Impact of congenital heart disease on clinical outcomes of oral propranolol therapy for Infantile hemangiomas: a propensity score-matched analysis

Abstract

Background

Oral propranolol is the first-line treatment for Infantile hemangioma (IH). However, there is a lack of studies indicating whether the congenital heart disease (CHD) makes a difference in oral propranolol therapy (OPT). Previous studies have only confirmed the safety and efficacy of OPT in asymptomatic CHD children.

Objective

This study aimed to evaluate and compare the clinical outcomes in pediatric patients with CHD and those without CHD following OPT and to assess risk factors that may contribute to recurrence.

Methods

A total of 310 children who received OPT for IH between January 2020 and January 2023 were included in a retrospective study. The study aimed to assess demographic data, clinical symptoms, auxiliary examinations, and treatment effects. To compare clinical outcomes between children with and without CHD, one-to-one propensity score matching (PSM) was applied.

Results

Among the 310 patients, 192 (61.94%) had CHD. After propensity-matching analysis, in the presence of CHD, the age at treatment withdrawal was significantly higher (15.0 months vs. 12.0 months, P < 0.05), the treatment duration was longer (12.0 months vs. 10.0 months, P < 0.05), and the recurrence rate was greater (32.9% vs. 5.3%, P < 0.001). No significant difference in the degree of regression was observed between the CHD and non-CHD groups.

Conclusion

Pediatric patients with CHD exhibit a later age at treatment withdrawal, a longer duration of therapy, and a higher recurrence rate compared to those without CHD.

Introduction

Infantile hemangioma (IH) represents the most common vascular tumor in children, with an estimated incidence ranging from 2 to 10% [1]. Oral propranolol therapy (OPT) is currently recognized as the first-line treatment for IH [2], achieving an efficacy rate exceeding 90% while maintaining a favorable safety profile [3,4,5]. However, 13.5 to 18% [6, 7] of the patients experienced recurrence after withdrawal, which indicates that the effect of propranolol is not irreversible. Recurrence may affect the confidence and adherence to the subsequent treatment. The premature treatment withdrawal is thought to elevate the risk of tumor recurrence, yet the optimal timing for withdrawal and the duration of therapy remains undefined. Therefore, identifying risk factors associated with recurrence of IH and developing individualized treatment protocols, including guidelines for the timing of medication discontinuation are crucial. This study aims to minimize recurrence resulting from insufficient treatment duration while avoiding unnecessary prolongation of therapy.

Prospective studies have shown that OPT can be safely performed in asymptomatic children with congenital heart disease (CHD), demonstrating a good to excellent response to IH [8]. However, there is a lack of studies indicating whether the CHD makes a difference in OPT.

As mentioned above, this study aimed to evaluate the clinical outcomes of OPT in IH children with CHD versus those without CHD, and to identify risk factors for recurrence.

Methods

Settings and children

This study was approved by the Institutional Research Ethics Board of Children’s Hospital of Chongqing Medical University (Date: 2024/No: 257), and complied with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. As the study was retrospective in nature and therefore did not change the clinical management of patients involved, the Institutional Research Ethics Board of Children’s Hospital of Chongqing Medical University waived the requirement for informed consent. This retrospective cohort study included all patients who received OPT for IH at the Children’s Hospital of Chongqing Medical University between January 2020 and January 2023. The included patients met the diagnostic criteria established by the International Society for the Study of Vascular Anomalies (ISSVA) [9]. Patients were excluded if they had received other treatments for IH before or during OPT; they had previously used propranolol for other medical indications such as arrhythmia or hypertension; the patients with contraindications to propranolol; they did not cooperate with the doctor’s advice; or the clinical data was incomplete.

Oral propranolol therapy (OPT)

The initial dose for hospitalized patients was 0.5 mg/kg/d, which was increased to a maximum dose of 2 mg/kg/d after 1 week if no significant adverse effects, and the initial dose for outpatients was 2 mg/kg/d, which is the maximum dose. The medication was administered twice daily, either during meals or within 30 min postprandially. Complete regression of the tumor or patient age greater than 1 year, with no significant changes observed over a 2-month period, may serve as criteria for treatment discontinuation. All patients were monitored through internet-based outpatient visits or hospital consultations for 6 to 12 months after withdrawal to evaluate the long-term effectiveness.

Study design

The data collected for each patient included (1) demographic data: sex, premature birth (< 37 weeks), multiple gestations, birth weight, delivery, feeding type, age at treatment start, age at treatment withdrawal; (2) clinical symptoms: multifocal, size, location, morphologic subtype, segmental and ulceration; (3) auxiliary examinations: color Doppler ultrasound, CT or MRI, abdominal ultrasound and echocardiography (CHD); (4) treatment effects: degree of regression, recurrence, time of recurrence, treatment methods following recurrence, and effectiveness of treatment after recurrence.

Clinical outcomes

We assessed clinical outcomes using several indicators: recurrence, degree of regression, age at treatment withdrawal and duration of therapy.

Recurrence is defined as the reappearance of superficial red spots, vascular changes, or other color alterations at the original tumor site, as well as an increase in the tumor’s surface area, volume, or texture. From the perspective of color Doppler ultrasound, recurrence is indicated by significant growth at the primary tumor site or the reemergence of abundant blood flow signals within the tumor [6, 7, 10].

The degree of regression at the withdrawal was independently assessed by two physicians using pre- and post-treatment photographs of the lesions, and the efficacy of treatment was evaluated using the Achauer classification system at the time of treatment cessation, categorized as follows: excellent (75%-100% regression), good (50%-75% regression), moderate (25%-50% regression), and poor (0%-25% regression) [11].

Statistical analysis

In our study, data were entered using Excel software and all statistical analyses were conducted with R software (version 4.3.1), along with MSTATA software (www.mstata.com). Univariate analysis and propensity score matching (PSM) with post-matching analysis were conducted. PSM was performed using the nearest-neighbor method with a caliper of 0.25, matching cases of CHD to control subjects in a 1:1 ratio, based on patient characteristics including sex, birth weight, delivery, premature birth, multiple gestations, feeding, age at treatment start, multifocal, location, morphologic subtype and segmental. The quality of the match was evaluated using the absolute standardized mean difference, with an acceptable threshold set at ≤ 0.20. Following PSM, statistical analyses were performed to compare patients with good outcomes and poor outcomes using the matched data.

The analysis included measures such as mean, median, standard deviation, range, and quartiles for continuous variables and frequency tables for categorical variables. Data distribution was assessed using normality tests, and appropriate descriptive statistics methods were applied to both normally and non-normally distributed variables.

For comparison between groups of categorical data, the Fisher exact test was used when expected frequencies were less than 5; otherwise, the Chi-squared test was applied.

Results

Study population

During the study period, a total of 310 patients met the inclusion criteria. The clinical data of these patients before OPT are presented in Table 1. Among them, 192 patients had CHD (61.94%), while 118 did not (38.06%). Among the 192 patients with CHD, 79 had a patent foramen ovale (PFO), 75 had an atrial septal defect (ASD), 2 had a ventricular septal defect (VSD), and 2 had a patent ductus arteriosus (PDA). Additionally, 34 patients had two or more coexisting abnormalities. The frequencies of delivery method, age at treatment start, and segment type were significantly higher in the CHD group compared to the control group (all P < 0.05). However, the two groups were comparable in terms of sex, birth weight, premature birth, multiple gestations, feeding type, morphologic subtype, location, and number. It is important to note that all the patients with CHD in this study are mild CHD cases and did not experience common CHD-related symptoms, such as palpitations, dyspnea, cyanosis, growth retardation, or poor exercise tolerance, during the treatment, and none required medical intervention.

Table 1 Comparison of clinical data of children with CHD before and after propensity matching

The objective of conducting propensity score matching (PSM) was to minimize allocation bias, thereby better reflecting the impact of concurrent CHD on the efficacy of OPT treatment (Table 1). Following a 1:1 PSM, 76 patients from the CHD group (39.58%) were successfully matched with 76 patients from the non-CHD group. The baseline characteristics of the matched patients were highly balanced (Figs. 1 and 2).

Fig. 1

The standardized mean difference of covariates between different groups before and after propensity score matching. Unadjusted: before matched covariate distribution; Adjusted: after matched covariate distribution

Fig. 2

Balance of covariates before and after propensity score matching. Unadjusted Sample: before matched covariate equalization; Adjust Sample: after matched covariate equalization

Comparison of clinical outcomes

Table 2 presents the treatment effects of OPT for patients with and without CHD. In the analysis of all patients, background differences between the two groups were unavoidable. The CHD group was significantly associated with a higher recurrence rate (22.4% vs. 6.8%, P < 0.001), a lower regression rate (72.9% vs. 88.1%, P = 0.001), and a longer duration of OPT (12.0 months vs. 9.5 months, P < 0.001). However, there was no significant difference in age at treatment withdrawal between the CHD and non-CHD groups.

Table 2 Comparison of clinical outcomes after OPT between non-CHD group and CHD group

Comparisons with propensity score-matched patients (Table 2) revealed that, in the presence of CHD, the age at treatment withdrawal was significantly higher (15.0 months vs. 12.0 months, P < 0.05), the treatment duration was longer (12.0 months vs. 10.0 months, P < 0.05), and the recurrence rate was greater (32.9% vs. 5.3%, P < 0.001). Nonetheless, there was no significant difference in the degree of regression between the CHD and non-CHD groups.

Discussion

Although numerous studies have examined the clinical outcomes of OPT for IH, and recent advancements in treatment strategies have significantly improved the prognosis of IH, recurrence after withdrawal remains a frequent occurrence [6, 7]. Timely recognition of adverse outcomes of IH treated with propranolol after withdrawal can provide effective clinical guidance to obtain optimal treatment options and get the best possible outcome. Recent studies on factors influencing IH recurrence after OPT, in addition to inadequate treatment course (stopping treatment before 12 months of age), have mainly concentrated on the hemangioma’s intrinsic characteristics, such as its location in the head, neck, and face (especially in the tip of the nose), its multiple, segmental distribution, and mixed or deep involvement [7]. However, the role of the cardiovascular system in IH has been largely overlooked. Only a limited number of studies have demonstrated the safety and efficacy of OPT in children with IH and asymptomatic CHD [8], and the potential impact of CHD on OPT outcomes remains unexplored, so we designed this study.

CHD is the most common congenital anomaly in neonates. It is primarily diagnosed through clinical presentation and physical examination, with echocardiography serving as the gold standard for definitive diagnosis [12]. CHD is typically classified hemodynamically into left-to-right shunt (acyanotic), right-to-left shunt (cyanotic), and non-shunt (acyanotic) types. The left-to-right shunt type is the most prevalent, with common conditions including ASD, VSD, PDA and PFO [13].

To reduce allocation bias and obtain a more precise assessment of the impact of CHD on OPT outcomes, we employed PSM. Following adjustment for the propensity score, all covariates, except CHD, were adequately balanced and comparable across groups. This approach allows the use of non-randomized subgroup data to explore the relationship between treatment factors and outcomes, thereby enhancing the robustness and validity of the findings. In our study, we found a significant association between the presence of left-to-right shunt CHD and the age at treatment withdrawal, the duration of OPT, and the recurrence rate. This study included a total of 310 patients with IH, of whom 192 (61.94%) had CHD, all characterized by left-to-right shunt types. Although CHD is not a risk factor for IH [14], more than half of the patients in this study had CHD. We hypothesize that this relatively high proportion may be influenced by potential confounding factors, such as the single-center design and the limited sample size, which could introduce population bias. Furthermore, the predominance of left-to-right shunt CHD in this cohort may reflect the lower overall prevalence of right-to-left shunt CHD. Due to sample size limitations, our study did not include any cases of right-to-left shunt CHD. Therefore, studies examining risk factors for clinical outcomes in IH undergoing OPT should prioritize the comorbidity of CHD and assess the risk–benefit ratio in relation to therapeutic outcomes.

The mechanism of action of OPT for IH is not fully elucidated and may involve multiple pathways. Notably, numerous studies have reported that CHD is associated with elevated levels of various biochemical markers and molecular factors that align with the known mechanisms of propranolol’s action in treating IH. For instance, Yasser et al. [15]. reported that serum levels of interleukin 6 (IL-6) and vascular endothelial growth factor (VEGF) were significantly elevated in children with CHD compared to their healthy peers, with higher levels observed in the cyanotic CHD (CCHD) group compared to the acyanotic CHD (ACHD) group. Ji et al. [16]. demonstrated that propranolol inhibits the proliferation of hemangioma-derived endothelial cells (HemECs) and induces apoptosis by decreasing VEGF expression, upregulating pro-apoptotic genes such as p53 and Bax, and downregulating the anti-apoptotic gene Bcl-xL. Furthermore, Chim et al. [17]. found that propranolol reduces hypoxia-inducible factor 1-alpha (HIF-1α) expression and inhibits hemangioma development by diminishing VEGF activity via the HIF-1α-VEGF-A angiogenesis axis, leading to downregulation of PI3K/Akt and p38/MAPK signaling pathways. Additionally, Reiner et al. [18]. reported that children with left-to-right shunt CHD exhibited activation of the neuroendocrine system, as evidenced by elevated levels of norepinephrine and plasma renin activity. Propranolol is known to competitively bind to adrenergic receptors with catecholamines such as norepinephrine, thereby blocking their pro-tumorigenic effects in hemangiomas [19, 20]. Propranolol also reduces the expression of other growth factors, including IL-6, which are elevated via the β-adrenergic receptor signaling pathway following stimulation by epinephrine or norepinephrine [21], thus further inhibiting hemangioma progression. Additionally, the renin-angiotensin system (RAS) has been proposed as an independent pathogenic factor in IH. DeJong et al. [22]. found that proliferative IH expressed various components of the RAS, concluding that propranolol may inhibit hemangioma growth through RAS inhibition.

Therefore, we hypothesize that patients with IH and concurrent left-to-right shunt CHD exhibit higher levels of VEGF, IL-6, norepinephrine, and plasma renin activity compared to those without CHD. These elevated factors may lead to increased proliferative activity of the tumor, thereby diminishing the inhibitory effects of propranolol on tumor proliferation. As a result, the rate of regression may slow, extending the time required for complete resolution or optimal therapeutic effect, which in turn prolongs the duration of OPT. Similarly, upon reaching standardized criteria for discontinuation of therapy, residual hemangioma tissue may remain present due to the diminished effects of propranolol, while levels of VEGF, IL-6, norepinephrine, and renin remain elevated. This persistence may result in renewed proliferative activity in the residual tissue, leading to the recurrence of hemangiomas. Although right-to-left shunt CHD is less prevalent, it is associated with even higher levels of VEGF, IL-6, norepinephrine, and renin activity, which may correlate with longer durations of OPT and increased recurrence rates [15, 18]. However, it is important to note that this study is a retrospective analysis, and the inclusion of elevated CHD-related biochemical markers is intended to help explain this phenomenon. This hypothesis is based on a reasonable assumption regarding the role of propranolol in the treatment of IH.

This study also provides guidance for medication regimens and monitoring strategies. Based on prior clinical experience, treatment withdrawal may be considered for children under 1 year old with complete tumor regression, regardless of the presence of CHD. For children older than 1 year whose tumors have not fully regressed, treatment withdrawal may be considered if no significant changes observed over a 2-month period. While the mean age of treatment withdrawal for children with CHD was 15 months old, this does not mean that all patients must continue OPT until 15 months old. Although extending the duration of OPT may reduce recurrence risk further, it is unlikely to benefit patients with a low risk of recurrence. In such cases, extending treatment would only unnecessarily prolong therapy. Therefore, we recommend that specialists and parents closely monitor children with CHD until 15 months old, with regular monitoring for recurrence. For patients who have had treatment withdrawn for 6 to 12 months, routine monitoring should be maintained.

Limitation

This retrospective cohort study was conducted at a single center with a relatively small sample size, which could have introduced potential biases in both the data collection process and subsequent analysis. Moreover, the study’s definition of recurrence, based on both clinical presentation and Doppler ultrasound results, is subject to interobserver variability, which is an unavoidable limitation. In addition, since the study exclusively included patients with mild CHD, its findings do not provide direct evidence regarding the potential impact of moderate-to-severe CHD on IH outcomes. Further research is needed to evaluate whether similar trends apply to this population. Furthermore, the study did not assess the impact of right-to-left shunt CHD on clinical outcomes associated with OPT, and the influence of specific CHD types on OPT efficacy remains unexplored. In order to comprehensively evaluate the effects of various shunt types in CHD on the clinical outcomes of OPT, future studies should prioritize more extensive prospective cohort studies. Although the clinical efficacy and safety of OPT have been established, current understanding of its mechanisms of action primarily derives from laboratory and animal models.

Conclusion

Based on available evidence, this study represents the first clinical investigation into the impact of CHD on OPT for IH from the perspective of clinical efficacy. We successfully identified associations between the age at treatment withdrawal, the duration of therapy, and the recurrence in relation to CHD. For children with CHD, if treatment withdrawal occurs before 15 months old due to meeting the cessation criteria, close monitoring should be maintained until 15 months, after which routine follow-up can be implemented. This research contributes to the development of personalized withdrawal protocols and monitoring strategies, enhancing the ability of healthcare providers to effectively communicate potential complications to patients and their families.