Infantile hemangioma (IH) is the most common vascular tumor in childhood [1].
As reported in a prospective study, IHs occur in up to 4.5% of infants, and risk factors include placental anomalies, prematurity, and low birth weight [2].
Generally, IHs undergo an initial proliferative phase, followed by an involutional phase and a final involutional phase [3]. Although IHs are self-limiting, a significant minority of untreated IHs can leave permanent disfigurement after involution. Therefore, in patients with IH, opportunities to intervene and prevent sequelae in the initial stage are lost.
Corticosteroids used to be the first-line treatment for IHs until 2008, when Léauté-Labrèze and colleagues first reported the efficacy of propranolol in the treatment of IHs [4]. Subsequent reports have demonstrated the efficacy of systemic propranolol in the treatment of problematic IHs [5, 6]. Now, oral propranolol has become the first choice for problematic IHs [7].
In addition, intralesional drug injection, pulsed dye laser (PCL), and topical timolol are also treatment options in some IHs [7]. However, residual lesions requiring additional treatments are common after these treatments. HIs commonly arise in the head and neck region, where the disfigurement significantly affects appearance.
Now, the sequelae of IHs gradually attract the attention of doctors. Currently, some studies have been carried out focusing on the long-term sequelae of IHs. The objectives of this systematic review were to thoroughly identify the sequelae of IHs, determine specific risk factors for sequelae, and learn how to improve the sequelae of IHs. The authors set out to answer the following questions: "What is the prevalence of skin sequelae in IHs?" "What are the factors that affect the appearance of skin sequelae in IHs?" and "How to improve the skin sequelae of IHs?"
| Methods |
> Search strategy and information sources
The authors conducted a systematic review following PRISMA guidelines [8]. They registered this systematic review in the PROSPERO database. Details of the protocol can be accessed at http://www.crd.york.ac.uk/PROSPERO/ (Registration number CRD42021251262). The authors searched from 2008 to the present in the following databases: PubMed, EMBASE, Cochrane Library, and Clinicaltrials.gov. They restricted the publication language to English.
The search chain was built in two parts: infantile hemangioma and sequelae. The full search strategy is included in the supplementary information. A reference list of all independently selected articles was examined to identify additional studies excluded in the initial search.
> Eligibility criteria and selection process
All interventional and observational studies that focused on long-term cutaneous sequelae of IHs were included. Articles were excluded if they were a review, meta-analysis, or case report involving fewer than four patients.
Studies focusing only on the short-term efficacy and safety of treatments, but not on skin sequelae, were excluded . The titles and abstracts of the studies were reviewed by three independent authors. Potentially eligible articles were then accessed for text review for eligibility. The reference lists were also manually screened for potentially relevant articles. Discrepancies were adjudicated by the senior author.
> Quality evaluation
Three reviewers independently assessed risk of bias in the included studies. Discrepancies were resolved through discussion. The Cochrane Collaboration tool was used to evaluate randomized trials [9], and MINORS (methodological index for non-randomized studies) was used to evaluate the quality of non-randomized intervention studies [10]. A modified “Newcastle-Ottawa Quality Assessment Scale” (NOS) was used to assess the quality of observational studies [11]. They added two items, “determination of infantile hemangiomas” and “definition of sequelae,” in NOS. For those cohort studies that did not have a comparison group, they removed the item “comparability of cohorts on the basis of design or analysis.”
> Data collection
Data collection was carried out using a pre-piloted template system. The extracted data included basic information, study methodology, anatomical location of IH, treatment, follow-up time point, cutaneous sequelae of IHs, specific factors affecting the occurrence of sequelae, and any other important information that had not been determined before.
> Statistical analysis
The general synthesis approach for this review will be narrative. If three or more studies report the same outcome, in the same way, they perform a meta-analysis for the given effect measure.
| Results |
> Search
The electronic database search yielded 4448 reports. After screening titles and abstracts, 62 were potentially eligible and five additional studies were identified from the reference search. After reading the full texts, 17 studies were included in this systemic review. There were no unpublished data that satisfied the authors’ criteria.
> Study characteristics
Of those 17 studies, 15 were cohort studies, among which most of them were uncompared studies. Two of these studies were nonrandomized intervention studies. The female-to-male ratio was 1.5:1 to 4.7:1. The most frequent locations of the sequelae were the head and neck. The type of HI is mainly superficial or mixed.
The mean duration of follow-up ranged from 1 to 7.38 years. Three studies reported long-term cutaneous outcomes of untreated IHs; Five studies evaluated the sequelae of patients treated with oral beta blockers, two studies followed the outcome of patients treated with LCP, and another eight studies reported the sequelae of IHs treated with various methods.
The results of the risk of bias assessment are shown in supplementary information.
> Sequels
Half of the studies reported the frequency of different types of sequelae, respectively, while the remaining half only reported an overall rate of sequelae.
The overall rate of sequelae varies from 5.3 to 93.5%. In most of the included studies, diagnoses were based on clinical manifestation and physical examination. Some studies evaluated residual lesions with ultrasound and MRI. None of those patients underwent a skin biopsy.
The most frequently mentioned sequelae were telangiectasia, fibroadipose tissue and pigment change.
Other cutaneous sequelae identified included scarring, skin atrophy, anetoderma, redundant skin, and hypotrichosis.
> Risk factors associated with sequelae
Risk factors associated with long-term sequelae were further classified into patient demographics, hemangioma characteristics (including type, appearance, ulceration, size, and location), and treatment factors.
Chang et al reported that patient age at initial consultation was negatively associated with the degree of sequelae [16]. Beyond age, no clear links between sequelae and patient demographics such as gender and race were reported in the included studies.
Five of the included studies showed a statistically significant correlation between the type of IH and sequelae; four of them reported that mixed hemangiomas led to significantly more residual lesions than other subtypes [18, 20, 25, 26]; One study reported that superficial nodular hemangiomas led to significantly more residual lesions than deep nodular hemangiomas [14].
The higher incidence of sequelae in mixed hemangiomas may be due to their tendency to result in both telangiectasias and fibroadipose tissue. The difference is that superficial nodular hemangiomas in their study are defined as IHs that had epidermal invasion, including mixed and superficial IHs.
Three studies investigated the relationship between the type of IH and the severity of sequelae, two of them found that significant and severe sequelae were observed more frequently in mixed IHs [13, 16], while Yu et al found that superficial hemangiomas resulted in significant or severe residual lesions in the majority of cases [20].
The difference is probably that their assessment of severity was self-assessed but not by objective indicators. Furthermore, Yu et al recognized that the superficial hemangiomas they included were mainly large, which left a wide range of skin damage affecting appearance and was classified as "significant."
Chang et al reported that complete regression of HI occurring in the central facial region was significantly lower than for those in the perifacial region [15, 16]. And this is the first and only study that reveals the association between sequelae and location.
There seems to be a consensus that larger IHs are associated with a higher rate of sequelae. However, most of the included studies did not measure IH size accurately mainly because IHs have multiple morphology; Most of the included studies did not report the correlation between IH size and sequelae.
Baselga et al mentioned that hemangioma size was related to sequelae, but did not explain how they measure the size of IHs [13]. Two of the included studies showed that the thickness of the superficial component is associated with sequelae [13, 25].
Furthermore, IH morphology (focal, segmental, and indeterminate) was rarely noted in the included studies. Only one study reports that focal IHs resulted in more sequelae requiring reconstructive surgery than segmental and indeterminate ones [25].
The characteristics of IHs determine not only the frequency and degree of sequelae but also the type of sequelae. Three studies reported a higher rate of fibrofatty tissue in deep and mixed IHs than in superficial ones [13–15].
The reason may be that both deep and mixed IHs contain subcutaneous invasion. A cobbled surface and a stepped edge of superficial or mixed HIs may be risk factors for anetodermic skin. Five studies found that ulcerated IHs generally left a scar [13, 14, 16, 17, 25].
In most studies that compared sequelae between treated and untreated patients, untreated patients left more sequelae than treated patients [21, 24, 25]. Jiang et al compared the sequelae of superficial IHs treated with laser and observation. They found that the sequelae in the laser group were significantly less than in the observation group. Chelleri et al also reported that patients treated with laser therapy alone had the lowest rate of residual lesions [24].
In a retrospective cohort study evaluating the sequelae of IHs treated with LCP, only 35 of 657 patients had sequelae [17]. Among all included studies, patients treated with laser had the fewest sequelae. However, the authors cannot conclude that LCP is the best to avoid sequelae because adverse effects such as ulceration and bleeding are commonly observed after LCP treatment, mainly due to excessive energy efficiency.
Several studies evaluated the effect of oral beta-blocker therapy in the treatment of IH, and were of the same opinion that β-blockers can decrease the rate of sequelae [12, 20, 24, 25, 28]. However, there was no comparison group in most studies.
A study comparing a cohort of patients treated with propranolol and a historical cohort of patients with IHs found that treatment with oral propranolol can reduce the risk of sequelae requiring correction. However, sequelae can be seen after treatment and subsequent therapy is still necessary to improve the aesthetic result. One study reported that surgical outcomes between medically treated patients and untreated patients had no significant differences [19].
> How to improve these skin sequelae?
Telangiectasia is generally defined as minimal to mild because it can resolve spontaneously or be improved by non-invasive procedures such as topical timolol or LCP [12, 20, 27, 28]. Pigment change can also improve spontaneously or be improved by LCP [12, 17, 18, 23, 28]. Fibrofatty tissue and redundant skin rarely regress spontaneously; Therefore, the treatment of these sequelae is mainly surgical resection [13, 19, 20].
| Statistic analysis |
The results are shown as reported in the original article. A meta-analysis was not performed due to significant heterogeneity between studies. They did not perform comparative statistical analysis as the results were reported using different outcome measures.
| Discussion |
The authors found that cutaneous sequelae in involuted IHs are very common.
Furthermore, the risk factors are very varied. Some sequelae, such as telangiectasia and pigment change, can resolve spontaneously, but some sequelae, such as fibrofatty tissue, require surgical intervention. Several limitations were identified in the included studies.
The prevalence of sequelae varied widely among the included studies because their study designs differed widely. Some of these studies only included HIs in a specific body part or type. Furthermore, there are many possible treatment methods, so the treatment of IHs can be multidisciplinary. Although sequelae have been reported from many therapeutic procedures, including oral propranolol therapy, LCP, and surgical interventions, no control group was included, the authors found no randomized controlled studies comparing cutaneous sequelae between intervention and observation, for so the evidence is insufficient.
For all included studies, selection bias could lead to an overestimation of the prevalence of HI sequelae because parents whose children suffer from HI sequelae might have been more eager to participate.
The follow-up time biased the result. There are excluded studies whose follow-up duration was not long enough or only focused on short-term adverse effects. However, the included studies also differ greatly in the duration of follow-up.
In Chen’s study, patients were evaluated before treatment and followed up for at least one year after the last treatment. At this time, IHs may still be in the proliferation or involution phase, meaning that these lesions are likely to completely regress over time. In addition, sequelae such as telangiectasia and pigment change may resolve over time, so if the follow-up time is long enough, the rate of sequelae may be lower.
Definitions of sequelae were missing or inconsistent. Most of the included studies did not illustrate the definitions of the different types of sequelae. A sequel can have various nomenclatures such as chromatosis and hyperpigmentation, discoloration and hypopigmentation, atrophic scarring and skin atrophy. Usually, there was no exact definition or legend in the texts, so it is not known if they have the same sequel. Furthermore, some studies did not include patients with sequelae that they believed could improve spontaneously such as telangiectasia and pigment change, which would underestimate the frequency of sequelae.
A limitation of this review is that it is difficult to quantitatively compare and summarize study results due to different patient inclusion criteria, duration of follow-up, definitions, and types of assessment used for sequelae. Furthermore, no randomized controlled trials were included because randomized controlled trials on the topic would be quite difficult to conduct.
Furthermore, they did not consider psychological health outcomes in this systematic review. There remains a need to evaluate long-term psychological and behavioral sequelae in patients with IH.
Although many studies focusing on IHs have been published, few reports have discussed long-term cutaneous sequelae. A review from last year has already preliminarily summarized several sequelae [29]. However, this systematic review summarized the frequency, risk factors, and treatments of sequelae in IHs.
| Conclusion |
Skin sequelae in involuted IHs are very common and can become underestimated causes of permanent disfigurement.
Some sequelae may improve spontaneously, but some people require invasive procedures to obtain a satisfactory aesthetic result.
The authors’ results have identified specific major factors leading to persistent sequelae of HI. More attention should be paid to prevention, early recognition and improvement of sequelae in patients with IH.
| Comment |
The present study highlights that infantile hemangiomas are the most common vascular tumors in childhood. These may retract spontaneously, but treated and untreated patients may suffer permanent skin sequelae. On the other hand, it is highlighted that the sequelae in involuted infantile hemangiomas are very common, and prevention, recognition, and improvement of these cutaneous sequelae are important to achieve a better quality of life in affected childhoods.
