Impact of sodium-glucose co-transporter inhibitors on cardiac autonomic function and mortality: no time to die

Abstract

Sodium-glucose co-transporter 2 (SGLT2) inhibitors have been shown to improve cardiovascular outcomes not only in patients with diabetes but also in those with heart failure, irrespective of diabetic status. However, the mechanisms underlying the cardioprotective effects of these newer anti-diabetic drugs remain to be fully elucidated. One exciting avenue that has been recently explored in both preclinical and clinical studies is the modulation of the cardiovascular autonomic nervous system. A reduction in sympathetic nervous system activity by SGLT2 inhibitors may potentially translate into a reduction in arrhythmic risk and sudden arrhythmic death, which may explain, at least partly, the cardioprotection shown in the cardiovascular outcome trials with different SGLT2 inhibitors. Although some of the data from the preclinical and clinical studies are promising, overall the findings can be contradictory. This highlights the need for more studies to address gaps in our knowledge of these novel drugs. The present review offers an in depth overview of the existing literature regarding the role of SGLT2 inhibitors in modulating cardiovascular autonomic function as one of the possible pathways of their cardioprotective effects.

Introduction

The development of sodium-glucose co-transporter 2 (SGLT2) inhibitors (SGLT2i; gliflozins) has been a fascinating story in recent therapeutics, which garnered worldwide attention due to the clinically impactful results that have been recognized since the publication of the first SGLT2 inhibitor cardiovascular outcome trial (CVOT) in 2015 (EMPA-REG OUTCOME).¹ The EMPA-REG OUTCOME study was a multicentre randomized controlled trial which showed that patients with Type 2 diabetes mellitus (T2DM) at high risk for cardiovascular events who received empagliflozin had a lower rate of cardiovascular mortality, hospitalization for heart failure and all-cause mortality. Notably, the CVOTs in patients with T2DM were introduced and are designed with the specific objective to demonstrate the cardiovascular safety of any new anti-diabetic drug as mandated by both the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA).^2,3 Remarkably, the unexpected positive findings from EMPA-REG OUTCOME were similarly shown in subsequent CVOTs with canagliflozin and dapagliflozin (CANVAS in 2017⁴ and DECLARE-TIMI58 in 2018,⁵ respectively) in patients with T2DM, suggesting a class effect which extends beyond blood glucose control. Moreover, two further SGLT2 inhibitor trials in 2019 demonstrated the benefit of these drugs in patients with chronic kidney disease and T2DM (CREDENCE⁶), and heart failure with reduced ejection fraction patients regardless of diabetes status (DAPA-HF⁷).

As SGLT2 receptors are predominantly expressed in the kidney and not the heart,^8,9 it is has been shown that SGLT2 inhibitor mediated cardioprotection occurs in an off-target manner.^10–13 As such, multiple mechanistic preclinical studies have been conducted, whilst clinical studies are also ongoing in an attempt to tease out the exact mechanistic pathways that may be responsible for the cardioprotection benefits observed in the aforementioned CVOTs. Indeed, the improved cardiovascular outcomes in the SGLT2 inhibitor CVOTs were driven by a reduction in hospitalization for heart failure, but the reasons behind this are unclear. An interesting observation from these trials was that the improvement in blood pressure occurred without compensatory improvement in heart rate, suggesting dampening of sympathetic nervous system (SNS) activity by SGLT2 inhibitors.^1,4,5,7 This was similarly shown in experimental studies where luseogliflozin significantly decreased blood pressure without any alterations in heart rate in hypertensive rats.¹⁴ Since autonomic dysfunction (characterized by hyperactivity of the SNS) and heart failure can both occur with diabetes, and this is linked with increased mortality,^15,16 it is possible that modulation of the cardiovascular autonomic function by SGLT2 inhibitors may account for the reduction in hospitalization for heart failure shown in the trials with different SGLT2 inhibitors. Furthermore, drugs that mitigate SNS activity, such as beta-blockers, have also been shown to improve outcomes in heart failure patients.^17–20

This review article explores in depth the existing preclinical and clinical studies focused on the role of SGLT2 inhibitors in modulating cardiovascular autonomic function as one of the possible pathways mediating their cardioprotective effects. A summary of the key findings from the preclinical and clinical studies investigating the role of SGLT2 inhibitors in modulating cardiovascular autonomic function is also shown in Table 1.

Current evidence from preclinical studies

Multiple preclinical studies have sought to evaluate the cardiovascular autonomic effects of SGLT2 inhibition. In this context, Chiba et al.²¹ showed that acute oral administration of dapagliflozin significantly suppressed noradrenaline turnover in the brown adipose tissue of mice within 18 h, which reflects the SNS activity in adipose tissue. Interestingly, Yoshikawa et al.²² found that ipragliflozin mitigated the arterial blood pressure lability (arterial blood pressure variability which frequently impacts negatively on the T2DM pathophysiology²³) in streptozotocin-induced diabetic rats. Of note, Matthews et al.²⁴ were the first to demonstrate that oral dapagliflozin reduced SNS activity markers (e.g. tyrosine hydroxylase and noradrenaline) in the kidneys and heart of mice fed a high-fat diet (HFD). Moreover, Wan et al.²⁵ showed that empagliflozin use in diabetic Otsuka Long Evans Tokushima (OLETF) rats improved the circadian rhythm of SNS activity (as measured by blood pressure recordings) via a sympatho-inhibitory effect during the sleeping period.

More recently, a series of elegant experiments further demonstrated the positive effects of SGLT2 inhibition on SNS activity.²⁶ In these studies, chemical denervation of the SNS with oxidopamine (6-OHDA) in a neurogenic hypertensive non-diabetic mouse model [Schlager (BPH/2J) mouse] was shown to reduce the expression of renal SGLT2 with improvements in both metabolic and haemodynamic parameters. The authors then introduced dapagliflozin (by oral gavage every 2 days for 2 weeks) to HFD-fed Schlager mice and found that in addition to improvements in blood glucose levels and blood pressure (similar to that shown in the relevant clinical trials), SGLT2 inhibition was associated with a reduction in cardiac and renal tyrosine hydroxylase staining (a reliable marker of SNS activity) and norepinephrine content.²⁶ In addition, SGLT2 inhibition was associated with improvement of endothelial function and reduced inflammation, documented by attenuated pro-inflammatory cytokines and increased cardiac and renal anti-inflammatory interleukin-10. In other words, the lower amounts of sympathetic activation with SGLT2 inhibitors may just be secondary to improvements in haemodynamics and metabolic stress. Collectively, the findings suggest a bidirectional relationship between SGLT2 inhibition and SNS activity.

Current evidence from clinical studies

Given the promising findings of the relevant preclinical studies, further studies aimed to evaluate the effects of SGLT2 inhibition on SNS activity in the clinical setting. Intriguingly, a relevant case report exists in the medical literature presenting the case of an 83-year-old patient with diabetic and chronic heart failure whose cardiac sympathetic nerve hyperactivity—a potentially useful parameter for evaluating the prognosis of heart failure,²⁷ which was assessed by I-123-metaiodobenzylguanidine cardiac-scintigraphy (I-123-MIBG)—was reduced following 12 months of ipragliflozin administration, with no heart failure related re-hospitalization 2 years after the initiation of this treatment.²⁸ More recently, two clinical studies that sought to evaluate the effects of an SGLT2 inhibitor on the cardiovascular autonomic function were published and are discussed in the following sections.

The EMPA-HEART CardioLink-6 Holter analysis study

The EMPA-HEART CardioLink-6 study was a randomized trial that compared empagliflozin vs. placebo on cardiac remodelling in patients with T2DM and established coronary artery disease (CAD), which documented that empagliflozin was associated with significantly decreased left ventricular mass at 6 months of treatment.²⁹ From this study, Garg et al.³⁰ extracted the collected Holter monitoring data (66 participants completed 24-h ambulatory Holter monitoring at baseline and 6 months) to conduct an exploratory analysis on the effects of empagliflozin on heart rate variability (HRV). These results were reported as the EMPA-HEART CardioLink-6 Holter analysis study which was the first clinical study to evaluate the effects of SGLT2 inhibition on cardiovascular autonomic function using HRV parameters from non-invasive Holter monitoring.³⁰ Of note, HRV is a measure of variations in time intervals between two consecutive sinus heart beats and constitutes a surrogate marker characterizing the sympathetic and parasympathetic aspects of the autonomic nervous system. For the clinical practice, a growing body of evidence has shown that a lower HRV is associated with poorer cardiovascular outcomes, such as myocardial infarction, heart failure, sudden cardiac death, and cardiovascular mortality.^31–33

Interestingly, the EMPA-HEART CardioLink-6 Holter analysis study found no significant difference in the HRV between the empagliflozin and placebo groups. Although the findings were contrary to the hypothesis that SGLT2 inhibition may have a positive impact on the SNS, as suggested by the aforementioned preclinical studies, these neutral findings may be attributed to a number of reasons. Indeed, since the study cohort consisted of subjects with stable CAD, it is plausible that more pronounced changes in the autonomic tone might be detected in patients with more advanced cardiac comorbidities, such as myocardial infarction and significant left ventricular dysfunction. Moreover, as this study presented a post hoc exploratory analysis, the analysed sample size was not powered to assess for pre-specified differences in the HRV parameters. Furthermore, although the high frequency component of HRV is largely influenced by the vagal tone, no HRV parameter is considered to be truly reflective of SNS activity, but instead the HRV parameters are felt to represent the interplay between the parasympathetic and sympathetic arms of the cardiovascular autonomic system.³⁴ Therefore, HRV parameters may not constitute the best method to assess primary changes in the SNS due to SGLT2 inhibition. Additional modalities, such as heart rate turbulence (HRT), heart rate recovery post-exercise, or imaging of the autonomic nervous system with radiotracers (as utilized in the aforementioned case report), could provide a more holistic assessment of the cardiovascular autonomic function, but understandably this was beyond the scope of the EMPA-HEART CardioLink-6 Holter analysis study. Finally, data on the doses of beta-blockers and anti-hypertensive agents were not obtained in the context of this study, and so it was not possible to determine if this could have affected HRV parameters in the studied participants.

The EMBODY trial

The EMBODY trial³⁵ was a more recent prospective, multicentre, randomized, double-blind, placebo-controlled trial which evaluated patients with T2DM and acute myocardial infarction (AMI) in Japan. In this study, 105 patients were randomized at 2 weeks post-AMI to either empagliflozin or placebo, and were subsequently followed up to 24 weeks. The EMBODY trial endpoints included changes in HRV, HRT, T-wave alternans (TWA), and late potentials (LP) as measured by 24-h ambulatory Holter monitoring, whilst cardiac sympathetic activity via I-123-MIBG myocardial scintigraphy was also assessed.

Based on the results of this study, there were no significant differences in the intergroup comparison; however, there were significant improvements in the parameters reflecting both sympathetic and parasympathetic nerve activity within the empagliflozin group. In particular, the standard deviation of all 5-min mean normal RR intervals (SDANN), square root of the mean of the sum of the squares of differences between adjacent RR intervals (r-MSSD) and high frequency were increased, reflecting improved parasympathetic nerve activity. Moreover, the low-to-high frequency ratio (a marker of sympathovagal balance) was reduced which reflects improved cardiac SNS activity. Significantly improved HRT was also noted in the empagliflozin group. Cardiac SNS activity, as reflected by I-123-MIBG myocardial scintigraphy, showed improvement in both study groups, reflecting the natural recovery course of AMI. Overall, the EMBODY trial was the first randomized clinical study to demonstrate improvement in both cardiac sympathetic and parasympathetic activity by SGLT2 inhibition. The difference in the underlying cardiac comorbidity of the recruited patients with T2DM—patients with stable CAD in the EMPA-HEART Link-6 Holter analysis vs. patients with acute coronary syndrome in the EMBODY trial—may be one of the reasons that could explain the differing findings from these trials.

Current evidence for reduction in sympathetic activity and arrhythmia risk with SGLT2 inhibitors

One of the potential translational effects of SNS activity reduction caused by SGLT2 inhibition is a subsequent reduction in arrhythmias. Diabetes is associated with the development of arrhythmias such as atrial fibrillation (AF), which constitute a key risk factor for sudden cardiac death and arrhythmic deaths.^36,37 Furthermore, the manifestation of arrhythmias in patients with diabetes is also associated with a worse prognosis.³⁸ The suggested mechanisms for the increased risk of arrhythmias include myocardial fibrosis, oxidative stress and inflammation leading to cardiac electromechanical and autonomic remodelling.³⁹ At present, there are still no SGLT2 inhibitor clinical trials or basic science studies that have investigated the drug as a pure anti-arrhythmic agent but the following observational studies attempt to address this gap in our knowledge.

Real-world data: a nationwide population-based longitudinal cohort study

As none of the CVOTs for SGLT2 inhibitors included the arrhythmic burden as a primary outcome, Chen et al.⁴⁰ attempted to assess the risk of new-onset arrhythmias and all-cause mortality with SGLT2 inhibitors by utilizing real-world data from the Taiwanese National Health Insurance Research Database. In the context of this study, empagliflozin and dapagliflozin were the SGLT2 inhibitors used at that time, whilst the range of studied arrhythmias included AF and other supraventricular and ventricular arrhythmias based on the International Classification of Diseases, 9th and 10th revision, Clinical Modification (ICD-9-CM and ICD-10-CM). The study only looked at two outcome measures: all-cause mortality and new onset arrhythmia.

Interestingly, the study revealed that SGLT2 inhibitors were associated with a 17% reduction in new-onset arrhythmias [adjusted hazard ratio of 0.830 (95% confidence interval, 0.751–0.916)] and 46% reduction in all-cause mortality [adjusted hazard ratio of 0.547 (95% confidence interval, 0.482–0.621)]. In terms of the breakdown of the new-onset arrhythmias, the occurrence of AF, supraventricular arrhythmias and ventricular arrhythmias were lower in the SGLT2 inhibitor group but did not reach statistical significance [adjusted hazard ratios—AF: 0.841 (0.662–1.068), supraventricular arrhythmias 0.815 (0.570–1.167), and ventricular arrhythmias 0.797 (0.525–1.208)].

Other studies

A protective effect of SGLT2 inhibitors on AF was also reported by other recent studies. Indeed, the post hoc analysis of the DECLARE-TIMI 58 trial showed that dapagliflozin was associated with a 19% reduction in AF and atrial flutter.⁴¹ This was independent of the presence of AF/atrial flutter, atherosclerotic cardiovascular disease or history of heart failure at baseline.

In addition, Bonora et al.⁴² interrogated a large pharmacovigilance database (the FDA adverse event reporting system) to show that the reporting of AF was significantly lower with SGLT2 inhibitors, suggesting a protective role against the onset of AF. Taken together, the findings of these clinical studies suggest the exciting potential of SGLT2 inhibitors having clinically relevant anti-arrhythmic properties.

More recently, a meta-analysis of 34 randomized controlled trials with 63 166 patients with either T2DM or heart failure found that SGLT2 inhibitors (canagliflozin, dapagliflozin, empagliflozin, or ertogliflozin) were associated with a 19% reduced risk of incident atrial arrhythmias and 28% reduction in sudden cardiac death.⁴³

SGLT2 inhibition may also have a positive impact in reducing the risk of developing ventricular arrhythmias. A recent post hoc analysis of the DAPA-HF trial found that dapagliflozin was associated with an improvement in composite outcome (5.9% compared with 7.4%, hazard ratio 0.79, 95% confidence interval, 0.63–0.99, P = 0.037) which consisted of the first occurrence of any serious ventricular arrhythmias, resuscitated cardiac arrest or sudden death.⁴⁴

However, the association found in the post hoc analysis of DAPA-HF was not observed in a very recent meta-analysis of 19 randomised controlled trials (RCT) with 55, 590 participants (45). Sfairopoulos et al showed that SGLT2 inhibitors were not significantly associated with a lower risk of sudden cardiac death [9 RCTs, risk ratio (RR) 0.74, 95% confidence interval (CI) 0.50-1.08; P = 0.12] and ventricular arrhythmias (17 RCTs, RR 0.84, 95% CI 0.66-1.06; P = 0.14).⁴⁵ However, it is worth noting that the number of sudden cardiac death and ventricular arrhythmia events were relatively few leading to wide confidence intervals and the point estimates suggested potential benefits, hence the need for further research in this area.

Ongoing clinical trials

The EMPYREAN study is an ongoing prospective, randomized, open-label, multicentre trial of 134 patients with T2DM in Japan.⁴⁶ Participants of this study are randomly allocated to receive either empagliflozin or sitagliptin (a dipeptidylpeptidase-4 inhibitor), whilst ambulatory Holter monitoring is performed at baseline and at 12 and 24 weeks of treatment. Similarly to the EMPA-HEART CardioLink-6 Holter analysis study, HRV will be analysed in the EMBODY trial with the primary endpoint being the change in the low-frequency (0.04–0.15 Hz) to high-frequency (0.15–0.4 Hz) ratio from baseline to 24 weeks.

Limitations

The autonomic nervous system is highly complex in terms of its anatomy and physiology and attempts to evaluate the effects of SGLT2 inhibitors on specific components of this sophisticated system has been challenging. Therefore, it is unsurprising that much of the conclusions drawn from the current clinical studies looking into the effects of SGLT2 inhibition on the cardiovascular autonomic nervous system modulation relied on indirect measures of autonomic function and the results have been at times, contradictory. It is also unclear if the improvement in autonomic nervous system parameters is the cause or simply an association to the improvement in cardiovascular function. In other words, the improvement in haemodynamic parameters and metabolism could either be an effect of the reduction in sympathetic activity or vice versa. In addition, there have been other mechanisms demonstrated by other research groups, which may confound the effects of autonomic nervous system modulation and reduction in arrhythmia risk observed in the clinical studies. For example, dapagliflozin (in animal studies) has been shown to reduce myofibroblast infiltration and myocardial fibrosis (which can lead to a reduction in arrhythmic substrate) via the activation of known mediator of cardioprotection, the janus kinase—signal transducer and activator of transcription 3 (JAK-STAT3) pathway.⁴⁷ It is also worth noting that unlike the SGLT2 inhibitor CVOTs, the clinical studies that attempted to evaluate the reduction in arrhythmia risk by SGLT2 inhibitors as described in this article were not prospective or randomized and not designed to look at the reduction in arrhythmia onset, burden or arrhythmic death. Unlike recent SGLT2 inhibitor trials like the DAPA-HF,⁷ EMPEROR-Reduced,⁴⁸ and EMPEROR-Preserved⁴⁹ which evaluated a specific heart failure patient population, similarly designed clinical trials are needed in a cardiac arrhythmia patient population.

Future pre-clinical and clinical studies should therefore aim to address the aforementioned limitations when designing their studies.

Future perspectives and conclusions

Although the exact mechanism(s) behind the cardioprotection noted with treatment with the different SGLT2 inhibitors have yet to be established, both the preclinical and clinical studies discussed in this review highlight the potential role of SGLT2 inhibitors in modulating cardiovascular autonomic function. Additional studies from bench to bedside are clearly required to understand how SGLT2 inhibitors exert their off-target effects on the autonomic nervous system, and how this modulation translates into improved cardiovascular outcomes. In terms of the potential of SGLT2 inhibitors to be repurposed as anti-arrhythmic agents, there is still a great deficiency in basic science and clinical studies to support this. In conclusion, more pre-clinical and clinical studies are required to address this gap in our knowledge on the role of SGLT2 inhibitors as modulators of cardiac autonomic function.

Acknowledgements

All authors have made an important scientific contribution to the study and have assisted with the drafting or revising of the manuscript, in accordance with the definition of an author as stated by the International Committee of Medical Journal Editors (ICMJE) at http://www.icmje.org/recommendations. We confirm that neither the manuscript nor any part of it has been published or is under consideration for publication elsewhere and that there are no conflicts of interests or funding to report. The guarantor for this article is Professor Faizel Osman.

Conflict of interest: none declared.

References