of the patients were
part of twins. For the patient group with an ulcerated IH this was
6.5%, for the patient group with non-ulcerated IH this was 5.3%.
This is not a significant difference (p = 0.4) (figure 3).
Chorionic villus sampling
In this study no chorionic villus sampling was performed in either
patient group.
Atopic dermatitis
In the patient group diagnosed with an ulcerated IH, 54.2% of the
children were also diagnosed with atopic dermatitis, in the patient
group with only non-ulcerated IH this was 46.5% (figure 3). This is not a
significant difference (p = 0.62). Overall, not all patient files
gave clear information about the presence of atopic dermatitis in
the patient. Only for 22.4% of the patients with an ulcerated IH
versus 12.0% of the patients with nonulcerated IH was it noted.
Discussion
In this study, ulceration was found in 23% of the IH patients
compared to 5-13% described in literature [1, 2]. This can be
explained by the fact this study was carried out in an academic
centre with a specialized multidisciplinary hemangioma team that
serves as a tertiary centre for patients with difficult
(ulcerating) IH.
When overviewing the characteristics described, there were no
differences concerning the number of IH per patient between the
patient group with one or more ulcerated IH versus the group with
just non-ulcerated IH. With respect to size-, localization and type
of IH, some differences could be recognized, comparable to the
recently published results of the American cohort studied by
Chamlin et al. in 2007 [4].
Ulcerated IH were significantly larger than non-ulcerated IH. In
line with this result, segmental IH, which in general cover a
substantial skin surface, ulcerated more frequently. Probably,
larger IH run a greater risk of becoming mechanically damaged by
friction resulting in ulceration. With respect to localization,
ulcerated IH were mostly localized in the head- and neck-region,
but the diaper area was also a localization with an enlarged risk
for ulceration. It seems that traumas like friction and
contamination/maceration are more common in these areas. With
respect to clinical types, IH with a superficial component were
frequently ulcerated as opposed to deeper IH that seldom
ulcerated.
Taken together, epidermal involvement and susceptibility to
trauma and maceration seem to play a role in pathogenesis of
ulceration in IH. These characteristics are connected with the
barrier function of the skin. If this is impaired, there seems to
be a greater risk of ulceration [2]. In view of this, it was
interesting to compare the incidence of atopic dermatitis, in which
the barrier function is also impaired, between the patient groups
with and without ulcerated IH, but no significant difference could
be found. Additional prospective studies need to be done to confirm
or reject this assumed interesting relationship.
The phase in which ulceration mostly took place was obviously
the proliferation phase. This is also known from the literature [2,
4, 5]. The possible reason given in literature is the outgrowth of
the blood supply in the fast growing IH, resulting in central
necrosis [6, 7]. The mean duration of ulceration turned out to be
about 8 weeks. It must be stated that, because our centre is a
referral centre for complicated IH, this might be overestimated in
this study. Besides this, the exact duration is often unknown
because of the difficulty of exact registration. Patients who do
better are often lost to follow up and/or return to the general
practitioner. Therefore a prospective study should be carried out
as well.
In addition, demographic characteristics were studied. From the
literature it is known that the female to male rate for IH patients
is 2.5-4 to 1 [2]. In our study it was found that the group with
ulcerated IH comprised a higher percentage of female patients
compared to the patients with non-ulcerated IH. In former studies,
hemangioma patients were more likely to be premature. Which could
be in relation with the fact that a greater percentage of premature
children are female [8-11]. In our study it was observed that the
patients with ulcerated IH were more often premature. The mean
gestational age of the children with ulcerated IH was also lower.
Besides this, the mean birth weight of patients with ulcerated IH
also seemed lower although with only a tendency to significance.
This last result is also known from the literature [12]. This could
mean that prematurity and low birth weight increase the risk of
ulceration.
It was striking that in all twin cases, the other twin was
unaffected. The number of twins in the studied cohort (5.6%) was
higher than the 1.6% twins in the general population, which is in
line with a higher incidence of prematurity and lower birth weights
in this group. Although ulceration was not more frequent in twins
with an IH.
Conclusion
Ulceration is a frequent complication affecting 23% of our studied
patients. This high incidence of ulceration makes this Dutch
cohort, containing high rates of complicated IH, very appropriate
for evaluating the characteristics of ulcerated IH. In summary,
larger IH with a superficial component in areas more predisposed
for contamination are more at risk for ulceration. We found a high
incidence of ulceration in the large group of IH in the head neck
region but also a statistical significantly higher percentage of
ulcerated IH in the perineum and buttock area. These results will,
however, give us more insight into the unknown and multifactorially
determinated pathogenesis of ulceration in IH. Besides this, it
will probably make it possible to predict whether or not an IH
might ulcerate.
Acknowledgements
Financial support: none. Conflict of interest: none.
References
1 Esterly NB. Cutaneous hemangiomas, vascular stains and
malformations, and associated syndromes. Curr Probl Pediatr 1996;
26: 3-39.
2 Frieden IJ, Haggstrom AN, Drolet BA,
et al. Infantile hemangiomas: current knowledge, future
directions. Proceedings of a research workshop on infantile
hemangiomas, April 7-9, 2005, Bethesda, Maryland, USA. Pediatr
Dermatol 2005; 22: 383-406.
3 Kim HJ, Colombo M, Frieden IJ. Ulcerated
hemangiomas: clinical characteristics and response to therapy. J Am
Acad Dermatol 2001; 44: 962-72.
4 Chamlin SL, Haggstrom AN, Drolet BA,
et al. Multicenter prospective study of ulcerated hemangiomas.
J Pediatr 2007; 151: 684-9; (689.e1).
5 Liang MG, Frieden IJ. Perineal and lip ulcerations
as the presenting manifestation of hemangioma of infancy.
Pediatrics 1997; 99: 256-9.
6 Thomas RF, Hornung RL, Manning SC, et al.
Hemangiomas of infancy: treatment of ulceration in the head and
neck. Arch Facial Plast Surg 2005; 7: 312-5.
7 Waner M, Suen JY. Hemangiomas and vascular
malformations of the Head and Neck. In: The natural history of
hemangiomas. New York: Wiley-liss, 1999: 13-45.
8 Amir J, Metzker A, Krikler R, et al.
Strawberry hemangioma in preterm infants. Pediatr Dermatol 1986; 3:
331-2.
9 Haggstrom AN, Drolet BA, Baselga E, et al.
Prospective study of infantile hemangiomas: demographic, prenatal,
and perinatal characteristics. J Pediatr 2007; 150: 291-4.
10 Miller T, Frieden IJ. Hemangiomas: new insights and
classification. Pediatr Ann 2005; 34: 179-87.
11 Powell TG, West CR, Pharoah PO, et al.
Epidemiology of strawberry haemangioma in low birthweight infants.
Br J Dermatol 1987; 116: 635-41.
12 Morrison J. Twin gestation and premature birth. J
Perinatol 2005; 25: 1-3.
|
Printable version
Version PDF
