A, Deep hemangioma that regressed without sequelae; B, superficial hemangioma that left only telangiectasia; C, mixed hemangioma that left anetodermic skin; D, mixed hemangioma that left redundant skin; and E, mixed hemangioma that left fibrofatty tissue.
A, Patient 21 at age 8 months (left), and age 3 years (right); B, patient 8 at age 8 months (left) and 3 years (right). Superficial hemangioma with a step border (patient 21, A) left more sequelae than a mixed hemangioma of the same size and location but with a progressive border (patient 8, B).
A, Patient 22 at age 6 months (left) and age 4 years (right); B, patient 31 at age 8 months (left) and 2 years (right). Hemangiomas with a smooth surface leave less sequelae (patient 22, A), than hemangiomas with a cobblestoned surface (patient 31, B).
Customize your JAMA Network experience by selecting one or more topics from the list below.
Baselga E, Roe E, Coulie J, et al. Risk Factors for Degree and Type of Sequelae After Involution of Untreated Hemangiomas of Infancy. JAMA Dermatol. 2016;152(11):1239–1243. doi:10.1001/jamadermatol.2016.2905
Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Can the clinical characteristics of hemangiomas predict the type and degree of sequelae that they leave after natural involution?
In this retrospective cohort study of hemangiomas left to natural involution, 54.9% left significant sequelae. Superficial and deep hemangiomas left significantly fewer sequelae than combined hemangiomas. Hemangiomas with a step or abrupt border and with a cobblestoned surface left significantly more sequelae.
Clinical characteristic of the hemangioma help to predict the risk of sequela and should guide treatment decisions.
Infantile hemangiomas involute to some extent, but they often leave sequelae that may cause disfigurement. Factors determining the risk of permanent sequelae after regression are of crucial importance in treatment decision making.
To describe the sequelae left by infantile hemangiomas after natural involution and to identify clinical characteristics that could predict the most severe or a particular type of sequelae.
Design, Setting, and Participants
Multicentric retrospective cohort study of images from 187 infantile hemangiomas that had not received systemic treatment and had follow-up pictures until regression that were selected from photographic files taken between 2003 and 2013 at 4 university hospitals with large vascular clinics in 3 different countries.
Main Outcomes and Measures
Outcome measures were the type of sequelae classified as residual telangiectasia, anetodermal skin, redundant skin, persistent superficial component, and the degree of sequelae ranging from 1 to 4.
A total of 184 hemangiomas were included. The overall incidence of significant sequelae was 101 of 184 (54.9%). The most common sequelae after involution were telangiectasias (145, 84.3%), fibrofatty tissue (81, 47.1%), and anetodermic skin (56, 32.6%). The average age at which hemangioma completed involution was 3.5 years. Superficial and deep hemangiomas left significantly fewer sequelae than combined hemangiomas (Mann-Whitney; superficial vs deep, OR, 1.6; 95% CI, 0.6-3.8; P = .81; superficial vs combined, OR, 3.3; 95% CI, 1.7-6.3; P < .001; deep vs combined, OR, 2.1; 95% CI, 0.9-5.1; P < .001). Hemangiomas with a step or abrupt border of the superficial component left more severe sequelae than those with a smooth border (χ2,OR, 3.4; 95% CI, 1.8-6.6; P < .001). Superficial hemangiomas with a cobblestone appearance or rough surface left more severe sequelae than those with a smooth surface (Kruskal-Wallis; α, 0.05; P < .001). Using multivariate analysis, combined hemangiomas with a superficial component and a step border were associated with more sequelae.
Conclusions and Relevance
In this retrospective study of sequelae in a large cohort of untreated infants, we quantified the prevalence of permanent scarring and identified clinical features predictive of permanent sequelae. Our observations provide useful information at a time when the treatment paradigm for hemangiomas has changed. Such knowledge may help primary care physicians predict the risk of sequelae and identify high-risk lesions to implement early treatment.
Infantile hemangiomas (IH) are the most common benign vascular tumors. They have a characteristic natural history marked by rapid proliferation in the first months of life, followed by slow involution that takes 5 to 9 years to complete. While all IH involute to some extent, they often leave permanent sequelae that may cause disfigurement. The extent, characteristics, and predictive factors for permanent sequelae after involution have not been well characterized. Poor recognition of the risk of sequelae by pediatric health care providers may encourage parents’ unrealistic expectations that the hemangioma will go away.
For decades, systemic corticosteroids were the mainstay of therapy for IH. Because of the poor risk-to-benefit ratio for oral corticosteroids, most IH were left untreated.1 Propranolol is now considered the treatment of choice for IH2-7 and has revolutionized our approach because complete regression without sequelae is obtained after 6 months of treatment in 60% of cases.1,8 Considering this high response rate, and better cosmetic outcomes compared with oral corticosteroids, indications for treatment of IH have been revised to include those at high risk of leaving permanent sequelae if left untreated. Identifying risk factors for permanent sequelae is therefore crucial for treatment decisions. This retrospective study of untreated infants quantifies the incidence of permanent sequelae and identifies predictive clinical features.
We retrospectively reviewed all images of 4 hospitals in 3 different countries from 2003 to 2013. Exclusion criteria were IH that had received systemic therapy.
Each center obtained approval from its institutional review board to participate in the study. Written informed consent was acquired from patients who continued to be followed at our clinics or whom we were able to contact by telephone. For patient lost to follow-up we had previous consent for keeping photographs in our files for later review. No compensation was given for participation in the study.
Using the online data capture program, Research Electronic Data Capture (REDCap), we collected data on the incidence and type of sequelae after hemangioma regression.
Data collected regarding IH characteristics is summarized in Table 1. We also assessed the characteristics of the lesion border and the surface. The border was classified as “progressive” if it rose gradually from normal skin and as “abrupt” or “stepped” if it was a clear, cut-off border. The surface of the superficial component was described as either smooth or cobblestoned. The height of the superficial component was also estimated from photographs.
Other data collected were age at which the hemangioma completed regression. A hemangioma was considered to have completely regressed if there was no change in appearance after 1 year of follow up, according to pictures.
The type of sequelae was collected from pictures and classified as: residual telangiectasia; anetodermal skin; redundant skin; or persistent change in color (Figure 1). We also quantified the degree of sequelae in a range from 1 to 4, from no sequela to severe sequelae that warranted consideration for surgical correction.
Statistical analysis was performed at Hospital de la Santa Creu i Sant Pau. Logistic regression was used to assess potential predictive factors for sequelae (none/minimal sequelae vs significant/severe sequelae). An odds ratio (OR) with corresponding 95% CIs and P values were calculated. The χ2 test was used for categorical variables. To study the correlation and dependence of variables, we used the Spearman correlation coefficient, Mann-Whitney test, and Kruskal-Wallis. All statistical analyses were performed using IBM-SPSS statistical software (version 22.0, SPSS Inc).
A total of 185 IH met inclusion criteria. Details regarding clinical characteristics of the IH and their sequelae are summarized in Table 1. The mean follow-up period was 4.1 years (median, 4 years; range, 1-12 years of age). Mean age at which no further involution was seen after 1 year of follow-up was 3.44 (median, 3 years; range, 1-7 years).
The incidence of significant to severe permanent sequelae after regression was 55%. The type of IH was significantly associated with the degree of sequelae (Kruskal-Wallis, P < .001). No or minimal sequelae were observed in 45 of 76 (59%) of the superficial IH, in 13 of 27 (48%) of deep IH, and in 25 of 81 (31%) of combined IH (χ2, P < .001).
The type of border of hemangiomas that had a superficial component (superficial and combined) was also significantly associated with the degree of sequelae (Kruskal-Wallis, P < .001). We found that 55 out of 78 (70.5%) IH with a step border left significant to severe sequelae compared with 33 out of 80 (41.3% ) of those with a smooth border (χ2, OR, 3.4; 95% CI, 1.8-6.6; P < .001) (Figure 2). Likewise, IH with a pebbled or cobblestoned surface left significantly more reduntant skin than those with a smooth surface (50/80 [62.5%] vs 38/78 [48.7%]) (Kruskal-Wallis, α, 0.05; P < .001) (Figure 3).
The height of the superficial component also correlated with the severity of sequelae (Spearman ρ, 0.316; P < .001). Size of the hemangioma also was related with sequelae: median size (none/minimal sequelae) 6 cm vs 9 cm (size for significant/severe sequelae) (Kruskall-Wallis, P = .002).
A multivariate analysis of superficial and combined IH (Table 2) showed that only border and type of hemangioma remained significant for significant/severe sequelae (both variables, P < .001). The IH with a step border had an OR (95% CI) of 5.76 (2.6-12.8) of leaving significant/severe sequelae and combined IH had an OR (95% CI) of 5.63 (2.5-12.4) (Hosmer and Lemeshow test; P = .73; area under the curve [AUC], 0.733).
The most frequent type of sequelae left by IH after involution were telangiectasias (145 of 172 [84.3%]), fibrofatty tissue (81 of 172 [47.1%]), and anetodermic skin (56 of 172 [32.6%]) (Figure 3). The least common were redundant skin (27 of 172 [15.7%]) and scar (22 of 172 [12.8%])
Infantile hemangioma characteristics also correlated with the type of sequelae. Deep and combined IH left more fibrofatty tissue than superficial IH (20 of 27 [74.1%] vs 51 of 80 [63.8%] vs 13 of 76 [17.1%]; χ2, P < .001) (Figure 2). Infantile hemangiomas with a superficial component with a cobblestoned appearance left more anetodermic skin than those with a smooth surface (37 of 80 [46.3%] vs 16 of 78 [20.5%]; χ2, P = .001). Infantile hemangiomas with a step border left more redundant skin (39 of 78 [28.2%] vs 14 of 80 [6.3%]; χ2, P < .001) and more anetodermic skin (50% vs 17.5%; χ2, P < .001) than those with a progressive border (Figure 1).
We were unable to find a correlation between IH location and the type of sequelae and the degree of sequelae owing to the limited number of IH at any particular location. However, 8 of the 11 IH (72%) on the tip of the nose and 9 of the 17 IH on the lips (52.9%) left significant to severe sequelae.
Data regarding pedunculated IH were not evaluated owing to the small number of cases. However these are known to require plastic correction owing to the inevitable fibrofatty residuum.
The results of our study show that more than half of untreated IHs have significant or severe sequelae. The most common sequelae were telangiectasias and fibrofatty tissue, present in more than 50% of cases.
We identified certain hemangioma features that constitute risk factors for sequelae and that determine the severity of sequelae. Apart from the type of hemangioma, the border and surface seem to be important factors. Sequelae were observed more often in combined IHs. On the other hand, IHs with a stepped border of the superficial component and those with a cobblestoned surface left more severe persistent sequelae than those with a progressive border and a smooth surface. As expected, ulcerated IHs always left a residual scar. The worst scenario for risk of sequelae, therefore, apart from an ulcerated IH, was a combined IH with a superficial component and a stepped border.
We also observed that the features of the IH determine not only the degree of sequelae but also the type of sequelae. Deep and combined IHs left significantly more fibrofatty tissue than superficial IH. Superficial IHs only left fibrofatty tissue if they were thick (data not shown). Infantile hemangiomas with a cobblestoned surface and a stepped border were more prone to leave anetodermic skin and redundant skin on regression than those with a progressive border and smooth surface. This may be important because fibrofatty tissue and redundant skin are sequelae that are readily visible and can only be corrected by surgery.
Our incidence of sequelae is similar to that obtained by Baulard et al.9 After studying 137 hemangiomas, these authors found residual lesions in 69% of cases. However, because they did not quantify sequelae, both mild and severe sequelae may have been included in the final analysis. The authors found that superficial nodular IHs left significantly more residual lesions than deep IHs. In contrast, in our series combined IHs left more severe sequelae than superficial and deep IHs. However, because Baulard et al9 did not present images in their article, it is not possible to determine if they used the term superficial nodular for hemangiomas that we call combined IH.
In our series, involution of IHs was complete by 3.5 years of age. These data are in line with those recently published by Darrow et al,1 who found that the involution of IH begins as the child approaches 12 months of age and is usually completed by the age of 4 years. These data contrast with the general belief and consensus shown in many dermatology textbooks and articles that involution is complete in 50 percent of cases by the age of 5 years, in 75 percent by the age of 7 years, and in 90 percent by the age of 9 years.10
This data gives primary care physicians better information not only to predict rate or prevalence of sequelae, but also to identify high-risk lesions and implement early treatment.
This study has several limitations. First, since it is a retrospective study without a systematic follow-up, in which photographs were used to perform analyses and measurements, it is possible that the IHs continued to improve after what we considered to be a final photograph. Second, visual assessment of border height might not always be exact. Third, selection bias may have occurred because small hemangiomas are less likely to be photographed and more likely to be lost to follow-up. And finally, the low number of cases prevented conclusions from being drawn about pedunculated IHs and about the specific locations of hemangiomas.
The information provided by our study may help clinicians and families in predicting the risk of sequelae and in the treatment decision-making process. A close follow-up during the first weeks of life is mandatory to detect early morphological changes that suggest a risk of significant sequelae and to consider therapy before unfavorable clinical features develop. In conclusion, this study demonstrates that more than half of IHs will leave significant or severe sequelae if left to involute spontaneously, and that their morphology may help predict the risk and nature of major sequelae. Furthermore, our data provide a well-characterized historical control cohort of children with untreated IH that can be used in future therapeutic interventional trials.
Corresponding Author: Eulalia Baselga, MD, Department of Dermatology, Hospital de la Santa Creu I Sant Pau, S. Antoni M Claret 167, 08025 Barcelona, Spain (email@example.com).
Published Online: August 17, 2016. doi:10.1001/jamadermatol.2016.2905
Author Contributions:Concept and design: Baselga, Roe, McCuaig.Acquisition, analysis, or interpretation of data: All Authors.Drafting of the manuscript: Baselga, Roe, McCuaig.Critical revision of the manuscript for important intellectual content: Baselga, Roe, Coulie, Muñoz-Garza, Boon, Hernández-Martin, Gich, Puig.Statistical analysis: Gich.Administrative, technical, or material support: Roe, Muñoz-Garza, McCuaig.Study supervision: Roe, Boon, Hernández-Martin.
Conflict of Interest Disclosures: Drs Baselga, McCuaig, and Boon have been involved in clinical trials of propranolol in the treatment if infantile hemangiomas sponsored by Pierre-Fabre Dermatology, and they have served as consultants for Pierre-Fabre.
Funding/Support: We had financial support from Pierre-Fabre Dermatology for a medical writer.
Role of the Funder/Sponsor: Pierre-Fabre Dermatology had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We are indebted to Carolyn Neway, language services coordinator at the Institut de Recerca Sant Pau, for her English grammar and style corrections. She was not compensated for her contributions. We thank the patients for granting permission to publish this information.
Create a personal account or sign in to: