Predictors of cardiac disease in duchenne muscular dystrophy: a systematic review and evidence grading

Abstract

Background

Duchenne muscular dystrophy (DMD) is a rare disease that causes progressive muscle degeneration resulting in life-threatening cardiac complications. The objective of this systematic literature review was to describe and grade the published evidence of predictors of cardiac disease in DMD.

Methods

The review encompassed searches of Embase, MEDLINE ALL, and the Cochrane Database of Systematic Reviews from January 1, 2000, to December 31, 2022, for predictors of cardiac disease in DMD. The certainty of evidence (i.e., very low to high) was assessed using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework.

Results

We included 33 publications encompassing 9,232 patients with DMD. We found moderate- to high-quality evidence that cardiac medication (i.e., ACE inhibitors [enalapril and perindopril], β-blockers [carvedilol], and mineralocorticoid receptor antagonists [eplerenone]) are significantly associated with preserved left ventricular ejection fraction (LVEF), left ventricular end-systolic volume (LVESV), and left ventricular circumferential strain (LVCS). DMD mutations in exons 51 and 52 were found to be significantly associated with lower risk of cardiomyopathy; deletions treatable by exon 53 skipping and mutations in the Dp116 coding region with improved LVEF and prolonged cardiac dysfunction-free survival; and exons 45–50 and 52 with early left ventricular systolic dysfunction (low/very low-quality evidence). We found high-quality evidence that glucocorticoids (deflazacort) are significantly associated with preserved LVEF and improved fractional shortening (FS), and low-quality evidence that glucocorticoids (deflazacort, prednisone, and/or prednisolone) are associated with improved ejection fraction (EF) and lower risk of cardiomyopathy, ventricular dysfunction, and heart failure-related mortality. Full-time mechanical ventilation was found to be significantly correlated with LVEF (low-quality evidence), muscle strength with FS (low-quality evidence), and genetic modifiers (i.e., LTBP4 rs10880 and ACTN3) with LVEF, lower risk of cardiomyopathy and left ventricular dilation (low-quality evidence).

Conclusion

Several sources of cardiac disease heterogeneity are well-studied in patients with DMD. Yet, the certainty of evidence is generally low, and little is known of the contribution of non-pharmacological interventions, as well as the impact of different criteria for initiation of specific treatments. Our findings help raise awareness of prevailing unmet needs, shape expectations of treatment outcomes, and inform the design of future research.

Background

Duchenne muscular dystrophy (DMD) is a rare, X-linked neuromuscular disease caused by mutations in the DMD gene resulting in progressive muscle degeneration, loss of independent ambulation, and life-threatening cardiac and respiratory complications [1]. In the past 50 years, advances in the medical management of DMD have dramatically improved prognosis. Children born in the 1960s seldom survived beyond their second decade of life, which may be compared with recent estimates of life-expectancy of patients receiving current standards of care–including glucocorticoid therapy, spine surgery, and mechanical ventilatory support–of about 30 years [2]. Yet, the unmet medical need and burden of illness remains substantial [3,4,5,6].

Following the introduction of the routine use of mechanical ventilatory support in advanced stages of the disease, cardiac involvement has emerged as one of the leading causes of morbidity and mortality in patients with DMD [7]. Dystrophin deficiency in the heart leads to myocardial damage which manifests as cardiomyopathy, resulting in compromised myocardium, potentially fatal rhythm abnormalities, and clinical heart failure. Features of cardiac dysfunction include sinus tachycardia, myocardial fibrosis, and left ventricular enlargement and systolic dysfunction [8]. However, symptoms of cardiac dysfunction (e.g., dyspnea, abdominal pain, fatigue, and inability to perform activities of daily living) are frequently unrecognized in individuals with DMD due to the severe physical impairment associated with the disease, particularly in adults [9]. For that reason, regular follow-up and monitoring is essential to the care strategy of cardiac disease in DMD [8].

Despite their importance for clinical management and prognosis, presently there is a lack of a comprehensive, up-to-date synthesis of predictors of cardiac disease in children and adults with DMD. These include, for example, pharmacological treatments (e.g., glucocorticoids, angiotensin-converting enzyme [ACE] inhibitors, and β-blockers), genetic modifiers associated with dystrophin deficiency and muscle degeneration (e.g., latent TGFβ binding proteins [LTBPs] and the ACTN3 gene encoding α-actinin-3), and DMD mutations [8, 10, 11]. The objective of this systematic literature review was to describe and grade the published evidence of predictors of cardiac disease in DMD.

Methods

Search strategy and selection criteria

The bibliographic searches were performed in the following databases: Embase, MEDLINE ALL, and the Cochrane Database of Systematic Reviews. We considered all records published between January 1, 2000 (to ensure relevance to current care practices) and December 31, 2022. We used the search terms “Duchenne muscular dystrophy” as a Medical Subject Heading or free text term, in combination with variations of the term “predictor” (full search strings are provided in eTable 1, eTable 2, and eTable 3 in the Additional file 1). We considered studies of any type, reported in any language, that included male patients diagnosed with DMD exposed to any treatments. We did not consider editorial letters or conference abstracts (as they lack details essential for meaningful synthesis) and did not formally include identified systematic reviews (but screened their reference lists for potential publications). We performed this systematic literature review using guidance from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [12].

Screening and data extraction

Screening was conducted independently by two investigators (EL and SA). Conflicts were designated to be resolved by a third reviewer (HL). We extracted the following data elements from included articles: Author; title; study year; geographical setting(s); study design; site(s)/data source(s); study period; sample population characteristics; case ascertainment; pharmacological interventions (incl. number of exposed, dose, and duration of exposure); outcome measures(s); method of analysis; and outcome results. We considered evidence of predictors of cardiac disease, defined as any factor–either endogenous (e.g., DMD mutations or genetic modifiers) or exogenous (e.g., pharmacological interventions, including exposure to ACE inhibitors and β-blockers)–significantly associated with cardiac health and function in DMD. We only considered mortality outcomes if the cause of death was established to be related to cardiac involvement. We did not seek to synthesize sources of cardiac variability stemming from cardiac features or assessments (e.g., magnetic resonance imaging or blood biomarkers). Upon identification of the relevant literature, two investigators (EL and SA) systematically screened reference lists of all included publications with the aim to identify additional records of interest not captured by the search strategy.

Level of evidence

We assessed the certainty of the identified evidence of predictors of cardiac disease in DMD using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework [13]. GRADE rates the overall certainty of evidence based on design limitations, risk of bias, consistency of the results across available studies, the precision of the results, directness, and likelihood of publication. The tool comprises of four levels of evidence, also known as certainty of evidence or quality of evidence: (1) very low (i.e., the true effect is probably markedly different from the estimated effect), (2) low (i.e., the true effect might be markedly different from the estimated effect), (3) moderate (i.e., the authors believe that the true effect is probably close to the estimated effect), and (4) high (i.e., the authors have a lot of confidence that the true effect is similar to the estimated effect). Per the GRADE manual, two investigators (EL and AA) independently provided an initial rating of all included records based on study type. Next, the certainty of evidence at the outcome level was rated down for issues or limitations pertaining to study limitations (e.g., risk of bias due to failure to develop and apply appropriate eligibility criteria, flawed measurement of exposure and/or outcome, failure to adequately control for confounding, and incomplete follow-up), inconsistency of results (i.e., an unexplained heterogeneity of results), imprecision (i.e., a low degree of certainty in reported point estimates), indirectness of evidence (stemming from, for example, differences between populations, differences in interventions, and/or differences in outcome measures), and publication bias (i.e., a systematic under- or over-estimation due to selective publication of studies), and/or rated up in case of a large magnitude of effect, a dose response, or if confounders are likely to minimize the effect. Finally, each investigator independently provided an overall GRADE certainty rating of each outcome and study [13]. All GRADE ratings were subsequently reviewed and confirmed by HL and KW.

Results

Upon completion of the bibliographic searches, we identified a total of 3,590 articles, of which 984 were duplicates. After full-text review of 85 records, 33 articles [14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46] were ultimately included. Figure 1 presents the PRISMA flow chart of the study selection process. Identified studies encompassed 9,232 patients with DMD from 11 countries (i.e., Brazil, Canada, China, Germany, France, Greece, Italy, Japan, South Korea, the United Kingdom (UK), and United States of America (US) (two multi-national studies [26, 28] did not explicitly disclose included countries) (Table 1). Yet, as some countries were represented by more than one study each, we cannot rule out that a proportion of patients might have been included more than once. In total, 15% (5 of 33) of articles described results from randomized research, 6% (2 of 33) from prospective cohort studies, 76% (25 of 33) from retrospective cohort studies, and 3% (1 of 33) from case series.

Fig. 1

PRISMA diagram of the selection process of the included publications

Table 1 Characteristics of included studies

Predictors of cardiac disease in DMD

Cardiac medication

We identified three randomized controlled trials (RCTs) reporting evidence of benefits of cardiac medication on left ventricular ejection fraction (LVEF) in patients with DMD (Table 2). Specifically, in the RCT and open-label extension by Duboc et al. [21], encompassing 57 French children with DMD (mean age: 11 years; range: 9–13), the proportion with LVEF < 45% after 60 months of follow-up was significantly lower among those treated with ACE inhibitors (perindopril) (initiated at a LVEF > 55%), in some cases administered in combination with β-blockers (agents not reported), compared with no ACE inhibitor treatment (4% vs. 28%, p = 0.02). Similarly, in the RCT by Silva et al. [40], treatment with ACE inhibitors (enalapril) (initiated at a LVEF > 50%) was found to be associated with slower myocardial fibrosis (MF) progression identified on cardiovascular magnetic resonance among 42 Brazilian patients (39 with DMD and three with Becker muscular dystrophy [BMD], a milder allelic condition also caused by in-frame mutations in the DMD gene; mean age: 12 years, range not reported) across 24 months of follow-up. Lastly, in the RCT by Raman et al. [37], involving 42 US participants with DMD (median age: 15 years, inter-quartile range [IQR] not reported), those treated with mineralocorticoid receptor antagonists (eplerenone) (initiated at a LVEF > 45%) were found to have significantly lower decline in LVEF after 12 months of follow-up (median change from baseline: -1.8% [treated] vs. -3.7% [untreated], p = 0.032). However, some cases were concurrently receiving ACE inhibitors (agents not reported), angiotensin receptor blockers (ARBs) (agents not reported), β-blockers (agents not reported), and/or loop diuretics (furosemide). Significant differences were also noted for left ventricular end-systolic volume (LVESV) (median change from baseline: -1.64 ml [treated] vs. 4.07 ml [untreated], p = 0.034), as well as left ventricular circumferential strain (LVCS) (median change from baseline: 1.0% [treated] vs. 2.2% [untreated], p = 0.020). Additionally, we identified one RCT reporting evidence of benefits of cardiac medication on heart rate (HR), PQ-interval, and P-wave in patients with DMD. Specifically, Dittrich et al. [20] examined the effects of combined treatment with ACE inhibitors (enalapril) and β-blockers (metoprolol) in a German cohort encompassing 38 children with DMD (mean age: 11 years, range: 9–13). At 19 months after randomization, patients receiving enalapril and metoprolol (initiated at a left ventricular fractional shortening [LVFS] ≥ 30%) were found to have significantly improved HR, P-wave, and PQ-interval compared with those treated with placebo (all p < 0.05).

Table 2 GRADE assessment of studies of predictors of cardiac disease in DMD

We identified one uncontrolled clinical trial, Kwon et al. [29], describing the efficacy of ACE inhibitors (enalapril) or β-blockers (carvedilol) in 23 Korean patients (22 with DMD and one with BMD; mean age: 13 years, range not reported). After 20.1 months of follow-up, fractional shortening (FS), left ventricular end-systolic diameter (LVESD), and left ventricular free wall (LVFW) systolic myocardial velocity were significantly improved compared with baseline values in patients treated with either enalapril or carvedilol (criteria for initiation not reported) (all p ≤ 0.046).

Looking at results from identified observational research, significantly improved LVEF was reported by Aikawa et al. [14] in their study of 34 Japanese patients (21 with DMD and 13 with BMD) treated with ACE inhibitors (cilazapril and enalapril) (initiated at different LVEF levels), in some cases in combination with β-blockers (bisoprolol) and/or ARBs (agents not reported); Jefferies et al. [24] in 69 US patients (62 with DMD and seven with BMD) treated with ACE inhibitors (enalapril, captopril, and lisinopril) (initiated at an LVEF ≥ 55% or evidence of left ventricular dilation) and/or β-blockers (carvedilol and metoprolol); and Kelley et al. [26] in a multi-national cohort comprising of 147 patients with DMD treated with β-blockers (agents not reported) (criteria for initiation not reported) (in some cases in combination with ACE inhibitors, ARBs, diuretics agents, anti-arrhythmics, and/or inotropes [agents not reported]).

We identified one retrospective cohort study, Porcher et al. [35], examining the impact of prophylactic use of ACE inhibitors (perindopril, enalapril, ramipril, or lisinopril) on the risk of hospitalization for heart failure among 576 French patients with DMD with normal left ventricular function. Compared with no treatment, ACE inhibitors (initiated at a LVEF ≥ 55%) were associated with a significant risk reduction (HR: 0.50, 95% CI: 0.26 to 0.99, p < 0.05; adjusted HR: 0.16, 95% CI: 0.04 to 0.62, p < 0.05; and HR propensity score-based analysis: 0.37, 95% CI: 0.20–0.68, p < 0.05).

Viollet et al. [44] showed ejection fraction (EF) improvement compared to baseline 12 months before initiation of therapy in 42 patients receiving either ACE inhibitor (lisinopril) only, or ACE inhibitor plus a β-blocker (metoprolol or atenolol) (p < 0.0001); however, ACE inhibitor plus β-blocker was not superior than ACE inhibitor alone.

Further evidence of benefits of cardiac medications in DMD include improved FS [25]; HR [25, 32]; left ventricular end-diastolic diameter (LVEDD) [24, 25]; LVEDD (Z-score) [25]; left ventricular end-diastolic volume (LVEDV) [17]; LVESD [29]; LVESV [17]; LVFW systolic myocardial velocity (cm/sec) [29]; left ventricular mass (LVM) [17]; left ventricular myocardial performance index (MPI) [24]; left ventricular sphericity index [24]; and death, deterioration of heart failure and severe arrhythmia (composite endpoint) [32]. One study also examined the impact of the timing of initiation of ACE inhibitors (cilazapril and enalapril) in terms of LVEF in patients with DMD or BMD [14].

DMD mutations

We identified four observational studies reporting evidence of effects of DMD mutations on cardiac disease in DMD (Table 2). Specifically, in a retrospective cohort study of 69 patients with DMD and BMD (of which 47 had genetic analysis of their deoxyribonucleic acid [DNA]), Jefferies et al. [24] investigated the association between DMD mutations and age at onset of cardiomyopathy (defined as EF < 55% or left ventricular dilation). Mutations involving exons 12, 14, 15, 16, and 17 (type not reported) were all shown to be associated with onset of cardiomyopathy, and exons 51 and 52 appeared to be associated with lower risk of cardiac involvement. In the retrospective cohort study by Yamamoto et al. [45], encompassing 181 Japanese children and adults with DMD, patients with mutations in the Dp116 coding region were found to have a significantly longer cardiac dysfunction-free survival than those with other dystrophin isoform deficiencies (p = 0.022). Moreover, Cirino et al. [19] found that exons 45, 46, 47, 48, 49, 50, and 52 were associated with early left ventricular systolic dysfunction (all p < 0.044) among 40 Brazilian patients with DMD. Finally, in a case series by Servais et al. [39], comprising of 35 non-ambulatory French patients with DMD, LVEF was estimated at 50.3% in patients with deletions treatable by exon 53 skipping (DMD-53), 63.6% in patients with mutations not treatable by exon 53 skipping (DMD all-non-53), and 66.7% in patients with deletions not treatable by exon 53 skipping (DMD del-non-53) at end of follow-up (DMD-53 vs. DMD all-non-53: p = 0.018; DMD-53 vs. DMD del-non-53: p = 0.028).

Genetic modifiers

We identified two observational studies reporting evidence of effects of genetic modifiers on cardiac disease in DMD (Table 2). Specifically, Barp et al. [16] studied genetic modifiers for dilated cardiomyopathy in a sample of 178 Italians with DMD and found that patients with the LTBP4 rs10880 CC/CT genotype had a higher risk of dilated cardiomyopathy compared with the TT genotype (p < 0.027). Moreover, Nagai et al. [34] described cardiac dysfunction (defined as LVEF < 53%) and left ventricular dilation (defined as LVEDD > 55 mm) by genotype in 77 Japanese patients with DMD. Median cardiac dysfunction-free survival was 13.4 years and 15.3 years (p = 0.041) in patients with the ACTN3 null genotype and ACTN3 positive genotype, respectively (HR: 2.78, 95% CI: 1.04 to 7.44, p < 0.05). The left ventricular dilation-free survival rate was different between patients with the RR, RX, and XX genotypes (p = 0.023) and lower in patients with the ACTN3 null genotype compared with the ACTN3 positive genotype (HR: 9.04, 95% CI = 1.77 to 46.20, p < 0.05).

Glucocorticoid exposure

We identified six observational studies reporting evidence of benefits of glucocorticoids on LVEF in patients with DMD (Table 2). Specifically, Biggar et al. [18] found the proportion of patients with LVEF < 45% at 18 years of age to be lower among those receiving glucocorticoid therapy (deflazacort) compared with no glucocorticoid therapy (10%, vs. 58%, p < 0.001); Mavrogeni et al. [33] estimated the median LVEF at end of follow-up (duration not reported) at 53% and 48% in patients with and without glucocorticoid treatment (deflazacort), respectively (p < 0.001); Kelley et al. [26] reported of improved LVEF in patients treated with glucocorticoids (agents not reported); Silversides et al. [41] estimated the proportion of patients with LVEF < 45% at end of follow-up (duration not reported) at 5% with glucocorticoid therapy (deflazacort) and 58% without glucocorticoid therapy (p = 0.001); and Schram et al. [38] estimated the mean annual rate of change in LVEF across follow-up at -0.43% for patients treated with glucocorticoids (deflazacort or prednisone) and -1.09% for those with no glucocorticoid treatment (p = 0.0101). Additionally, Tandon et al. [42] found a significant negative association between duration of glucocorticoid therapy (deflazacort or prednisone) and LVEF in a retrospective cohort study of 98 US patients with DMD. Specifically, an increased glucocorticoid treatment duration was associated with an LVEF decline of 0.43% per year of treatment (p < 0.0001).

We identified six observational studies reporting evidence of benefits of glucocorticoids on FS in patients with DMD. Specifically, Biggar et al. [18] estimated the mean FS at 18 years of age at 33% and 21% in patients with and without glucocorticoid treatment (deflazacort), respectively (p < 0.002); Houde et al. [23] estimated the mean FS at end of follow-up (duration not reported) at 30.8% in participants treated with glucocorticoids (deflazacort) (in some cases in combination with ACE inhibitors [agents not reported]) compared with 26.6% in those not receiving glucocorticoid therapy (p < 0.05); Markham et al. [30] found the mean FS to be higher in patients with DMD treated with glucocorticoids (deflazacort or prednisone) compared to those who were not treated (34% vs 26%, p < 0.001); Schram et al. [38] estimated the mean FS at end of follow-up at 29% for patients treated with glucocorticoids (deflazacort or prednisone) and 23% for untreated participants (p = 0.0043); Silversides et al. [41] estimated the mean FS at 33% and 21% with and without glucocorticoid treatment (deflazacort), respectively (p = 0.002); and Trucco et al. [43] estimated the mean annual rate of decline at 0.53% in those treated with glucocorticoids (deflazacort or prednisone) and at 1.17% in patients not treated (p < 0.01). Additionally, in two separate retrospective cohort studies, Markham et al. [30, 31] evaluated the frequency of ventricular dysfunction (defined as FS < 28%) after glucocorticoid treatment. The authors found that those receiving glucocorticoids (deflazacort or prednisone) had a significantly lower risk of ventricular dysfunction compared to untreated patients (all p ≤ 0.02).

We identified five observational studies describing the effects of glucocorticoids on cardiomyopathy outcomes. Specifically, Houde et al. [23] found the proportion of participants with dilated cardiomyopathy (defined as FS < 28% or LVEDD > 95th percentile) to be lower among those treated with deflazacort (in some cases in combination with ACE inhibitors [agents not reported]) than those who were untreated (32% vs. 58%, p < 0.05). Similarly, in a multi-national cohort comprising of 5,345 patients with DMD, Koeks et al. [28] reported that 42% and 60% of patients ≥ 20 years of age with and without glucocorticoid exposure (deflazacort, prednisone, or prednisolone), respectively, had evidence of cardiomyopathy at end of follow-up (p = 0.0035). The prevalence of cardiomyopathy among patients previously treated with glucocorticoids was 62%. Moreover, Schram et al. [38] investigated the risk of cardiomyopathy (defined as EF < 45%) in 86 Canadian patients with DMD and found glucocorticoids (deflazacort or prednisone) to have a protective effect (HR: 0.38, 95% CI: 0.16 to 0.90, p = 0.0270). In line with these results, Trucco et al. [43] found that patients not treated with glucocorticoids (deflazacort or prednisone) had a higher risk of cardiomyopathy (defined as FS < 28%) compared with their treated counterparts (neither group exposed to any cardiac medication) (HR: 2.2, 95% CI: 1.1 to 4.6, p < 0.05). In the study by Barber et al. [15], involving 462 US participants with DMD, a significant inverse association was observed between glucocorticoid duration and timing of onset of cardiomyopathy (defined as FS < 28% or EF < 55%). Specifically, the probability of developing cardiomyopathy decreased by 4% for every year of treatment with glucocorticoids (p < 0.001). In contrast, the study by Kim et al. [27], comprising of 660 US patients with DMD, reported an increased risk of cardiomyopathy (defined as FS < 28% or EF < 55%) in participants treated early with glucocorticoids compared to those who were untreated (HR: 2.1, 95% CI: 1.2 to 3.5, p < 0.01), as well as in those treated early vs. late (HR: 2.1, 95% CI: 1.2 to 3.5, p = 0.01).

We identified one observational study reporting evidence of effects of glucocorticoids on heart failure (HF)-related mortality. Specifically, among 86 Canadians with DMD, Schram et al. [38] found that the proportion of patients who died from HF-related causes was 0% in those treated with glucocorticoids (deflazacort or prednisone) and 22% in untreated patients (p = 0.0010) (all of whom also received cardiac medication).

Further evidence of benefits of glucocorticoids in DMD include improved EF [23]; LVEDD [30, 38]; LVEDV [33]; LVESD [38, 41]; meridional wall stress (mWS) [30]; systolic blood pressure [41]; summed rest score (SRS) [46]; velocity of circumferential fiber shortening (VCFc) [30]; and ventricular dysfunction [31].

Muscle strength

We identified one retrospective cohort study examining the relationship between muscle strength and FS in patients with DMD (Table 2). Specifically, Posner et al. [36] presented evidence of significant correlations between subjective arm and leg strength and total quantitative muscle testing, respectively, and FS (p ≤ 0.01), among 77 US children and adults with DMD.

Ventilation support

We identified one retrospective cohort study describing an effect of mechanical ventilation on LVEF (Table 2). Specifically, Fayssoil et al. [22] reported a significant inverse relationship between full-time mechanical ventilation and annual rate of LVEF decline among 101 French adults with DMD (p = 0.012).

Rating of the certainty of the evidence

Per the manual of GRADE, we initially attributed included RCTs a high rating, observational studies a low rating, and case reports a very low rating. Next, we downgraded the rating for Aikawa et al. [14], Jefferies et al. [24], and Kajimoto et al. [25], Kwon et al. [29], and Silva et al. [40] due to indirectness (as the studies also included patients with diseases other than DMD); Duboc et al. [21] and Dittrich et al. [20] due to inconsistency of results; and Jefferies et al. [24], Markham et al. [30], Mavrogeni et al. [33], and Cirino et al. [19] due to small sample sizes (overall and/or by examined strata). Finally, we provided an overall rating of the certainty of the evidence of each study (Table 2).

Discussion

Across the past couple of decades, the successful dissemination of a coordinated, multidisciplinary approach to the clinical management of DMD has realized remarkable improvements to prognosis. Yet, as patients walk and live longer, new challenges have emerged, especially for cardiologists. Indeed, the development of therapeutic strategies responding to the additional strain on the heart associated with prolonged ambulation, as well as increased life-expectancy, has emerged as one of the most pressing clinical issues in this heavily burdened patient population. A key component to this effort, relevant to both clinical practice and research, is an increased understanding of sources of cardiac heterogeneity. To that end, in this systematic literature review, encompassing a total of 33 studies involving 9,232 patients from 11 countries, we synthesized and graded the body of evidence of predictors of cardiac disease in DMD.

Exposure to cardiac medication, including ACE inhibitors, β-blockers, and mineralocorticoid receptor antagonists, has been shown to have a significant effect on a wide range of commonly evaluated cardiac outcomes in patients with DMD. However, in many studies, the individual contribution from these pharmacological agents remains to some degree unknown, since they are commonly prescribed in combination. For example, in the study of β-blockers by Kelley et al. [26], some patients were concurrently treated with ACE inhibitors and/or ARBs, diuretics, anti-arrhythmics, and inotropes, and many were receiving glucocorticoids, which also are associated with cardiac disease in DMD (as discussed below). We also found few estimates pertaining to specific features of pharmacological cardiac intervention, such as the comparative effect of different doses or regimens, but one study examined the impact of the timing of initiation of ACE inhibitors in patients with DMD or BMD, reporting of a significant effect only among those treated at LVEF < 55% [14]. Similar negative findings have been more recently reported from a RCT of children with DMD (mean age: 9 years) with normal ventricular function treated with ACE inhibitors and β-blockers for 36 months [47] (which is not surprising given that cardiac dysfunction is not expected at these ages in patients with DMD [8]). Considering the increased importance of cardiac management in DMD following prolonged ambulation and survival, further research is warranted to help understand optimal treatment algorithms of cardiac medication in this patient population, including benefits and harms of prophylactic intervention.

Several identified studies focused on the genotype–phenotype association with dystrophin-deficient cardiomyopathy. Mutations in exons 51 and 52, deletions treatable by exon 53 skipping, and mutations involving the Dp116 coding region, have been shown to have a comparatively protective effect against cardiomyopathy [24, 39, 45]. However, in terms of mutations associated with a higher risk and early onset of cardiac disease, we found some potential inconsistent results. While some authors observed that particularly distal or downstream mutations were associated with early left ventricular systolic dysfunction [19], other authors reported that more proximal or upstream mutations were associated with an early onset of cardiomyopathy [24]. Furthermore, Jeffries et al. [24] found that mutations in exon 52 were protective against cardiomyopathy; while Cirino et al. [19], reported an early onset of left ventricular systolic dysfunction with involvement of this same exon. In addition, other genes than DMD have also been linked to cardiac outcomes and have been mentioned as potential prognostic factors. Particularly, LTBP4 and ACTN3 polymorphisms and genotypes have been proposed to be associated with a higher risk of dilated cardiomyopathy [16, 34]. Concerning the interpretation of the synthesized evidence of DMD mutations and DMD genetic modifiers, it is important to keep in mind that the field of genetics/genomics in DMD is still advancing. As such, some publications of this topic report results from relatively small pilot studies of low certainty. However, this does not mean that the potential importance of DMD mutations and DMD genetic modifiers is low, or that further investigation of DMD mutations and DMD genetic modifiers is not warranted. Instead, our synthesis should be viewed as the current state of the evidence-base, expected to be amended by future research, through which our understanding and certainty of the evidence of specific genetic factors in DMD is expected to be greatly enhanced.

Glucocorticoids have a significant, positive effect on a wide range of cardiac outcomes in DMD. Yet, similar to cardiac medications (discussed above), little is known of the comparative impact of specific agents or regimens. In most studies, it is also difficult to elicit the effects specific to glucocorticoids, since they are commonly prescribed together with, for example, ACE inhibitors and β-blockers. Interestingly, Kim et al. [27] found that patients treated early with glucocorticoids had worse outcomes than those who remained untreated or treated late. A possible explanation for this finding includes confounding by indication, in which those treated early are clinically different from those not treated or treated late, for example, by being subject to a particularly aggressive disease trajectory (which could trigger early intervention). Nonetheless, the impact of different timings of, or criteria for, treatment initiation on cardiac disease in DMD remains largely unknown and warrants further study.

Fayssoil et al. [22] reported full-time mechanical ventilation support to be significantly associated with more favorable cardiac progression. Although not yet replicated in other samples of patients with DMD, as noted by the authors, these findings are supported by previous research showing that ventilatory support can help increase intrathoracic pressure and thus decrease left ventricular afterload. Yet, it is important to keep in mind that similar to most studies in this review, Fayssoil et al. [22] studied patients also receiving ACE inhibitors, β-blockers, and diuretics. It is therefore not possible to quantify the specific contribution of ventilatory support on cardiac disease based on the reported data.

Our findings have several implications for clinical practice and research. First, understanding predictors of cardiac disease, including phenotypic variability as part of the natural disease evolution, is important for tailoring patient-specific treatment algorithms, as well as to shape expectations of realistic treatment outcomes. Second, evidence of predictors of cardiac disease is critical also to the design RCTs of new pharmaceutical interventions in DMD to ensure adequate internal and external validity. Indeed, pooling patients with vastly different disease trajectories, in particular those exhibiting extreme phenotypes (either protective or detrimental) is likely to produce estimates of treatment effects that are challenging to interpret and difficult to generalize. Third, and last, the data synthesized as part of this review would also be expected to help inform matching algorithms and similar statistical procedures employed to indirectly compare and contextualize evidence obtained from single-arm trials to outcomes observed in natural history studies. This is likely to become increasingly important as the pipeline of new experimental treatments, including gene therapies, is reaching testing in human clinical trials in the coming decade [48].

Our study is subject to a few limitations. First, to ensure relevance to current clinical care practices, we limited the search to account for records published from the calendar year 2000. Although unlikely, we might thus have missed some data applicable to the review topic. Second, it is important to emphasize that we were unable (based on the reported evidence) to compare the impact of specific predictors of cardiac disease, for example genetic versus therapeutic effects. That being said, from our review, it is clear that such an analysis would be quite challenging to perform because of the number of potential predictors the typical patient with DMD simultaneously is subject to at a given time (e.g., genetic modifier, cardiac medication, and glucocorticoids). Large studies of predictors of cardiac disease in DMD might help delineate some of the individual effects; yet, from an epidemiological point of view, eliciting the causal effects of individual predictors is likely to remain a challenge, in particular for less common genetic expressions. Third, in concordance with the review objective, we did not account for predictors of progression of myocardial disease based on cardiovascular assessments (e.g., the relationship between myocardial fibrosis, MRI, and/or blood biomarkers, respectively, and the development of systolic dysfunction and heart failure), since this would necessitate an independent search strategy encompassing dedicated search criteria, and also considering the scale of the current review (as adding numerous additional factors, identified via a separate protocol, would greatly expand the scope and complexity of the study). Identifying this evidence is, however, an important topic for future research. Fourth, and last, we did not recognize and include certain factors known to impact cardiac health and function more generally (e.g., exercise and obesity), as our review, by design, focused on evidence derived from populations of patients with DMD.

Conclusions

Several sources of cardiac disease heterogeneity have been delineated in patients with DMD, including cardiac medication (moderate- to high-quality evidence), DMD mutations (low/very low-quality evidence), DMD genetic modifiers (low-quality evidence), glucocorticoid exposure (high-quality evidence), muscle strength (low-quality evidence), and ventilation support (low-quality evidence). Yet, little is known of the contribution of non-pharmacological interventions, as well as the impact of different criteria for initiation of specific treatments. Our findings help raise awareness of prevailing unmet needs, shape expectations of treatment outcomes, and inform the design of future research.