FormalPara Key Points

This review and network meta-analysis found a combination of different drugs and conventional treatment could improve the efficacy of DCM treatment in clinical practice.

Carvedilol and ivabradine were found to be relatively high in several efficacy indicators and had strong therapeutic application value.

Beta-blockers play an important role in the treatment of dilated cardiomyopathy.

1 Introduction

Dilated cardiomyopathy (DCM), characterized as the dysfunction of dilatation and movement in the left or both ventricles [1], is one of the most prevalent myocardial diseases in clinical practice. This disease is considered one of the most common causes of heart failure (HF) [2], accounting for 1:250–400 among patients with HF and 1:2500 among the general population [3]. According to another study [4], there are 5–7 cases of DCM per 100,000 people annually. The disease develops slowly, the mechanism is complex, and many consequences might emerge in the advanced stage [5]. The prognosis of DCM is poor, with a 5-year survival rate of about 50% and a 10-year survival rate of 25% [6, 7].

The treatment of DCM centers on improving cardiac function and myocardial metabolism, staving off infection, and preventing embolism and other complications, and there is a lack of specific treatment. The endpoint of treatment is to avoid HF and arrhythmias [3, 8]. The majority of patients who receive a typical HF regimen report better DCM symptoms and a longer survival time [9]. However, after 2 years of treatment, conventional anti-HF therapy has a 40% failure rate, which may result in refractory HF requiring heart transplantation [10]. Hence, more novel medications are required to extend the survival of DCM patients. In recent years, numerous clinical trials have tried various unconventional and adjuvant drugs to treat DCM, including agents for improving myocardial metabolism, non-dihydropyridine calcium channel blockers, and statins. However, there is still a lack of comparison between the efficacy of different drugs in treating DCM.

Conventional meta-analyses cannot compare various kinds of interventions and select the most effective intervention. Network meta-analyses may combine evidence from direct and indirect comparisons to determine the optimal treatment option, resulting in evidence-based medical evidence for drug selection in clinical practice [11, 12]. This network meta-analysis compared the efficacy of conventional and adjuvant drugs in the treatment of DCM to provide a reference for clinical application.

2 Materials and Methods

Our investigation followed the PRISMA 2020 statement [13, 14], and the checklist is included in Table S2.The study was pre-registered on PROSPERO (# CRD42022339080 [15]).

2.1 Literature Search and Screening

The studies were obtained from PubMed, Embase, the Cochrane Library, and Web of Science between the databases’ inception and 27 June 2022. The language of the studies was limited to English. The main search terms used were “cardiomyopathy, dilated;” “cardiomyopathies, dilated;” “dilated cardiomyopathies;” and “dilated cardiomyopathy.” Detailed information is presented in Table S1. Two researchers independently performed literature screening. If there was any dissent, a third reviewer was consulted and resolved the discrepancy by discussion. According to the type of studies, clinical characteristics of the research subjects, and interventions, the titles and abstracts were reviewed for preliminary screening. Irrelevant literature was excluded, and all possibly and definitely relevant studies were collected. When the researchers reviewed the literature using the full text, the author was contacted if the information was missing.

2.2 Inclusion and Exclusion Criteria

Inclusion criteria: (1) Research subjects: adults diagnosed with DCM without limitation on sex, age, disease course, etc. (2) The study design was a randomized controlled trial (RCT). (3) Intervention: for the treatment group, study drugs combined with conventional treatment (including diuretics, digoxin, angiotensin-converting enzyme inhibitors, or angiotensin II receptor antagonists [8]); for the control group, conventional treatment or conventional treatment combined with study drugs. (4) Outcome measures: improvement of left ventricular ejection fraction (LVEF), left ventricular end-diastolic dimension (LVEDD), left ventricular end-systolic dimension (LVESD), New York Heart Association classification (NYHA), and heart rate (HR).

Exclusion criteria: (1) No comparability between groups; (2) for duplicated publications or similar literature, inclusion of only the one with the largest sample size and the most complete information; and (3) obvious errors in data that could not be corrected.

2.3 Data Extraction and Quality Evaluation

Two researchers extracted the data and assessed the quality. The extracted data included researchers, study type, patient characteristics, treatment methods, and measurement indicators. Quality assessment was conducted according to the Cochrane tool for assessing the risk of bias of RCTs, including the method of randomized allocation, concealment of grou** scheme, blinding, completeness of outcome data, selective reporting of study results, and other sources of bias. If there was any dissent, a third author was consulted for evaluation.

2.4 Statistical Analysis

R4.1.3 software was used to perform network meta-analysis and generate the forest plot, SCURA plot, and league table. The Revman 5.4 software was employed to assess the bias. I2 was used to calculate heterogeneity quantitatively. If I2 < 50%, there was no statistical heterogeneity among the studies, and a fixed-effects model was used for meta-analysis. The source of heterogeneity should be investigated if I2 was 50% or above, indicating statistically significant heterogeneity among trials, and a random-effects model was used for meta-analysis. The gemtc package (version 1.0-1) in R4.1.3 software was used to conduct a network meta-analysis and create network relationship diagram. For Bayesian inference, Markov Chain Monte Carlo (MCMC) was utilized, and the random-effects model was used for analysis. Initially, four Markov chains were set for simulation, and the number of iterations was set as 25,000 times. The first 5000 times were for annealing to eliminate the influence of the initial values, and the iteration history diagram was plotted to evaluate the convergence of the model. The convergence was represented by the potential scale reduction factor (PSRF). If the PSRF value converged to 1, it indicated that the model had adequate convergence and that the related findings in this model could be used. surface under the cumulative ranking curve (SUCRA) value was calculated to rank the efficacy of the intervention. A higher SUCRA value suggested a better effect of the intervention.

3 Results

3.1 Results of Literature Retrieval

A total of 5640 papers [PubMed (n = 827), Embase (n = 1698), Cochrane (n = 1089), Web of Science (n = 2026)] were initially retrieved, and 2121 duplicated publications were excluded. According to the inclusion and exclusion criteria, 3243 articles were excluded by reading the titles and abstracts, and 224 were excluded after reading the full text. Finally, 52 RCTs [16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67] were included. The detailed literature screening process and results are shown in Fig. 1.

Fig. 1
figure 1

Flow chart of the study selection

Full size image

3.2 Basic Information of the Included Studies

A total of 52 RCTs [16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67] involving 3048 patients with DCM were included. There were 25 drugs included in this study (l-thyroxine, ivabradine, dipyridamole (DIP), trimetazidine, perhexiline, carvedilol, metoprolol, rosuvastatin, n-3 polyunsaturated fatty acids (n-3 PUFAs), verapamil, atorvastatin, low-molecular-weight heparin (LMWH), nebivolol, amiodarone, cerivastatin, simvastatin, spironolactone, pentoxifylline, recombinant human growth hormone (rhGH), diltiazem, enalapril, bucindolol, bevantolol, captopril, coenzyme Q10). The network meta-analysis was performed on LVEF, LVEDD, LVESD, HR, and NYHA (Fig. 2). Detailed data are in Fig. S1. All the included studies reported baseline information (including gender, age, NYHA cardiac function class, and intervention time), which was comparability. Table 1 depicts the basic features of the included studies.

Fig. 2
figure 2

Network relationship diagram of different outcomes of treatment. It combines the relationship between the reticulation and some of the comparative efficacy in the league table, with arrows pointing in the direction of worse efficacy and statistically different, and dots representing no statistical difference

Full size image
Table 1 Baseline characteristics of studies included in the network meta-analysis
Full size table

3.3 Quality Evaluation of Included Studies

The studies were evaluated by using the Cochrane tool for assessing the risk of bias of RCTs. Nine studies providing a detailed description of random sequence generation and allocation concealment were rated as low risk [17, 23, 33, 34, 43, 49, 51, 61, 62]. A total of 7 studies [16, 25, 46, 52, 63, 65, 66] described detailed single-blinding method, and 11 [20, 21, 23, 26, 39, 47, 50, 51, 55, 56, 61] studies described detailed double-blinding method. Thus, the 18 studies were considered low risk. Since none of the included RCTs had missing data, the risk of bias in “incomplete outcome data” was low. Five studies did not report all pre-specified primary outcome indicators, so their risk of bias in “selective reporting” was unclear. Other risks of bias were rated as unclear risks (Fig. 3).

Fig. 3
figure 3

Risk of bias. A Risk of bias summary; B risk of bias graph

Full size image

3.4 LVEF Improvement

LVEF improvement was the difference of LVEF before and after treatment. A total of 43 articles [16,17,18,19,20,21,22,23, 26,27,28,29, 31,32,33,34,35,36,37,38,39,40,41, 43, 45,46,47,48, 50,51,52,53,54,55, 57,58,59,60,61,62, 64,65,66] involving 22 drugs reported LVEF. The I2 of the meta-analysis was 38%, and the fixed-effects model was used. Because there was little heterogeneity across these studies, subgroup and sensitivity analyses were not undertaken. The PSRF value was 1, indicating satisfactory convergence. Additionally, no publication bias test was performed because the number of comparisons between all these studies was lower than 10. The league table demonstrated that, in 43 groups of drugs among all the interventions, the effects of drugs in the line on improving LVEF were significantly better than the drugs in the row. The effects of drugs in the line on improving LVEF were significantly worse than those in the row in 36 groups of drugs (Data Sheet 1). However, considering the SUCRA value and forest plot, carvedilol [versus control (mean difference (MD)): −10.41, credible intervals {CrI}: −11.72, −9.10)] showed the best effects among these interventions. Verapamil [versus control (MD: −10.64, CrI: −13.90, −7.39)] and trimetazidine [versus control (MD: −8.98, CrI: −11.03, −6.93)] also showed good effects, while rhGH [versus control (MD: 0.90, CrI: −2.67, 4.52)] showed the worst effects on improving LVEF (Table 2; Fig. 4A).

Table 2 SUCRA of LVEF, LVEDD, LVESD, NYHA, and HR rate
Full size table
Fig. 4
figure 4

Forest plot of different outcomes of treatments. A LVEF; B LVEDD; C LVESD; D NYHA; E HR

Full size image

3.5 LVEDD Improvement

LVEDD improvement was the difference of LVEDD before and after treatment. A total of 27 studies [16, 17, 19, 21, 26, 27, 29, 31,32,33, 37, 41, 42, 44,45,46, 51,52,53,54,55, 57, 58, 63,64,65,66] involving 18 drugs reported LVEDD. The I2 of the meta-analysis was 11%, and the fixed-effects model was used. Due to the low heterogeneity among these studies, subgroup and sensitivity analyses were not undertaken. The PSRF value was 1, indicating satisfactory convergence. Moreover, no publication bias test was performed, because the number of comparisons between all these studies was lower than 10. The league table showed that the effects of drugs in the line on improving LVEDD were substantially better than the effects of drugs in the row in nine drug groups across all interventions. The effects of drugs in the line on improving LVEDD were significantly worse than the drugs in the row in 1 group of drugs (Data Sheet 2). However, considering the SUCRA value and forest plot, ivabradine [versus control (MD: 5.00, CrI: 3.53, 6.49)] showed the best effects among these interventions. Bucindolol [versus control (MD: 4.93, CrI: 3.72, 6.13)] and verapamil [versus control (MD: 5.13, CrI: 0.78, 9.46)] similarly had good effects on improving LVEDD, while enalapril [versus control (MD: 0.54, CrI: −6.77, 7.89)] had the worst effect (Table 2; Fig. 4B).

3.6 LVESD Improvement

LVESD improvement was the difference of LVESD before and after treatment. LVESD was reported in 16 papers [16, 17, 19, 21, 26, 29, 31,32,33, 41, 44, 46, 51, 55, 57, 63], including 10 drugs. The I2 of the meta-analysis was 5%, and the fixed-effects model was used. Due to the low heterogeneity among these studies, subgroup and sensitivity analyses were not undertaken. The PSRF value was 1, indicating satisfactory convergence. Furthermore, no publication bias test was performed because the number of comparisons between all the above studies was lower than 10. The league table indicated that the effects of medications in the line on improving LVESD were considerably better than drugs in the row in six drug groups among all interventions. The effects of drugs in the line on improving LVESD were significantly worse than the drugs in the row in 11 groups of drugs (Data Sheet 3). However, considering the SUCRA value and forest plot, ivabradine [versus control (MD: 9.31, CrI: 7.68, 10.96)] showed the best effects among these interventions. l-thyroxine [versus control (MD: 9.23, CrI: 3.55, 14.9)] and atorvastatin [versus control (MD: 5.95, CrI: 0.84, 11.09)] also showed good effects, while diltiazem [versus control (MD: 0.10, CrI: −3.02, 3.21)] showed the worst effects on improving LVESD (Table 2; Fig. 4C).

3.7 NYHA Improvement

NYHA improvement was the difference of NYHA before and after treatment. A total of 30 articles [16,17,18,19,20,21,22,23,24,25, 31,32,33,34,35,36,37, 39,40,41, 46, 50, 51, 53, 55, 63,64,65,66,67] involving 18 drugs reported NYHA scores. The I2 of the meta-analysis was 44%, and the fixed-effects model was used. Due to the low heterogeneity among these studies, subgroup and sensitivity analyses were not undertaken. The PSRF value was 1, indicating satisfactory convergence. In addition, no publication bias test was performed, because the number of comparisons between all the above studies was lower than 10. The league table indicated that the effects of medications in the line on improving NYHA score were considerably better than drugs in the row in 34 drug groups across all interventions. The effects of drugs in the line on improving NYHA score were significantly worse than the drugs in the row in 26 groups of drugs (Data Sheet 4). However, considering the SUCRA value and forest plot, trimetazidine [versus control (MD: 0.86, CrI: 0.70, 1.01)] showed the best effects among these interventions. Pentoxifylline [versus control (MD: 0.80, CrI: 0.51, 1.09)] and bucindolol [versus control (MD: 0.70, CrI: 0.59, 0.81)] similarly had good effects on improving NYHA score, but DIP [versus control (MD: −0.13, CrI: −0.70, 0.44)] had the worst effects (Table 2; Fig. 4D).

3.8 HR Improvement

HR improvement was the difference of HR before and after treatment. HR was reported in 31 papers [16, 17, 21, 22, 26, 29, 30, 33, 34, 36, 39, 41, 42, 44,45,46,47, 49, 51,52,53, 55,56,57, 59,60,61,62, 64, 65, 67], including 15 drugs. The I2 of the meta-analysis was 11%, and the fixed-effects model was used. Due to the low heterogeneity among these studies, subgroup and sensitivity analyses were not conducted. The PSRF value was 1, indicating satisfactory convergence. No publication bias test was performed because the number of comparisons between all the above studies was lower than 10. The league table showed that the effects of medications in the line on improving HR were considerably better than the effects of drugs in the row in 15 drug groups across all interventions. The effects of drugs in the line on improving HR were significantly worse than the drugs in the row in 31 groups of drugs (Data Sheet 5). However, considering the SUCRA value and forest plot, ivabradine [versus control (MD: 13.00, CrI: 10.06, 15.95)] showed the best effects among these interventions. Carvedilol [versus control (MD: 9.07, CrI: 6.18, 11.96)] and bucindolol [versus control (MD: 9.05, CrI: 6.44, 11.73)] similarly had good effects on improving HR, but captopril [versus control (MD: −11.11, CrI: −17.63, −4.62)] had the worst effects (Table 2; Fig. 4E).

3.9 Sensitivity Analysis

After reviewing the 52 studies included in our research, we discovered that 2 of them [35, 36] were ischemic DCM, while the rest were non-ischemic DCM. We then conducted a sensitivity analysis by removing these two articles and re-running the network meta-analysis. The results showed no obvious difference compared with before the deletion. We have included these results in our supplementary materials. The findings of this study are best explained by patients with non-ischemic DCM. Further research is required to analyze patients with ischemic DCM.

4 Discussion

In this systematic review and network meta-analysis, we comprehensively summarized the efficacy of various drugs in patients with DCM. Echocardiography is the first-line imaging method for assessing DCM, and it can provide critical information for DCM diagnosis, risk stratification, and treatment [68]. About 40% of DCM patients have left ventricular reverse remodeling (LVRR). LVRR is one of the main determinants for the prognosis of DCM and should be regarded as the main target in the treatment [69]. Typical changes in LVRR can be demonstrated by echocardiography [3]. Decreasing HR can improve left ventricular filling and maintain the balance of myocardial oxygen supply and demand in patients with HF, lowering mortality and cardiovascular events [70,71,72]. Therefore, we comprehensively summarized the efficacy of various drugs used in DCM patients on LVEF, LVEDD, LVESD, and HR in this systematic review and network meta-analysis.

The results showed that the combination of adjuvant therapy and conventional anti-HF therapy was more effective than conventional therapy alone in improving LVEF, LVEDD, LVESD, NYHA, and HR. In improving LVEF, carvedilol, verapamil, and trimetazidine showed the best efficacy, while ivabradine showed significant effects in improving LVEDD, LVESD, NYHA, and HR.

In individuals with DCM, the circulating catecholamine level is usually elevated, indicating an overactive sympathetic nervous system [73]. This can aggravate left ventricular dysfunction and LVRR. Carvedilol, with unique vasodilating and antioxidant properties, is the most effective one among beta-blockers [74]. Another study [75] found that changes in Ca2+ homeostasis are a common and important potential mechanism resulting in arrhythmia, which can lead to sudden cardiac death or congestive heart failure. Carvedilol and verapamil can pharmacologically modulate abnormal Ca2+ to improve left ventricular function. Ivabradine is a specific inhibitor of funny current in the sinus node. It was mainly used for treating patients with LVEF ≤ 35% and resting HR ≥ 70 beats/min after being treated with beta-blockers (maximum dose or maximum tolerated dose) [76]. Ivabradine, unlike beta-blockers, has no effect on myocardial contractility or intracardiac conduction, even in individuals with impaired systolic function [77]. A study [78] showed that the risk of HF could be significantly reduced by using Ivabradine in combination with guideline- and evidence-based treatment. Ivabradine, when taken properly, can not only alleviate tachycardia caused by sympathetic nerve hyperactivity but also increase parasympathetic nerve activity, improving cardiovascular autonomic regulation and reversing ventricular remodeling [79]. The results also indicated that HR control is of great significance in the prognosis of DCM. Insufficient energy of myocardial cells is the main cause of HF. Trimetazidine, a kind of drug that improves myocardial metabolism, can reduce fatty acid α-oxidation and increase glucose oxidation, which results in higher adenosine triphosphate (ATP) production, thus alleviating the metabolic disorders in myocardial cells and subsequently improving cardiac function [80, 81].

Anti-HF therapy is currently the primary treatment for DCM. Conventional anti-HF drugs, such as β-blockers, angiotensin-converting enzyme inhibitors (ACEI), angiotensin II receptor antagonists (ARB), aldosterone receptor antagonists (MRA), and angiotensin II receptor enkephalinase inhibitors (ARNI), target the pathophysiological mechanism of HF, which involves three major systems in the human body (sympathetic nervous system, renin-angiotensin-aldosterone system, and natriuretic peptide system) [8, 82]. This systematic review and network meta-analysis included a variety of adjuvant medicines used for DCM therapy, such as statins, non-dihydropyridine calcium channel blockers, agents to improve myocardial metabolism, and so on. One study [83] has demonstrated that statins could change the pathophysiological process of lipid oxidation, inflammation, immunomodulation, and endothelial activity, which had potential therapeutic effects on HF induced by DCM. Due to the inhibition effects of calcium channel blockers on myocardial contractility, such drugs are generally avoided by DCM patients with relatively severe cardiac insufficiency. However, for DCM patients at an early stage, non-dihydropyridine calcium channel blockers, such as verapamil and diltiazem, can effectively improve the short-term exercise performance and diastolic filling as well as preserve the systolic function in patients with HF related with abnormal diastolic function [27]. Beta-blockers, which are commonly used in DCM therapy, appear to be more effective than other types of medications in reversing left ventricular dilation, with a closer relationship between the dose and effect [84, 85]. This is beneficial in protecting cardiac function and maintaining sufficient blood supply to organs, thus relieving clinical symptoms and improving quality of life. Nevertheless, it should be used with caution when cardiac output is low and severely restrictive disorder exists [2]. Coenzyme Q10 also improves the metabolism of myocardial cells via participating in oxidative phosphorylation and energy production. Coenzyme Q10 may enhance cardiac energy production by altering mitochondrial redox signaling, thereby preventing the vicious metabolic cycle in HF [86].

This study also included other adjuvant drugs for DCM treatment, among which l-thyroxine, LMWH, and pentoxifylline showed remarkable efficacy. l-Thyroxine exhibited significant and potentially positive effects in enhancing the efficiency of myocardial oxygen consumption while concurrently decreasing systemic vascular resistance. The metabolism of thyroid hormones is altered in numerous patients with acute and chronic heart disease (including HF). When thyroid stimulating hormone (TSH) and free T4 (FT4) levels remain normal, free T3 (FT3) levels in HF patients may be lower [2]. A further study established a link between FT3 levels and myocardial injury. Patients with low FT3 levels tend to have a higher risk of abnormal myocardial metabolism, where oxygen consumption in the heart and myocardial contractility decrease [87]. LMWH may provide additional clinical benefits for patients with chronic HF, which is possibly associated with the synergistic effects of its anticoagulant activity [31]. Pentoxifylline is a xanthine derivative that has the ability to inhibit the generation of tumor necrosis factor-alpha (TNF-α) [88]. Pentoxifylline can hinder the synthesis of TNF-α by inhibiting gene transcription [89]. Another study later presented that pentoxifylline could significantly improve cardiac function and left ventricular function, and it is related to decreased plasma concentration of TNF-α [41]. A previous meta-analysis [90] demonstrated that pentoxifylline had a beneficial effect on NYHA cardiac function classification, ejection fraction, and the mortality of HF.

According to this study, conventional therapy ranked lower in the multi-group comparison of multiple outcome measures mentioned above. This finding demonstrated that combining adjuvant medicines with conventional therapy might be more effective than conventional therapy alone in improving DCM. Nevertheless, considering the methodological shortcomings of previous research and the impact of various biases, the credibility of this conclusion is substantially diminished. Therefore, the generation of a random allocation sequence and the concealment of the random schemes should be reported in detail in future research. Blinding should be used to the extent possible, and the estimates of sample size should be reported, including detailed reporting of dropouts from the trial, thus making an intention-to-treat analyses.

4.1 Advantages and Limitations

This network meta-analysis combined information from several RCTs to highlight present concerns and aspects that need to be improved. The findings of this study provided the most thorough data currently available, and they might be used to guide early decisions about DCM treatment for adult patients. There were also some limitations in this study. First, our study may be intractable since most therapies were compared indirectly, resulting in a variety of confounding factors that we could not control. Second, despite our best efforts, the quality of the included RCTs was relatively poor. Although all the trials reported that the patients were randomly assigned into different groups, only some studies among the 52 RCTs described the specific methods of generating random sequences, such as random number tables or parallel random grou**. Third, some of the included studies were not pre-registered. Fourth, 2 of the 52 studies we included were on ischemic DCM. The results of this study can best be explained from patients with non-ischemic DCM. More studies on patients with ischemic DCM are needed for further analysis.

5 Conclusions

A combination of various drugs and conventional treatment could improve the efficacy of DCM treatment in clinical practice. Carvedilol and ivabradine, in particular, were relatively high in several efficacy indicators and had strong therapeutic application value. We found that beta-blockers have beneficial effects on ventricular remodeling, cardiac function, and clinical efficacy in DCM patients. Therefore, we believe that beta-blockers, especially carvedilol, should be used as much as possible. If LVEF and HR do not meet the standard, ivabradine can also be given in conjunction with other therapies. However, due to the limitations of the studies’ quality and quantity, large sample size, and multi-center status, high-quality randomized controlled trials are still required to corroborate our findings.