Sodium-glucose cotransporter-2 inhibitors in type 2 diabetes: a magic potion to reduce heart failure?

KEYWORDS:

• Heart failure
• hospitalisation
• sodium-glucose cotransporter-2 inhibitors
• treatment
• type 2 diabetes mellitus

1. Introduction

It is four decades since the Framingham study revealed the strong association between heart failure and type 2 diabetes mellitus (T2DM), with evidence supporting a bidirectional relationship [1,2]. Initially, research focused on the reduction of the risk of new-onset DM with pharmacological agents used in heart failure, notably candesartan, valsartan, ramipril and enalapril [2]. This changed in 2015 with the results of the EMPA-REG trial, showing for the first time that a glucose-lowering agent, empagliflozin, reduced the relative risk of hospitalization for heart failure (hHF) by 35% [3]. Thus, sodium glucose co-transporter 2 inhibitors (SGLT-2is), initially developed to target hyperglycemia, began to emerge as potential drugs to prevent heart failure in populations with and without DM [4].

Of note, cardiovascular outcome trials (CVOTs) were initiated in 2008 to prove non-inferiority as compared with placebo [5]. This started in response to the association of rosiglitazone with myocardial infarction [6]. Consequently, large CVOTs primarily focused on atherosclerosis, whereas heart failure was considered a secondary outcome.

2. Body

Indeed, the EMPA-REG trial included 7020 T2DM participants, virtually all (99.9%) with established cardiovascular disease (CVD) [3]. Similarly, the CANVAS Program assessed canagliflozin use among 10142 T2DM subjects, 65.6% of whom with established CVD [7]. In both these CVOTs, the primary outcome was the 3-point major adverse cardiovascular (CV) event (i.e. non-fatal myocardial infarction, non-fatal stroke, and CV mortality), whereas hHF was a secondary outcome [3,7]. More recently, the DECLARE-TIMI trial used two co-primary endpoints: (a) 3-point major adverse CV event; (b) CV death or hHF [8]. The latter trial was the largest including 17,160 participants (40.6% with CVD) who were randomized to 10 mg dapagliflozin or placebo [8]. These three CVOTs differed in terms of their primary outcome; however, results on hHF were consistent across all three studies, suggesting a significant reduction in hHF and providing evidence of a potential class effect [3,7,8].

These promising findings persisted irrespective of baseline heart failure status and were followed by a large systematic review and meta-analysis of the three aforementioned CVOTs [4,9]. Overall, SGLT-2is reduced hHF risk by 31% with a similar magnitude of the effect in patients with and without atherosclerotic CVD or a history of heart failure [9]. Conversely, the reduction in major adverse CV events was apparent only in subjects with established atherosclerotic CVD [9]. Such beneficial effects became more evident in another meta-analysis of 236 clinical trials including 176 310 participants (42.9% on SGLT-2is, 60.0% on glucagon-like peptide 1 receptor agonists [GLP-1RAs] and 53.8% on dipeptidyl peptidase 4 inhibitors [DPP-4is]) [10]. As compared to (a) placebo or no treatment, (b) dipeptidyl peptidase 4 inhibitors and (c) glucagon-like peptide 1 receptor agonists, SGLT-2is were associated with robust decrease in heart failure events (hazard ratio [HR]: 0.62, 95% confidence interval [CI]: 0.54–0.72; HR: 0.55, 95% CI: 0.46–0.67; HR: 0.67, 95% CI: 0.57–0.80, respectively), whereas the remaining antidiabetic agents failed to produce any significant effect as compared with placebo/no treatment [10].

Importantly, this benefit was replicated in two large real-world studies [4,11]. The CVD-Real study analyzed data from 154,528 participants (13% with baseline CVD) receiving SGLT-2is (53% canagliflozin, 5% empagliflozin and 42% dapagliflozin) and demonstrated a 39% reduction in hHF in comparison with other antihyperglycemic agents [4]. Similarly, the EMPRISE trial with 16,443 subjects (who were initiated on empagliflozin or sitagliptin), of whom 25% had baseline CVD, documented a 49% decrease in hHF, as compared with DPP-4is [11].

The most recent work by Singh et al. was the first to retrieve both randomized controlled trials (RCTs) (34,322) and observational studies (1,536,339) with a minimal duration of 24 weeks among T2DM subjects that explicitly reported hHF outcome [12]. Their systematic review and meta-analysis of all studies (1,570,661) documented a robust (34%) reduction in hHF risk vs. comparators (placebo and other anti-diabetic agents including sulfonylureas, DPP-4is, and GLP-1RAs) [12]. Separate analysis suggested further significant hHF reduction both in RCTs (30%) and in observational studies (36%). As previously shown, in sub-group meta-analysis of the RCTs, the benefit in hHF outcomes persisted irrespective of the baseline heart failure status [12].

The authors pointed to the need for conservative enthusiasm and suggested three important limitations in the major CVOTs [12]. First, the number of participants with concomitant heart failure and T2DM was too limited to reflect the actual clinical practice (10% in EMPA-REG, 14.4% in CANVAS Programme and 10% in DECLARE-TIMI) [12]. However, in the large real-world observational CVD-REAL and EMPRISE studies, only 3% and 5%, respectively, of participants exhibited baseline heart failure [4,11]. Secondly, there was a lack of objective heart failure verification [12]. Arguably, EMPA-REG, CANVAS Programme, and DECLARE-TIMI were not initiated as original heart failure clinical trials but were designed to demonstrate CV safety [3,7,8]. Unavoidably, heart failure was thus an exploratory outcome. By contrast, a number of studies provide evidence of the effects of SGLT-2is on ultrasound-measured parameters and biomarkers [13,14]. A systematic analysis of transthoracic echocardiography in T2DM subjects with established CVD suggested left ventricular reverse remodeling and diastolic function improvement following 3-month empagliflozin treatment (10 mg daily) [13]. In another study, 2-year canagliflozin therapy (100 mg and 300 mg daily) was associated with a significant delay in the rise of serum N-terminal pro-B-type natriuretic peptide, as compared with placebo (difference in median percent change −26.8%, p < 0.05) among 714 older (55–80 years) T2DM subjects [14].

Finally, and most importantly, the authors highlighted the fact that heart failure type was not specified in the three major CVOTs [12]. Indeed, no information was provided on ejection fraction, either preserved (≥50%) or reduced (<50%), and only a uniform reduction in the risk for hHF was shown [3,7,8]. Regarding this particular endpoint, data from studies that have compared hHF risk between diabetic subjects with preserved and those with reduced ejection fraction are inconsistent [15]. In this context and until the impact of DM on different ejection fraction groups will be clearly defined, the risk reduction in hHF with this class of anti-diabetic agents has a significance that cannot be overruled. One step forward, moving towards the treatment of heart failure, the new-designed, dedicated trials including EMPEROR-Preserved, EMPEROR-Reduced, and DAPA-HF apply standard definitions and specifically target patients with heart failure [4]. Their results are eagerly awaited to provide answers on the mechanism of action of these agents and whether their use can be expanded to non-diabetic populations.

Until definite answers are given, it appears that SGLT-2is exert their favorable heart failure outcome by pleiotropic mechanisms beyond the reduction of glucotoxicity and diuresis [16]. These include improvement of cardiac energy metabolism through ketone utilization, attenuation of oxidative stress, reduction of arterial stiffness and anti-inflammatory actions [5,17,18]. More recently, reduction in plasma volume without concomitant compensating neurohormonal activation and a mitochondrial protective effect (in vitro and in vivo) has been postulated as potential mechanism mediating the reduction of heart failure events [19,20].

In conclusion, SGLT-2is are relatively new antidiabetic agents in T2DM, and there is no doubt that hitherto available data encourage their use [5,21]. Even more importantly, in the new era after the EMPA-REG trial, non-inferiority vs. placebo is no longer enough [3]. This has led to the now prevailing preference for the use of SGLT-2is as second-line treatment after metformin monotherapy to prevent or delay the onset of heart failure [4]. In the light of this evidence, the 2018 joint consensus of the American Diabetes Association and the European Association for the Study of Diabetes has also suggested using empagliflozin or canagliflozin as second-line treatments in patients with established heart failure [22]. Although the cardiovascular benefits should always outbalance potential risks pertaining to genital candidiasis, euglycemic diabetic ketoacidosis, fractures and lower limb amputations (the latter two only recorded during canagliflozin treatment), the aforementioned hitherto available evidence from randomised trials, meta-analyses and real-world studies points to a favorable effect of the three SGLT-2is (dapagliflozin, canagliflozin and empagliflozin) on heart failure [3,4,7–12]. One step forward, results of the ongoing trials specifically addressing heart failure are eagerly awaited to elucidate the mechanisms of action of SGLT-2is and to show whether these agents can prove useful in the treatment of heart failure, irrespective of the presence or not of T2DM [4]. Overall, both existing and accumulating data increase clinicians’ optimism about the ever-improving efficacy of the treatments for T2DM and its complications [23,24].

Article highlights

• Three large cardiovascular outcome trials have shown a reduction in the hospitalization for heart failure (hHF) with the use of three sodium-glucose cotransporter-2 inhibitors (SGLT-2is): empagliflozin, canagliflozin, and dapagliflozin.

• This benefit has been demonstrated in a meta-analysis of randomized controlled trials and confirmed in real-world observational studies.

• More recently, a meta-analysis of randomized controlled trials and observational studies has arrived at the same conclusion.

• Importantly, the benefit in hHF outcomes persisted irrespective of the baseline heart failure status.

• We are now eager to acquire robust data on the mechanisms of action of SGLT-2is and to show whether these agents can prove useful in the treatment of heart failure, irrespective of the presence or not of diabetes mellitus. Ongoing trials are addressing these issues.

Declaration of interest

NP has been an advisory board member of TrigoCare International, Abbott, AstraZeneca, Boehringer Ingelheim, Elpen, MSD, Novartis, Novo Nordisk, Sanofi-Aventis, and Takeda; has participated in sponsored studies by Eli Lilly, MSD, Novo Nordisk, Novartis and Sanofi-Aventis; received honoraria as a speaker for AstraZeneca, Boehringer Ingelheim, Eli Lilly, Elpen, Galenica, MSD, Mylan, Novartis, Novo Nordisk, Pfizer, Sanofi-Aventis, Takeda and Vianex; and attended conferences sponsored by TrigoCare International, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Novartis, Novo Nordisk, Pfizer and Sanofi-Aventis.

Reviewer disclosures

A reviewer on this manuscript has disclosed

● Advisory boards: AstraZeneca; Boehringer Ingelheim; Bristol-Myers Squibb;

Eli Lilly; Intarcia Therapeutics; Johnson & Johnson; Merck Sharp & Dohme; Novartis; Novo Nordisk; Sanofi Aventis

● Lecture fees: AstraZeneca; Boehringer Ingelheim; Bristol-Myers Squibb;

Eli Lilly; Merck Sharp & Dohme; Novartis; Novo Nordisk; Sanofi Aventis

● Research Grant Recipient: Novo Nordisk, Boehringer Ingelheim

Peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose.

Additional information

Funding

This paper was not funded.