Differential Pathophysiological Mechanisms in Heart Failure With a Reduced or Preserved Ejection Fraction in Diabetes

State-of-the-art Review

Milton Packer, MD


JACC Heart Fail. 2021;9(7):535-549. 

In This Article

Differential Mechanisms of Action of SGLT2 Inhibitors in HFrEF and HFpEF

If the mechanisms (outlined in the previous text) lead to HFrEF and HFpEF in type 2 diabetes, then drugs that reduce the risk of HFrEF and HFpEF in patients with diabetes should act to mitigate these mechanisms.

The only class of drugs that reduces the risk of serious heart failure events in type 2 diabetes are the SGLT2 inhibitors (Table 1, Supplemental Appendix). In 4 trials of patients with type 2 diabetes who largely did not have heart failure at enrollment, SGLT2 inhibition with empagliflozin, dapagliflozin, canagliflozin, and ertugliflozin reduced the risk of heart failure hospitalizations. In 3 trials of patients with chronic kidney disease, SGLT2 inhibition with canagliflozin, dapagliflozin, and sotagliflozin reduced the risk of heart failure hospitalizations in patients who had diabetes. In 3 trials of patients with established heart failure, SGLT2 inhibition with dapagliflozin, empagliflozin, and sotagliflozin reduced the risk of heart failure hospitalization among participants with diabetes (and in those without diabetes). The magnitude of the benefit on heart failure events across the ten trials was strikingly similar—a 30%-35% reduction in risk.

Much of the evidence for a benefit of SGLT2 inhibitor on heart failure events pertains to patients with HFrEF. Participants with HFrEF responded most favorably to dapagliflozin in a trial of type 2 diabetes,[92] and patients with HFrEF (with or without diabetes) benefited from treatment with dapagliflozin and empagliflozin (Table 1). There is also evidence for a reduction in heart failure events in patients with HFpEF in a subgroup analysis of a trial with canagliflozin and in a subgroup analysis of a heart failure trial with sotagliflozin.[93] A recent press release has indicated that a large-scale trial with empagliflozin in patients with well-characterized HFpEF achieved its primary endpoint;[94] a similar trial with dapagliflozin is ongoing.

None of the large-scale trials with SGLT2 inhibitors were designed to identify the pathophysiological mechanisms by which these drugs reduce the risk of heart failure events. However, experimental and clinical studies have described effects of SGLT2 inhibitors: 1) to enhance nutrient deprivation signaling and autophagy; 2) to inhibit NHE1 in the heart and NHE3 in the kidney; and 3) to reduce epicardial adipose tissue volume and proinflammatory adipocytokine secretion, and thereby, mitigate cardiac inflammation, microcirculatory dysfunction, and fibrosis (Central Illustration).

Central Illustration.

Sodium-Glucose Cotransporter 2 Inhibitors Interfere With the Principal Mechanisms by Which Diabetes Can Promote the Development and Progression of Cardiomyopathy
Sodium-glucose cotransporter 2 inhibitors are capable of interfering with all 3 of the primary pathophysiological mechanisms by which type 2 diabetes can lead to heart failure with a reduced or preserved ejection fraction. These include the following actions: 1) up-regulating nutrient deprivation signaling (eg, sirtuin-1 and its downstream effectors); 2) interfering with the actions of sodium-hydrogen exchangers in the heart and kidneys; and 3) reducing the mass and proinflammatory activity of epicardial adipose tissue. The net result of all 3 effects is to slow the development of cardiomyopathy and reduce the risk of serious heart failure events.

Effect of SGLT2 Inhibitors on Nutrient Deprivation Signaling and Autophagic Flux

SGLT2 inhibitors act on SGLT2 protein, a nutrient surplus sensor located in the proximal renal tubule. The resulting renal glycosuria leads to a substantial loss of calories in the urine, inducing a state of perceived starvation, which then triggers the up-regulation of nutrient deprivation signals (particularly SIRT1) and an increase in autophagic flux.[35,95–98] SGLT2 inhibitors have been shown to stimulate SIRT1 and its downstream effectors in a broad range of tissues, including the heart, kidney, and liver.[36,99–105] Activation of SIRT1 is likely responsible for the gluconeogenesis and ketogenesis seen with SGLT2 inhibitors, because SIRT1/PGC-1a/FGF21 regulates the rate-limiting enzyme for these metabolic pathways.[106] Furthermore, SIRT1-mediated activation of hypoxia inducible factor-2α in the kidney and liver promotes the production of erythropoietin,[107] an action that explains the increase in hematocrit seen in trials with SGLT2 inhibitors.[108]

How can SGLT2 inhibitors to promote SIRT1 signaling in the heart if SGLT2 is not expressed in the myocardium? SGLT2 inhibitors have profound effects in many organs that do not express SGLT2, including cardiac and skeletal muscle, liver, and adipose depots.[99,105,109–112] The urinary loss of calories triggers a fasting-like transcriptional paradigm in tissues throughout the body,[98,105] and activation of SIRT1 is a cornerstone of the organismal response to starvation.[106] Additionally, because SGLT2 may act as a central sensor for a general state of nutrient surplus, suppression of its activity may led to the up-regulation of SIRT1 even if there is only modest glycosuria.[36,100] Interestingly, the fasting-like transcriptional paradigm induced by SGLT2 inhibitors may involve nutrient deprivation signals other than SIRT1/PGC-1a/FGF21. SGLT2 inhibitors may inhibit Akt/mTOR,[105] and certain SGLT2 inhibitors may also activate AMPK.[113,114] However, it seems unlikely that AMPK plays a dominant role in mediating the effects of SGLT2 inhibitors, because AMPK suppresses gluconeogenesis, ketogenesis, and erythrocytosis,[115,116] a profile that is opposite of that produced by SGLT2 inhibitors.

The concerted systemic activation of nutrient deprivation signaling (particularly SIRT1 and its downstream effectors) likely underlies the ability of empagliflozin, dapagliflozin, and canagliflozin to promote autophagic flux in many organs, including the heart,[117–121] both in diabetic and in nondiabetic models. Such an action may explain why SGLT2 inhibitors reduce oxidative stress, normalize mitochondrial structure and function, and mute proinflammatory pathways in the stressed myocardium.[118,121–123] Enhanced autophagy may also account for the ability of SGLT2 inhibitors to ameliorate ischemia-reperfusion injury and postinfarction remodeling and to minimize microvascular dysfunction, hypertrophy, and fibrosis; enhance contractile performance; and ameliorate the course of experimental cardiomyopathy.[109,110,123]

Effect of SGLT2 Inhibitors on Renal NHE3 and Cardiac NHE1

In the proximal renal tubule, SGLT2 colocalizes with and functionally interacts with NHE3,[124,125] which is responsible for the majority of sodium tubular reuptake following filtration. Knockdown of NHE3 interferes with the expression of SGLT2, and conversely, SGLT2 inhibitors interfere directly with the actions of NHE3.[126,127] The natriuresis following NHE3 inhibition is limited by a compensatory increase in the absorptive capacity for sodium in distal parts of the nephron, a response that can theoretically be blocked (in part) by the use of loop diuretic agents. However, in the clinical setting, SGLT2 inhibitors have produced no or little change in urinary sodium excretion during short-term treatment, with or without furosemide.[128–131] In large-scale clinical trials, changes in body weight with the use of SGLT2 inhibitors have been modest, even when these drugs have been administered to patients with chronic heart failure who are volume overloaded and receiving loop diuretic agents (Supplemental Appendix). Patients with heart failure and recent volume overload are not more sensitive to the ability of SGLT2 inhibitors to reduce heart failure hospitalizations, suggesting that diuresis is not a dominant mode of action of these drugs to reduce the risk of serious heart failure events.[132] Therefore, the decline in body weight during SGLT2 inhibitor use may be related to the loss of calories in the urine rather than an increase in sodium excretion,[133,134] because there is no relation between changes in body weight and changes in circulating natriuretic peptides during treatment with these drugs in patients with chronic heart failure.[132]

It is intriguing to postulate that SGLT2 and NHE are intertwined in the heart as they are in the kidneys. Several studies have shown that SGLT2 inhibitors reduce intracellular sodium concentrations in cardiomyocytes, an action that has been attributed to an effect of these drugs to inhibit NHE1 in the myocardium, possibly because NHE1 has a docking site for an interaction with SGLT2 inhibitors.[135,136] SGLT2 inhibitors reduces both intracellular sodium and calcium,[135–137] an effect that is blocked by treatment with the NHE1 inhibitor, cariporide. Inhibition of NHE1 might be expected to mitigate cardiomyocyte injury and attenuate the development of cardiac hypertrophy, fibrosis, remodeling, systolic dysfunction, and heart failure.[58,59] NHE1 inhibition may also ameliorate the abnormalities in the vasculature that may contribute to HFpEF;[48,49] of note, SGLT2 inhibitors have favorable effects on coronary arterial endothelial function, arterial stiffness, and diastolic cardiac filling in diabetes.[138–140] The actions of SGLT2 inhibitors to inhibit NHE1 have been confirmed by some investigators,[137,140–142] but not by others.[143] It is therefore noteworthy that the reduction in intracellular sodium produced by SGLT2 inhibitors may represent an indirect effect of enhanced nutrient deprivation signaling rather than a direct drug action.[123]

Effect of SGLT2 Inhibitors on Epicardial Adipose Tissue and the Secretion and Actions of Proinflammatory Adipocytokines

The expansion of epicardial adipose tissue in type 2 diabetes may act to focus the effects of systemic metabolic disorders onto the adjoining myocardium; the resulting secretion of proinflammatory adipocytokines leads to inflammation, microcirculatory dysfunction, and fibrosis, the hallmarks of HFpEF. Although the urinary loss of calories following SGLT2 inhibition causes only modest decreases in body weight, it causes a striking shrinkage of fat depots throughout the body, with a particular effect on visceral fat depots.[144–146] These effects are not related to an antihyperglycemic effect, because they are not seen with drugs that have more pronounced effects to lower blood glucose.[146,147] Importantly, the action of SGLT2 inhibitors on visceral fat depots includes an effect to reduce the mass of epicardial adipose tissue, an effect that has been observed in clinical trials with different members of the drug class.[111,148–152] The reduction in epicardial adiposity may underlie observations that SGLT2 inhibitors are capable of mitigating myocardial inflammation, microcirculatory dysfunction, and fibrosis and improving LV diastolic filling dynamics in both experimental models and in the clinical setting.[123,153–157]

Leptin is secreted by biologically active adipocytes; circulating leptin levels are correlated with epicardial fat mass[87] and are increased in patients with HFpEF;[88,89] and leptin can contribute to the pathophysiological abnormalities of HFpEF by effects on calcium handling, cardiac fibrosis, and sodium reabsorption in the renal tubules.[66,90,91] It has been hypothesized that—as a result of their effects on adipose tissue mass and biology—SGLT2 inhibitors may function as leptin antagonists.[158] Experimentally, SGLT2 inhibition can suppress the secretion of leptin by adipocytes,[159] and treatment with SGLT2 inhibitors reduces circulating levels of leptin in clinical trials of patients with type 2 diabetes or other insulin-resistant states, an effect that is disproportionate to the magnitude of weight loss.[160–162] In a study of aptamer-based proteomics, empagliflozin changed the expression of 43 of 3,713 proteins, one of which was the leptin receptor.[163] Some have suggested that a homeostatic relationship exists between the activity of leptin and that of glucose transporters, which may be influenced by SGLT2 inhibitors.[164]