Oral Atenolol versus Propranolol in the Treatment of Infantile Hemangioma: A Systematic Review and Meta-Analysis.

Background: Infantile hemangioma (IH) is the most common benign vascular tumor of infancy. Propranolol is considered first-line therapy for IH. However, it is associated with side effects. Therefore, there was a need for alternative therapy. Atenolol, a selective b1-blocker may be free from such side effects. Hence, the present study aims to develop a more accurate estimate of the safety and efficacy of atenolol compared to propranolol in the treatment of IH. Methodology: A search of various literature databases (PubMed, Embase, Ovid, Scopus, Cochrane Central, CINAHL, Web of Science, and Google Scholar) was done to identify studies which compared propranolol versus atenolol in the treatment of IH. The combined odds ratio along with corresponding 95% confidence intervals (CIs) were evaluated using a fixed-effects model.
Results: A total of 300 articles were screened of which five studies including 116 patients in atenolol arm and 138 patients in the propranolol arm were analyzed. Atenolol was comparable to propranolol in terms of efficacy as no significant difference was seen between both the treatment arms in terms of hemangioma activity score (mean difference 0.25 [95% CI;‒0.21, 0.71]) and complete response (odds ratio [OR] =0.43; 95% CI; 0.17, 1.11; P = 0.08,). Atenolol therapy was better than propranolol in terms of safety, i.e., serious/potentially serious side effect, (OR = 0.11; 95% CI; 0.02, 0.51; P = 0.005) and wheezing/bronchial hyperreactivity (OR = 0.11; 95% CI; 0.02, 0.51; P = 0.005).
Conclusion: The present meta-analysis provides evidence that atenolol has got a comparable efficacy and better safety profile with propranolol. Copyright:

INTRODUCTION

Infantile hemangioma (IH) is the most common benign vascular tumor of infancy with the incidence of 5%–10%.[1] There occurs benign proliferation of endothelial cells in the form of telangiectasia or discoloration. Usually, IHs grow swiftly during the initial 3–12 months of age and may take around 3 to7 years to involute spontaneously.[2] Most of these IHs do not need intervention and resolve naturally. IHs will need intervention in the proliferative phase in 10%–15% of lesions in case of painful ulceration, bleeding, infection, visual/airway impairment, and cosmetic disfigurement if the anatomical location is on the face, glabella, nasal tip, etc.[345] Different medical management options are available, but there are no uniform guidelines regarding the management. These management options included therapies with systemic and intralesional corticosteroids, chemotherapeutic agents such as interferon, vincristine, bleomycin, and surgical treatment. However, these drugs had long-term adverse effects and a variable response rate. Then came the era of beta-blockers in the management of IHs after the serendipitous findings of Léauté-Labrèze et al. when he used propranolol to treat cardiopulmonary conditions in 2008.[6] Unfortunately, even the usage of propranolol is not risk-free. Adverse effects such as bronchial obstruction, hypoglycemia, sleep disturbances, seizures, hypotension, and gastrointestinal symptoms are present even with propranolol.[78] These can be attributed to the direct effects of β2 receptors blockade as well as its ability to cross blood–brain barrier due to its lipophilicity. Hence, there is a need for an alternative treatment with improved safety and efficacy which led to the evaluation of the effect of atenolol in the therapy of IH. In 2011, Raphael et al. first stated that two IH patients withdrew propranolol therapy because of bronchial hyperreactivity and sleep disturbances. Later, switching on to atenolol the authors stated outstanding results with no adverse effects.[9] Consequently, various studies have reported that oral atenolol can be an alternative therapy to propranolol in IH.[1011]

Nonselective β-blockers may lead to airway obstruction and severe exacerbations in asthmatic patients. This is because of the blockade of β2 receptors along with β1 receptors by nonselective β-blockers. However, cardioselective β-blockers such as atenolol carries less risk because of selective action on β1 receptors and can therefore be used in reversible airway disease.[12] Nonselective β-blockers interfere with the regulation of gluconeogenesis, glycogenolysis, and lipolysis. Being less lipophilic, atenolol tends to produce less sleep disturbances and long-term cognitive effects as it is less likely to cross blood-brain barrier.[11]

Although both atenolol (selective β1 blocker) and propranolol (nonselective beta-blocker) are being used in the treatment of IH, to date no meta-analysis has addressed regarding the comparative safety and efficacy of both these agents. Hence, the present meta-analysis aims to develop a more precise estimate of the safety and efficacy of atenolol compared to propranolol in the therapy of IH.

METHODOLOGY

The preferred reporting items for systematic reviews and meta-Analyses (PRISMA) guidelines were followed throughout the conduct of our systematic review.

Objectives

Evaluation of comparative efficacy and safety of atenolol versus propranolol in IH in terms of

Response to therapy (complete and partial response combined)

Hemangioma activity score (HAS)

Adverse/side effects.

Complete response

complete resolution or >90% resolution of IH. Telangiectasia and redundant tissue was also deemed as CR. CR will be considered if a score of 5 is attained in the physician global assessment scale.[13]

Partial response

Some reduction in size or change in color or consistency which did not meet the CR criteria or attainment of physician global score of 3 or above.

Response to treatment

For our study purpose, we defined response to treatment as any clinically significant response (either partial or complete response).

Selection

The studies included met the following inclusion criteria: (i) Diagnosis of IH. (ii) Compared atenolol and propranolol (iii) Study design: Comparative observational studies and randomized controlled trials. In case propranolol or atenolol initiated with other anti-hemangioma, in that case, the articles were excluded from the study. In case efficacy endpoints were not defined, the articles were excluded from the study.

Database search

We searched various literature databases (PubMed, Embase, Ovid, Scopus, Cochrane Central, CINAHL, Web of Science, Google Scholar and PGIMER, Chandigarh library) with keywords IH, atenolol, propranolol, and synonyms. There was no language restriction. Initially, the conception/idea for the article was made by SP. The database search was done independently by HK and SP. The database files were extracted using endnote and duplicates were removed. After the removal of duplicates, the database files were uploaded to Rayyan QCRI and the title and abstracts were screened by HK, VM, MA, and SP. Full text of eligible articles was further screened as per predefined inclusion/exclusion criteria by PS, MM, NP, and SP. In case of discrepancy, RS and BM were consulted and the issue was resolved. Final drafting and critical revisions were made by VM and PS.

Keywords

We used keywords such as hemangioma and its synonyms (congenital hemangioma, IH, strawberry birthmark, and capillary hemangioma), atenolol and its synonyms (tenormin, and normiten), and propranolol and its synonyms (avlocardyl, dociton, and dexpropranolol) for different database search. Boolean operators were used to combine these different search terms for the search of different databases.

Data extraction

Required data were extracted from the included articles using the predefined data extraction form. We extracted data which included first author's name, total number of cases as well as control, year of publication, sample population characteristics, treatment modalities, outcome of treatment, length of time as well as follow-up. Details of Characteristics of studies comparing Atenolol and Propranolol is given in Table 1.

Table 1

Characteristics of studies comparing atenolol and propranolol

Author, year	Study type/sample size	Atenolol regimen	Propranolol regime	Scoring outcome, time point at which evaluated	Efficacy	Adverse effects
De Graaf et al., 2013	Prospective Sample size: Atenolol=30, propranolol=28	0.5 mg/kg/day starting, 1-3 mg/kg/day therapeutic dose	1-2 mg/kg/dose in three daily doses	VAS, HAS Average duration was 11.5 months (0.5-28 months)	No statistical difference in VAS (P=0.34) or change in HAS (P=0.39)	Atenolol: Restless sleep-4, constipation-2, diarrhoa-2 Propranolol: Bronchial hyperactivity-4, hypoglycemia-2, Hypotension-1 and restless sleep-11
Bayart, 2017	Retrospective Noninferiority study Sample size: Atenolol=27, propranolol=53	Initial dose 0.25 mg/kg, 0.5-2 mg/kg therapeutic dose	Initial dose 0.33 mg/kg, increased 1 mg/kg for first week and then increased to 2 mg/kg in 3 doses daily	HAS Duration-12-15 months	No statistically significant difference HAS (P=0.60)	Atenolol – transient diarrhea 4 Propranolol - wheezing/bronchial hyperactivity-6 Diarrhea-2
Abarzua-Araya et al., 2014	RCT Sample size: Atenolol=13, propranolol=10	1 mg/kg/day in a single daily dose	2 mg/kg/d in 3 daily doses	CR PR No response End point – 6 months	CR – 6 months Atenolol - 53% Propranolol - 60% (P=0.68)	No adverse effects
R. Ashraf, 2019	RCT Sample size: Atenolol=20, propranolol=20	Initial dose 0.5 mg/kg once daily and increased to 1 mg/kg/day after 24 h, if tolerated well	Initial dose of 1 mg/kg/dose increased to 2 mg/kg/dose twice daily	PGA HAS Utrasongraphy-Change in volume End point – 9 months or CR	No statistically significant difference PGA5 at 9 months-(P=0.196) HAS (P=0.941)	Atenolol - increased stool frequency -9 Fall in BP-5 Sleep distubance-1 Propranolol - increased stool frequency -6, fall in BP-4
M. Dakoutrou, 2019	Prospective comparative study Sample size: Atenolol=26, propranolol=28	0.5 mg/kg/day once daily which was further increased up to 2 mg/kg/day	Initial dose of 1 mg/kg/dose increased to 2 mg/kg/dose twice daily	CR PR Relapse - average endpoint was till resolution - 7 months	No statistically difference atenolol CR 84.6% versus propranolol CR 96.2% Relapse - 11.5% atenolol versus 10.7% Propranolol	Atenolol - diarrhea-2 sleeping difficulties-2 Propranolol hypotension-4

CR: Complete response, PR: Partial response, PGA: Physician global assessment, HAS: Hemangioma activity score, VAS: Visual Analogue Scale, RCT: Randomized controlled studies, BP: Blood pressure

Statistical analysis

In the case of continuous data (HAS, Change in HAS), the mean difference was calculated using R. In the case of dichotomous data (complete response, partial response, and adverse events), OR was used as appropriate for the meta-analysis. We also quantified the effect of heterogeneity using the I2 statistic, I2 values of 25, 50, and 75% were considered as low, moderate, and high estimates, respectively. Depending on statistical and clinical heterogeneity, we used fixed effects model (Mantel-Haenszel method) to calculate pooled effect estimates.

Risk of bias

The risk of bias of the studies included was assessed using the “Cochrane risk of bias tool for randomized controlled studies (RCTs)” in case of randomized controlled trials and Newcastle-Ottawa scale in case of observational studies.

RESULTS

We searched databases (a total of 300 articles) and a total of 174 articles were retrieved after the removal of duplicates. From these, after the title and abstract screen, 163 articles were excluded, and 11 articles were chosen for further full-text evaluation for possible inclusion. Of these 11 studies, five studies were included in our meta-analysis.[310111314] Six articles were omitted after reading full text. Of these six articles, four were single-arm studies which evaluated atenolol only.[15161718] Study characteristics are shown in Table 1. The PRISMA flow chart of the included studies is shown in Figure 1.

Figure 1

PRISMA (2009) flow chart

Efficacy parameters

Two studies (n = 116) compared atenolol and propranolol in terms of HAS and change in HAS from baseline. No significant difference was observed between both treatment arms with regard to HAS and change in HAS from baseline. Data are shown in Figure 2. A total of three studies (a total of 60 patients in the atenolol arm and 57 patients in the propranolol arm) described response to therapy (either complete or partial response). No significant difference was seen between the two arms in treating IH's. Two studies compared atenolol and propranolol in terms of relapse after treatment, no significant difference was seen in both arms in treating IHs. Data showed in Figure 3.

Figure 2

Forest plot - Comparison of Hemangioma activity score both groups.

Figure 3

Forest plot-comparison of response and relapse to therapy

Safety profile

A total of five studies (a total of 116 patients in the atenolol arm and 138 patients in the propranolol arm) were included to compare serious/potentially serious side effect that occurred in the atenolol versus propranolol groups while treating IHs. In terms of serious/potentially serious side effects, atenolol therapy was better than propranolol in treating IHs (odds ratio [OR] =0.11; 95% confidence interval [CI] 0.02, 0.51; P = 0.005). As the heterogeneity among studies is low, we used fixed-effect model (χ2 = 0.02, I2 = 0%, P = 0.99, df = 2). Data are shown in Figure 4. Four studies were included to compare Wheezing/Bronchial hyperreactivity that showed atenolol therapy was better than propranolol with respect to safety profile in treating IHs (OR = 0.11; 95% CI; 0.02,0.51; P = 0.005). In terms of diarrhea, sleep disturbances, and hypoglycemia, no significant difference between atenolol and propranolol was observed. Data are shown in Figure 4. Other adverse effects mentioned in the atenolol arm are constipation and acrocyanosis in one patient.[11] Excessive crying in both arms.[13]

Figure 4

Forest plot-Comparison of adverse effects between both groups

RISK OF BIAS

Among the two RCTs included, risk of bias was high among the domains of allocation concealment and random sequence generation. Risk of bias was unclear in the case of blinding of participants and personnel (performance bias). In other domains, risk of bias was either low or unclear. Newcastle-Ottawa scale [Table 2] was used to study the quality of nonrandomized studies.

Table 2

Risk of bias (nonrandomized studies) Newcastle-Ottawa Scale

Author	Selection	Comparability	Outcome
Bayart, 2017	***	*	***
Graff, 2013	***	*	***
Dakoutrou, 2019	**	*	***

Each study has been categorized into 3 criteria: selection, comparability and outcome and stars have been awarded to each item based on the quality. 3 or 4 stars in selection domain AND 1 or 2 stars in comparability AND 2 or 3 stars in outcome domain indicate that the study is of good quality. 2 stars in selection domain AND 1 or 2 stars in comparability domain AND 2 or 3 stars in outcome/exposure domain indicate fair quality study. Poor quality is 0 or 1 star in selection domain OR 0 stars in comparability domain OR 0 or 1 stars in outcome/exposure domain.

Publication bias

As the number of studies included in this review is <10, evaluation of publication will not have sufficient power to detect a real asymmetry in the funnel plot.[19] Hence, publication bias evaluation using funnel plot was not conducted Details of Risk of Bias is given in Figure 5.

Figure 5

Cochrane risk of bias tool for randomized controlled studies

DISCUSSION

In the past, steroids were used in the treatment of IH as they are believed to inhibit angiogenesis, though the exact mechanism remains unclear.[20] However, the usage of steroids was associated with many adverse effects.[21] Serendipitous discovery of propranolol in IH has overcome this drawback. Since then plenty of RCT, meta-analysis has shown its benefits over steroids, due to which propranolol has become the first line of management of IHs. However, even propranolol is not completely safe and associated with different adverse effects as mentioned in Table 1. Atenolol (selective beta-blocker) was first tried in patients who did not tolerate propranolol in the therapy of IH, exhibited similar results with fewer adverse effects.[9] The exact mechanism by which propranolol is effective in IH is not clear. The catecholamine's norepinephrine and epinephrine act on the β1 (situated in the heart and kidneys) and β2 (peripheral blood vessels, skeletal muscle, and lungs) receptors. Norepinephrine stimulates hypoxia-inducible factor (HIF) 1-alpha (HIF-1-alpha), which causes an increase in vascular endothelial growth factor (VEGF), MMP-2, and MMP-9 levels in tumor cells. Nonselective beta-blocker, propranolol impedes the action of catecholamine's in the cell, thereby decreasing VEGF and MMP levels.[2223] The present concept of propranolol in treating IH includes effect on endothelial cells, vascular tone, angiogenesis, and apoptosis.[24] The study by Stiles et al. suggests that propranolol blocks VEGF receptor-2 (VEGFR-2) reduces proliferation and migration of cells and leads to apoptosis.[25]

In terms of efficacy, our study does not demonstrate any significant difference between atenolol and propranolol in terms of HAS, change in HAS from baseline, response to therapy, partial response rate, and CR rate.

Comparative safety

In this study, there was no difference in the occurrence of any adverse event (irrespective of severity, total numbers of the patients suffering from adverse effects) between the two arms; however, less number of patients in the atenolol arm suffered from serious/potentially serious adverse events. Wheezing/bronchial hyperreactivity which is considered a serious adverse event was less in the atenolol group in comparison to the propranolol group. This can be attributed to the selective action of atenolol on β2 receptors, thereby lowering the risk of bronchospasm.[14] Atenolol also offers the advantage of including IH patients with obstructive bronchial pathology in the trial whereas such patients would have been excluded in few studies using propranolol to treat IH.

The major adverse events noted in the atenolol arm were gastrointestinal symptoms and sleep disturbances. However, fewer patients experienced adverse effects when compared to the propranolol group. The major adverse events noted in the propranolol arm were hypoglycemia, bronchial hyperreactivity, hypotension, mild side effects, restless sleep, constipation, and diarrhea. There was no significant difference between both the groups in terms of any adverse effects, diarrhea, sleep disturbance, hypotension, and hypoglycemia. Hence, atenolol can be a preferred agent compared to propranolol in the treatment of IHs. Both the groups had comparable efficacy, but atenolol exhibited better safety profile when compared to propranolol. Furthermore, atenolol can be easily used in cases where propranolol is contraindicated. The above findings in this meta-analysis reveal that both propranolol and atenolol have comparable efficacy with better safety profile in the case of atenolol, with special reference to the occurrence of serious/potentially serious adverse events. Another additional advantage of atenolol is that it requires less frequent dosing because it has a longer half-life compared to propranolol.[26] This can increase the patient compliance with atenolol, particularly the pediatric age group where medication administration is a quite troublesome process; however, propranolol needs to be taken 2–3 times daily. In the studies, included in our meta-analysis, it is observed that the dose of propranolol is comparatively higher than that of atenolol. Most of the authors have started atenolol with 0.25 mg/kg/day[11] or 0.5 mg/kg/day,[10] slowly increased to 1 mg/kg/day to 3 mg/kg/day[10] if well tolerated whereas in case of propranolol it is 1–2 mg/kg/day in divided doses.[3] This variation in dose might have failed to show a better efficacy of atenolol when compared to propranolol. Hence, for any therapy to be considered as successful should fulfill both the safety and efficacy criteria. Keeping in mind both the parameters atenolol is considered better than propranolol in treating IH.

Our study also has few limitations. To conclude that atenolol is better than propranolol in terms of safety and efficacy, the quality of the study design chosen is quite important. Limitations in all studies were small number of participants and long-term follow-up was lacking. RCT are the best to include in a meta-analysis. A key, but the unavoidable drawback was the heterogeneity of the studies included in the study. Due to the lack of enough RCTs, we had to include RCTs as well as prospective and retrospective study designs.

There is lack of adequate studies regarding the effect of different doses of atenolol in treating IH. There is lack of enough data regarding the exact age to initiate the therapy, optimal dosage and duration of therapy as well as the criteria for discontinuing the therapy. Therefore, large, multicenter, and well-designed randomized controlled trials are warranted to draw an exact conclusion regarding the safety and efficacy of atenolol in treating IH compared to propranolol.

CONCLUSION

The present meta-analysis provides evidence that atenolol has got a comparable efficacy and better safety profile with propranolol.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.