Background
Myocarditis, an inflammatory disease of the myocardium, remains a clinical challenge in pediatrics as it challenges practitioners in all aspects including diagnosis, methods of intervention, and follow-up counseling [
1]. It is a not rare and heterogenous disease in the pediatric age group in both developing and developed countries [
2], as a study indicated the evidence of myocarditis in 12% of the autopsied adolescent and young adult patients with sudden death [
3]. Myocarditis can cause serious long-term morbidity in children including diminished cardiac function, cardiac failure and potentially necessitating aggressive circulatory support [
4]. It can also lead to the development of a chronic dilated cardiomyopathy (DCM), which is a common indication for cardiac transplantation in children older than 1 year [
5]. Though patients of all ages may be affected by myocarditis, teenagers and infants, especially children in the first year of life, account for the majority of cases [
6]. A recent review on the epidemiology of myocarditis addressed this bimodal age distribution of myocarditis patients without any conclusive reasons [
7].
One other characteristic feature of pediatric myocarditis is the multiple etiologies, including viral, immune-mediated, and toxin-mediated [
8]. However, the direct pathogen leading to myocarditis in many patients may remain undetermined or mistakenly assigned to viruses, especially PVD19 and HHV6 [
5,
9]. At the same time, the diagnosis of myocarditis can be challenging due to the wide spectrum of clinical signs and symptoms [
10]. Though the majority of patients with myocarditis would present to their emergency department with a recent previous illness, some patients may have no history of illness before development of symptoms ranging from chest pain, respiratory symptoms, to cardiogenic shock and sudden death [
4,
11,
12].
Even though the understanding of pathogenesis in acute myocarditis has been constantly improving, treatments for myocarditis remain controversial and generally focus on supportive care with attention to guideline-directed treatment of heart failure and arrhythmia [
10,
13]. Therapies directed towards modulating the immune responses in patients have been considered beneficial in order to antagonize the autoimmune injury to myocytes which exacerbates myocardial dysfunction [
14]. Immunosuppressive therapies using steroids in pediatric patients have been conducted in observational and controlled clinical trials but led to confronting conclusions [
15‐
17]. Similar outcomes were also observed in trials conducted on the adult population [
18‐
20].
Previous systematic reviews failed to conclude clinically significant improvements upon administration of immunosuppressive agents against acute myocarditis, largely due to the small number of subjects, especially pediatric studies [
17,
21‐
23]. At the same time, most of the included studies also focused on adult patients. Chen et al. [
23] included 8 studies. However only 3 out of those 8 focused on children. The only previous systematic review focusing on the effects of corticosteroids on children with acute myocarditis was done 19 years ago by Hia et al. [
17]. At this point, more observational and controlled trials have been conducted on pediatric patients, which provided us with more evidence on revising the effects of immunosuppression in children with myocarditis. Therefore, the objective of this systematic review and meta-analysis was to provide an updated evaluation of the effects of immunosuppression therapies on the outcome of acute myocarditis.
Material and methods
Data sources and literature search
In this systematic review and meta-analysis of randomized controlled trials, we performed a systematic literature search with no date limits using PubMed, Embase and Cochrane library without a language limitation. The search was performed to October 1, 2022. The search was performed with the following MeSH terms: “Myocarditis”, “Anti-Inflammatory Agents”, “Immunosuppressive Agents”, “Glucocorticoids” and “Adrenal Cortex Hormones”; and key words: “carditis”, “acute viral myocarditis”, “corticosteroid”, “steroid”, “prednisone”, “dexamethasone”, “hydrocortisone”, “methylprednisone”, “betamethasone”, “budesonide”, “fludrocortisone” and “mineralcorticoids”. The search strategy was as follows: (“Myocarditis” OR “carditis” OR “acute viral myocarditis”) AND (“Anti-Inflammatory Agents” OR “Immunosuppressive Agents” OR “Glucocorticoids” OR “corticosteroid” OR “steroid” OR “prednisone” OR “dexamethasone” OR “hydrocortisone” OR “methylprednisone” OR “betamethasone” OR “budesonide” OR “fludrocortisone” OR “mineralcorticoids”). In addition, two reviewers independently performed manual search based on references from these articles and other review articles.
Inclusion criteria
The criteria used in the selection of studies for inclusion in this systematic review were as follows: (1) randomized or quasi-randomized controlled trials (RCTs or quasi-RCTs), observational studies, or retrospective studies; (2) human studies; (3) trials included children no more than 18 years old; (4) studies consisting of a minimum of 2 arms, one arm receiving at least one type of corticosteroids and at least one only conventional anti-failure medication.
Both authors independently performed the literature research and evaluation of the retrieved data. After data extraction and collection, we resolved discrepancy in exclusion criteria by discussion based on literatures available in the database. The studies included by one author were examined by the other author. Quality of the clinical trials were assessed with Jadad score [
24].
The information extracted from each trial included: authors, year of publication, sample size, study duration (only for RCTs), mean age or age range, intervention, dosage of intervention, comparator, and dosage of comparator.
Assessment of methodological quality
We followed the guideline given in Cochrane Handbook for Systematic Reviews of Interventions [
25]. The likely magnitude and direction of the bias and whether these biases impacted the findings according to the criteria. The two authors (QY and SZ) independently assessed the quality of studies without blinding to authorship or journal. We resolved discrepancy by discussion.
Statistical analysis
All statistical analyses were conducted using Review Manager 5.4 statistical software. Dichotomous variables were analyzed using risk ratios (RRs) with 95% confidential interval (95% CI) based on the fixed effects model. Continuous variables were analyzed using mean difference (MD) with 95% CI. Statistical heterogeneity of the included trials was assessed by the
I2 statistics. The random effect model was applied when compared trials had high heterogeneity (
I2 > 50%). We assessed publication bias of included studies using Egger test [
26].
Discussion
Cardiovascular disease is still a major cause of adverse outcomes in young individuals even though a unique, but still insufficiently characterized risk profile of cardiovascular disease presents in this group [
32]. Among these diseases, the true incidence of acute myocarditis in pediatric patients is even more unpredictable owing to sub-clinical presentations estimated to be around 1 per 100,000 [
33]. This number can still be undermined as pediatric acute myocarditis can be misdiagnosed initially as respiratory illness such as bronchiolitis or pneumonia, as myocarditis in children may not be presented with classic signs and symptoms [
11]. The optimum treatment methods against pediatric acute myocarditis are still debatable, while the current treatment of myocarditis is still largely supportive. Corticosteroids, which are proposed to be immunosuppressive, have been widely used in immunological suppression therapeutics against myocarditis which has an autoimmune nature [
23]. Although several different systematic reviews [
23,
34,
35] have analyzed the role of corticosteroids in cardiovascular inflammatory disease treatment, the only systematic review focusing on pediatric acute myocarditis remained to be the one conducted by Hia et al. [
17], a study conducted 20 years ago. As more clinical trials and retrospective studies in recent years [
2,
28‐
31] provided us with more evidence of the clinical effectiveness of corticosteroid anti-inflammatory treatment of pediatric acute myocarditis, and thereby the basis of an updated analysis.
Our meta-analysis did not demonstrate a significant reduction in mortality rate of corticosteroid anti-inflammatory therapy compared to conventional anti-failure treatment without any other intervention (
P = 0.53; RR = 0.87; 95% CI = 0.58 to 1.33). On the other hand, a significant enhancement of patients’ left ventricular function was identified in the meta-analysis of LVEF in the pediatric patient treated with corticosteroids compare to patients on supportive therapy (
P = 0.0009; MD = 11.93%; 95% CI = 4.87% to 18.99%). Though high heterogeneity (
I2 = 82%,
P = 0.004) was observed with the inclusion of the study conducted by Camargo et al. [
27], removal of this study from the meta-analysis still yielded a significant recovery in pediatric patients’ LVEF measurements (
P < 0.0001; MD = 9.00%, 95% CI = 7.48% to 10.52%). Meanwhile, we also tried to conduct analysis on the clinical endpoint and LVEDD improvement in pediatric patients. However, in both meta-analyses, high heterogeneity was seen within 2 included studies. Therefore, we cannot draw any conclusion out of these meta-analyses.
As more and more recent reviews have pointed out that myocarditis is more common, and can be more severe in children than in adults [
5,
7,
12], we think it is necessary to evaluate the immunosuppressive treatment of pediatric myocarditis rather than only relying on current data which mixed the clinical outcomes in populations of all ages. Two decades ago, when the public awareness on acute myocarditis in pediatric patients was still limited, Hia et al. [
17] tentatively suggested that insufficient evidences can be used to conclude whether immunosuppressive therapy can lead to significantly improved outcomes in children with acute myocarditis. Since then, though Chen et al. [
23] took a look at the effects of corticosteroids in general patients without a specific age range, not a single literature provided an insight on the effectiveness of anti-inflammatory therapy against acute myocarditis in pediatric patients. Unlike Hia et al. [
17] which made a more qualitative measurement on improvement after immunosuppressive treatment, this systematic review is the first to provide insights on quantitative assessment of mortality and left ventricular function in children with acute myocarditis. At the same time, all the studies included in this systematic review are controlled studies which included at least 1 arm receiving only supportive treatment, while the majority of case studies [
36‐
38] included in Hia et al. [
17] had small sample sizes (no more than 10 patients), and lacked a control group, and thereby were likely to have biases in estimating the treatment effects.
Our study has several limitations. First, most of the studies included in this systematic review had small sample sizes (3 studies had a total sample size less than 50). The likelihood of overestimation of treatment effects is proportional to the percentage of small trials included in the meta-analysis. Adding the fact that not all included studies are RCTs, there could be potential issues with the methodological control of the available evidence. Second, in some of the meta-analyses performed in this study, significant heterogeneity was observed. The high heterogeneity, especially the comparison made for clinical endpoint after treatment, is highly likely due to the heterogeneous clinical features and patient background, such as the quality of care provided by the intensive care clinics in different countries and areas. Finally, none of the comparisons made in this study has included all trials and patients in this systematic review, as different studies provided different evidences to characterize the clinically important improvement of disease activity. This fact led to reduced sample sizes in all comparisons, and thereby increasing the difficulty of the assessment of bias in comparisons in which only 2 studies were included.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.