Background
Diabetes mellitus (DM) and heart failure (HF) commonly coexist. About 40 % of hospitalized HF patients have DM, and these figures are expected to grow with the general aging of the population [
1]. Results from more than 100,000 patients in the Acute Decompensated Heart Failure National Registry suggested that 44 % of HF patients had DM [
2]. Results from a health maintenance organization show that about 12 of 10,000 patients with DM had HF at baseline, and 3.3 % of the rest developed HF during each year of follow-up [
3]. There is now a large number of epidemiological and clinical data supporting the strong association between HF and DM [
4]. Patients with HF can have insulin resistance, which increases their risk of developing type 2 DM [
5]. It was shown in an Italian observational study that 28 % of elderly patients with HF developed new-onset type 2 DM in 3 years and HF is an independent risk factor for type 2 DM (OR 3.3; 95 % CI 2.6–4.0) [
6]. Patients with HF are not only at increased risk of developing DM but patients with DM also have a greater probability of developing HF [
7]. In patients with DM, every unit increase in glycosylated hemoglobin (HbA1c) is associated with a 10 % to 15 % increased risk of developing HF [
8]. Recent literature suggested that co-existence of DM and HF can lead to increased morbidity and mortality [
9]. Hospitalized HF patients with DM have an even worse prognosis with increased rates of cardiovascular (CV) mortality and post-discharge HF hospitalization [
10]. Recently, a subgroup analysis of the results indicated that during standard treatment, side effects were most likely to appear in hospitalized HF patients with DM compared to those without DM [
11]. Thus, treating coincident HF and DM is still a challenge.
Mineralocorticoid receptor antagonists (MRAs) are powerful treatment agents for patients with cardiovascular disease [
12]. Morbidity and mortality benefits from treatment with MRAs have been demonstrated in HF patient and MRAs have become part of standard medical therapy for HF [
13,
14]. Similar to HF patients without DM, treatment with MRAs is associated with improved outcomes in patients with DM [
15]. However, associated adverse events including hyperkalaemia, gynecomastia, menstrual irregularities, and acute kidney injury can not be ignored [
16]. The effects of MRAs on glycaemic control are still uncertain. The results of some studies have demonstrated that spironolactone significantly elevated HbA1c levels or worsened glycaemic control [
17,
18], while one study has shown that spironolactone may have a beneficial effect on serum insulin and HOMA-IR in patients with non-alcoholic fatty liver disease [
19]. A few studies support the view that MRAs, whether spironolactone or eplerenone, did not have a significant effect on glucose levels [
20‐
22]. Furthermore, results of a small direct comparative trial have shown that spironolactone increased HbA1c in patients with DM and HF, but eplerenone did not [
23].
Aldosterone is a mineralocorticoid hormone that activates the apical epithelial sodium channel and the basolateral Na
+ /K
+ ATPase pump, and controls sodium excretion at the level of the distal tubules to exert an action on sodium homeostasis [
24]. However aldosterone can have harmful effects on the cardiovascular system [
25]. By blocking the mineralocorticoid receptor in the distal tubule of the kidney, MRAs prevent the activation of sodium channels and lead to diuresis with reduced excretion of potassium [
26]. MRAs can prevent vascular inflammation, myocardial fibrosis and ventricular remodelling, and improve endothelial function [
14]. HF and DM are characterized by a high level of oxidative stress and it has been reported that MRAs can reduce oxidative stress [
18].
A better understanding of the efficacy and safety of MRAs in patients with HF and DM is needed. To date, no meta-analysis has been conducted concerning the relationship between treatment with MRAs and outcomes in those patients. Therefore, it is worth undertaking a systematic review to assess the efficacy and safety of treatment with MRAs in patients with concomitant HF and DM.
Methods
This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (Additional file
1) guidelines [
27].
Eligibility criteria
Types of studies
A study was included if it was either a randomised or non-randomised controlled trial, a prospective or retrospective cohort study, or a case–control study that evaluated the efficacy and safety of MRAs in patients with HF and DM.
Types of participants
Patients of any age, gender or race with HF and DM were included.
Types of interventions
Patients in the treatment groups were given MRAs while the control groups were given placebo or had no intervention. The studies were included regardless of the follow-up duration and dosage of treatment.
Types of outcome measures
The primary outcome measures were all-cause mortality and hyperkalaemia. The secondary outcome measures were CV mortality or HF hospitalization, death from CV causes and change in estimated glomerular filtration rate (eGFR).
Literature search
A computerized literature search was conducted using Medline (1966–2015), Embase (1980–2015) and the Cochrane Central Register of Controlled Trials (1991–2015) from their inception to July 2015. The search strategy in Medline is shown in Additional file
2. In addition, we manually searched reference lists of all included studies and relevant review articles. Furthermore, we contacted experts and authors of included studies to retrieve potentially relevant published or unreported material. The retrieval was not restricted by language or quality of study.
Study selection and data collection
Two reviewers (G.S.P. and Y.H.M.) independently screened the titles and abstracts to select potential references according to a data extraction form established above, including patients, methods, interventions and outcomes. For eligible studies, two reviewers (G.S.P. and Y.H.M.) extracted the data independently. Disagreements were settled through discussion or consultation with a third author (F.M.D.).
Quality assessment
The quality of the included studies was evaluated using the nine-star Newcastle Ottawa scale (NOS) [
28] by two investigators. Each study was evaluated based on eight items, and the scale ranged from zero to nine stars, the more stars the higher the methodological quality.
Data synthesis and analysis
Data were analyzed using Review Manager (version 5.2). A fixed-effects model or random-effects model was used across the studies; relative risk (RR) and 95 % confidence intervals were calculated for dichotomous outcomes. A standard chi-square test and I2 statistic were calculated to evaluate the heterogeneity between trial results. All P values were two sided and P values of less than 0.05 for any test were considered to be statistically significant.
Discussion
Four studies with 4742 individuals were identified. The main findings of the present study were that MRA treatment was associated with improved clinical outcomes compared with those regimens without MRAs in patients with DM and HF. From a safety perspective, the most serious adverse effect of MRAs (spironolactone and eplerenone) is the development of hyperkalaemia. However, as adverse effects were mentioned in only 50 % of the studies, there was not sufficient evidence to draw conclusions on the issue of safety.
There are a number of limitations in this study. First, only four studies were available; the included studies were all observational and none of them were randomized controlled trials. There were three post-hoc sub-analyses of randomized controlled trials concerning MRAs in heart failure and just one prospective cohort study. Two students scored the quality of each of the articles, and quality scores for each study are shown in Table
2. However, the quality of individual research included in our analysis was not necessarily high. Eschalier et al. [
30] did not provide the detailed information about characteristics of the subject investigated. Khosraviani et al. [
32] failed to offer an elaborated description of the study design. Thus, the level of evidence for this meta-analysis does not seem to be high.
Secondly, any systematic review may suffer from publication or selection bias. Publication bias can be assessed graphically with a funnel plot, which could not be assessed owing to having just four available studies. Thirdly, methodology defects have been found in some of these studies, including failure to collect data prospectively and inadequate baseline comparisons. Some baseline characteristics were different among the studies. For example, studies used different MRAs (spironolactone, eplerenone), different controls (placebo control or blank control), different dosages of the active and control medicines, and different follow-up times and background therapies. Lastly, significant heterogeneities were detected when the effects of MRAs on CV mortality or HF hospitalization were evaluated. The heterogeneity among the studies might be affected by various factors, such as study designs, study quality and patient characteristics. Therefore, a random-effect model was used to synthesize the data. But because only two studies reported these events, we could not perform meta-regression meta-analysis to examine the source of the heterogeneity.
Due to the limited quantity and quality of the included studies, it is premature to draw conclusion about the efficacy of MRAs in patients with HF and DM. However, it should be noted that the absence of sufficient scientific evidence does not mean that the treatment is ineffective. The safety of MRA therapy for those patients remains to be further determined. The possible electrolyte trouble could have a multifactorial origin. We observed that when using MRAs, many patients also used others drugs, including ACE inhibitors, angiotensin receptor blockers (ARBs), beta-blockers, and aspirin, but medical therapy was not different at baseline between the two groups. When the treatment with MRAs was undertaken, ACE inhibitors or ARBs were generally not discontinued, but patients underwent a modulation of their dosages or their dose was left unchanged. This suggests that ACE inhibitors or ARBs could be associated with the increased risk of hyperkalaemia; MRAs alone might not have caused this adverse effect. At the stage of the addition of MRA to therapy, a gradual up-titration of the MRA dose or a concomitant thoughtful reduction of the dose of ACE inhibitor or ARB could prevent the occurrence of hyperkalaemia, except in cases of severe impairment of kidney function. In addition, DM is an independent risk factor for the development of hyperkalaemia [
33].
Several implications for research arise from this review. First, more randomized controlled trials of higher quality, with larger sample size and adequate follow up are needed to further evaluate the effect of MRAs. Second, the only extensively researched MRAs to date are spironolactone and eplerenone, further research on other MRAs such as canrenone is encouraged in these disease areas. Third, we should closely observe the incidence of adverse events, such as electrolyte abnormalities, gynecomastia, menstrual irregularities, acute kidney injury and the effects on serum glucose levels. Fourth, further study can focus on developing novel MRAs that have similar outcomes as spironolactone but lower rates of hyperkalaemia such as finerenone [
34,
35]. Fifth, further study is needed with a focus on evaluating the outcomes of efficacy and safety in patients with heart failure associated with chronic renal failure, especially in patients with diabetic nephropathy, which are better represented in the literature regarding MRA efficacy and safety.
Acknowledgments
This work is supported by the National Natural Science Foundation of China (NSFC)(No. 81200630), Natural Science Foundation of Zhejiang Province, China (No. LQ12H07001), Wenzhou City Science & Technology Bureau, China (No.Y20140033), Medical Technology Foundation of Zhejiang Province, China (No.2015KYA159), Opening Project of Zhejiang Provincial Top Key Discipline of Clinical Medicine.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
The following authors participated in various ways: in developing the idea into a concept (CMD, ZC), writing the initial protocol (DSS, CNY, ZJJ), developing the search strategy (GSP, FMD, YHM), identifying, selecting and appraising the included studies (ZXH, WLX, LCY), drafting the manuscript (CMD), reviewing and approving the final version of the manuscript (all authors).