Strategies of Unloading the Failing Heart from Metabolic Stress

Abstract

We propose a unifying perspective of heart failure in patients with type 2 diabetes mellitus. The reasoning is as follows: cellular responses to fuel overload include dysregulated insulin signaling, impaired mitochondrial respiration, reactive oxygen species formation and the accumulation of certain metabolites, collectively termed glucolipotoxicity. As a consequence, cardiac function is impaired with intracellular calcium cycling and diastolic dysfunction as an early manifestation. In this setting, increasing glucose uptake by insulin or insulin sensitizing agents only worsens the disrupted fuel homeostasis of the heart. Conversely, restricting fuel supply by means of caloric restriction, surgical intervention, or certain pharmacologic agents will improve cardiac function by restoring metabolic homeostasis. The concept is borne out by clinical interventions, all of which unload the heart from metabolic stress.

Introduction

Work performed over the last few decades has elucidated several mechanisms associated with the development of heart failure in the setting of obesity and diabetes (1) (2) (3). In this review, we briefly summarize current knowledge on the pathophysiology of non-ischemic heart failure in the state of metabolic dysregulation. Based on the hypothesis that insulin resistance is a marker, but not a mediator of heart failure, we propose a unifying concept for the effective management of heart failure in obesity and diabetes, which targets restoration of metabolic homeostasis by restricting fuel supply at its source. We propose a concept of metabolic unloading which is rooted in the literature on cardiac metabolism.

Fuel metabolism in the healthy heart

The healthy heart utilizes a wide spectrum of energy providing substrates (4). Ultimately, cardiomyocytes generate the majority of ATP via oxidative phosphorylation of ADP from reducing equivalents in the mitochondria. Insulin plays a modulating role in the metabolism of cardiomyocytes when multiple substrates are present (4) (5). While glucose is the preferred substrate postprandially, the heart’s preferred substrates for respiration in the fasted states are free fatty acids (FFA) (6). In addition, during protracted fasting, ketone bodies and amino acids also contribute to the heart’s production of ATP (6). This metabolic flexibility and adaptive transition from one fuel substrate to another is a defining feature of myocellular homeostasis, where concentrations of cellular metabolites remain largely unperturbed while metabolic flux rates change in response to environmental stimuli (Figure 1A).

Figure 1:

A. Schematic diagram summarizing the principle of fuel homeostasis in the heart. In a series of heavily regulated enzyme catalyzed reactions (blue arrow), energy providing substrates (fuel) are converted to ATP which supplies energy (orange arrow) for pump function of the heart (contraction/relaxation) (4) (6). B. Disrupted Homeostasis. Excess fuel supply resulting in the accumulation of non-oxidized metabolites in the heart and other organs (fuel toxicity) (72) (10). C. Metabolic unloading through restriction of fuel supply (73) (74) (75). D. Metabolic unloading through reduced production or diversion of excess fuel (76) (77).

Disrupted fuel homeostasis in the metabolically stressed heart

Metabolic flexibility and adaptation are impaired when the heart is metabolically challenged by obesity, diabetes and inflammation (7). Although high FFA levels suppress glucose uptake by the heart (and vice versa), glucose and FFA are both elevated in obese individuals with uncontrolled diabetes (Figure 1B) (8). In short, there is an oversupply of substrates.

When increased glucose uptake by the myocardial cells is not matched by increased glucose oxidation, it results in the production of reactive oxygen species (ROS), oxidative stress, and glycosylation of proteins. AMP-activated protein kinase (AMPK) is inhibited by the increased ATP/AMP ratio, which subsequently suppresses FFA oxidation and favors intracellular accumulation of fatty acid metabolites. Glucose overload leads to increased glycogen deposition and activation of the toxic polyol, and hexosamine pathways. Prolonged exposure of proteins to hyperglycemia causes a series of non-enzymatic reactions which promotes the formation of advanced glycation end products (AGEs) (9). In addition, increased levels of malonyl-CoA prevent fatty acyl-CoA (FACoA) entry into the mitochondria and direct its use for complex lipid synthesis, such as glycerolipids (phospholipids, diacylglycerols, and triglycerides) and ceramides. The latter is reflected in lipid droplets and endoplasmic reticulum distress. Accumulation of ROS and toxic lipids, as well as reduced AMPK activity contribute to activation of the inflammasome, causing further cardiac injury (10). Lastly, myosin isoform switches and impaired myocyte calcium homeostasis occur as well in setting of fuel overload (11) (12). Collectively, the above mechanisms, termed as glucolipotoxicity, eventually lead to decreased ATP production, impaired contraction and relaxation, myocellular hypertrophy, and fibrosis (Figure 1B) (13). We propose that restoring fuel homeostasis early in the process presents a window of opportunity for potentially preventing the maladaptive cardiac remodeling.

The hypothesis of insulin resistance as an adaptive cardioprotective response

It is known that in response to hyper-alimentation, muscle becomes insulin resistant, resulting in diversion of nutrients to adipose tissue (14). Insulin resistance is the result of dysregulated insulin signaling, decreased glucose transporter protein 4 (GLUT4) expression, decreased GLUT4 translocation, and impaired glucose uptake by the myocyte. In addition to abnormal insulin receptor substrate (IRS) phosphorylation, there is activation of protein kinase C (PKC) isoforms by ROS and accumulation of FFA metabolites including diacylglycerol, fatty acyl-CoA, and ceramides (15).

Although the presence of insulin resistance is a prominent feature in heart failure, no causality has been established. Rather, we and others have proposed that insulin resistance protects the heart from fuel overload and the deleterious effects of glucolipotoxicity (16). Furthermore, overriding the defense mechanism of insulin resistance by iatrogenically increasing insulin sensitivity or administering exogenous insulin may be harmful and might explain the unexpected results of recent clinical trials we describe below. With continued nutrient excess, the protective mechanisms of insulin resistance are lost and functional and structural changes ensue.

Exacerbation of Myocardial Fuel Toxicity by Insulin, and Insulin Sensitizing Agents

The clinical goal for treatment of patients with type 2 diabetes mellitus is to achieve and maintain a hemoglobin A1c close to the physiologic range (<7%). Maintaining hemoglobin A1c at this level prevents microvascular disease, but whether it prevents macrovascular disease or heart failure may depend on the type of antidiabetic treatment employed (17).

Exogenous insulin administration has a multifaceted impact on metabolism, atherogenesis, and inflammation. The vast majority of clinical trials evaluating the effect of insulin treatment on cardiovascular outcomes have failed to show benefit (18). Only the 10 year United Kingdom Prospective Diabetes Study [UKPDS]) showed a modest late reduction in myocardial infarction and death, with insulin being inferior compared to metformin (19).

From a metabolic perspective, the promotion of myocardial glucose uptake by insulin may be unfavorable for the failing heart, especially when aggressive glycemic goals are targeted. This was first noted in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) study which highlighted the previously poorly recognized harm of intense glycemic control in high-risk patients with type 2 diabetes (20). Of note, glycemic targets in ACCORD study were achieved mostly with insulin and insulin secretagogues, while the undesired outcomes were unrelated to hypoglycemia. The finding was further supported by subsequent studies examining the association of HbA1C with mortality in diabetic patients with heart failure, showing increased mortality in aggressively treated patients (21).

Thiazolidinediones (TZD) modulate the family of ligand activated receptors PPARs and achieve hypoglycemia through enhanced insulin-stimulated glucose uptake (22). Despite improving glucose control and insulin resistance, TZDs have been linked to greater rates of heart failure hospitalization in multiple clinical trials (23) (24). The adverse impact of TZDs came initially as a surprise, but it can be explained if we consider that they augment insulin signaling, leading to metabolic overload and expose the myocytes to the deleterious effects of glucolipotoxicity (25) (26). Antidiabetic therapeutics that stimulate insulin release or increase insulin sensitivity forcing the substrates into the cells, actually worsen glucolipotoxicity.

Strategies for Metabolic Unloading

A growing body of experimental and clinical evidence, supports the fuel overload hypothesis, and proposes a different view on the management of heart failure in the context of type 2 diabetes mellitus. Here, the role of nutrient excess and subsequent metabolic stress is emphasized and addressed with similar attention as hemodynamic and neurohumoral stress. With disrupted fuel homeostasis, it is the cellular fuel supply that requires modulation rather than the downstream effects of such excess. In principle, those goals can be achieved either by restricting nutrient provision (Figure 1C) or by eliminating excess fuel through the kidneys (Figure 1D). The former deploys mostly non-pharmacological interventions, the latter amplifies physiologic glycosuria as a means of pharmacological unloading of the heart from metabolic stress.

Restriction of Fuel Influx - Non-Pharmacologic Interventions

A. Caloric restriction, Intermittent Fasting

The benefit of intermittent fasting was first shown in rat studies (27) (28). Mechanisms responsible for these benefits in otherwise healthy individuals are not entirely understood but may include interference of gut microbiota (29). Caloric restriction has also positive effects on metabolic health and body composition, improves metabolism efficiency, and reduces oxidative stress both in rodents (30) and humans (31). Time-restricted feeding specifically, is associated with reduction in body weight, total cholesterol, triglycerides, interleukin-6 (IL-6) and TNF-α (32) as well as improvement in insulin sensitivity (33) (34). The above result in maintenance of blood glucose in physiologic range, depletion or reduction of glycogen stores, mobilization of fatty acids and generation of ketone bodies, reduction of circulating leptin and often elevation of adiponectin levels (35). Significant reductions in circulating lipids, blood pressure, and carotid intima-media thickness have also been noted (36).

As early as 1916, Dr. Elliot Joslin postulated that in patients temporary periods of under-nutrition are helpful in the treatment of diabetes (37). Since that time, observations have been made supporting the hypothesis that both insulin resistance and beta cell failure are reversible (38). Caloric restriction in individuals with type 2 diabetes results in reduced pancreatic and liver triglyceride content and to normalization of both beta cell function and hepatic insulin sensitivity (39). Thus dietary intervention aimed at weight loss is now the first line treatment of type 2 diabetes.

Currently, only limited data are available to quantify the therapeutic effects of fasting on cardiac structure and function. In one animal model of metabolic syndrome, caloric restriction attenuated cardiac remodeling and diastolic dysfunction (40). In a smaller study involving 12 obese patients with type 2 diabetes subjected to a very low calorie diet (450 kcal/day), a BMI reduction from 36 kg/m² to 28 kg/m² resulted in a reduction in myocardial triglyceride content, and improvement in left ventricular diastolic function (41).

B. Bariatric Surgery

Bariatric surgery (Figure 1C) is currently offered to patients with clinically severe obesity and offers the most effective modality for metabolic unloading (42). The benefits of weight loss surgery clearly extend beyond any mechanical aspects. Weight loss surgery profoundly impacts the physiology of energy balance and insulin sensitivity (43), is associated with microbiome and bile acid composition changes (44), and leads to extensive alterations in gut function (45). The bowel microbiota interacts with the host’s energy metabolism and immune system. Enteral bacterial composition is substantially altered in subjects undergoing Roux-en-Y Gastric Bypass (RYGB), and fecal transplantations from RYGB-treated mice and human subjects to germ-free mice induced weight loss compared to fecal transplants from sham-operated mice (43) (46). In addition, enteral hormones such as glucagon-like peptide-1 (GLP-1), glucagon-like peptide-2 (GLP-2), glucose-dependent insulinotropic peptide (GIP), oxyntomodulin (OXM) and peptide-YY (PYY) are all involved, but their exact interaction is not quite clear.

In line with the principles of disrupted fuel homeostasis, there is an inverse relationship between plasma FFA concentrations and left ventricular diastolic function in individuals with severe obesity (47). Following weight loss surgery, there is a reversal of insulin resistance, improved left ventricular systolic and diastolic function, and dramatic changes in skeletal muscle metabolism including alterations of fatty acid oxidation, which are mediated by adipose-derived hormones and cytokines that exert control over metabolic gene expression (48). Over a period of two years after bariatric surgery, patients with clinically severe obesity lost weight and exhibited a progressive decrease in ventricular wall thickness and left ventricular mass index (49).

Although the mechanisms are not entirely clear, evidence has accumulated to support use of bariatric surgery in patients with diabetes mellitus. The Surgical Treatment and Medications Potentially Eradicate Diabetes Efficiently (STAMPEDE) trial, showed that gastric bypass and sleeve gastrectomy were superior to intensive medical therapy in type II diabetes mellitus with impressive rates of glycemic control after gastric bypass (42% of patients) (50). In the Scandinavian Obesity Surgery Registry including > 25,000 individuals, gastric bypass surgery was associated with lower heart failure incidence compared to intensive lifestyle modification (HR 0.54, 95% CI 0.36–0.82) (51). In another study of 524 individuals with obesity and heart failure, bariatric surgery was associated with reduced incidence of subsequent heart failure exacerbation (52). Collectively, the experience with bariatric surgery strongly supports the hypothesis that restriction of nutrient fuel supply improves both, cardiac function and longevity (Figure 1C).

Restriction of Fuel Supply - Pharmacologic Strategies

A. Inhibiting gluconeogenesis: Metformin

Metformin inhibits hepatic glycogenolysis and gluconeogenesis. Metformin also changes several steps in glucose metabolism (53). Inhibition of glucose 6-phosphatase and phosphoenolpyruvate carboxykinase (PEPCK) results in further reductions in glucose production (54). Furthermore, metformin modulates the composition of gut microbiota and bowel wall permeability likely modulating gluconeogenesis signaling via AMPK-dependent pathway and insulin sensitivity (55). A novel formulation of metformin has shown that metformin acts on the bowel itself, including the L cells which are a source of glucose-regulating and satiety-regulating hormones including glucagon-like peptide-1 (GLP-1) and peptide YY (PYY) (56).

Although no randomized controlled trials have been performed to assess the role of metformin in heart failure, several observational data support benefit. Compared with sulfonylureas, metformin use in individuals newly treated for diabetes was associated with reduced rates of heart failure (57). Similarly, multiple observational studies of individuals with established heart failure have shown that metformin is associated with reduced mortality (58). The use of metformin has also been associated with echocardiographic improvement of diastolic function (59), a hallmark of diabetic cardiac dysfunction than has been used as an indirect measurement of metabolic overload, as it precedes the decrease in left ventricular ejection fraction in diabetic patients.

B. Modulation of gastric emptying: Glucagon-like peptide (GLP-1) analogs and dipeptidylaminopeptidase-4 (DPP-4)

Another pharmacologic strategy to reduce fuel supply is the use of incretin analogues. These gut-derived metabolic hormones are released by the gut in response to nutrients and amplify insulin secretion response and inhibit glucagon secretion and gluconeogenesis. GLP-1 exerts its function through GLP-1 receptors in the pancreas, liver, heart, gastrointestinal tract, adipose tissue as well as central nervous system and, importantly reduces appetite and food intake as well as delays gastric emptying (60). GLP-1 agonists operate by several mechanisms of action and some of the encouraging effects may result from restriction of substrate availability (61). On the myocardial level, 6-month treatment with liraglutide improves diastolic function in individuals with type 2 diabetes mellitus compared with those treated with other hypoglycemic drugs (62). In individuals with type 2 diabetes mellitus with established or high-risk for cardiovascular disease enrolled in the LEADER trial, liraglutide was associated with reduced major adverse cardiovascular events (63). Although initial studies in animals (64) and humans have suggested potential beneficial effects in those with established systolic heart failure randomized controlled trials of liraglutide in those with advanced systolic heart failure failed to show a benefit in clinical outcomes and showed a trend towards more adverse clinical events (65) (66).

The majority of the trials with dipeptidyl-aminopeptidase-4 (DPP-4) inhibitors, which prevent degradation of native GLP-1, have shown non-inferiority to placebo in cardiovascular outcomes, but have not shown superiority (67). Furthermore, most of the trials have not demonstrated increased heart failure risk with DPP-4 inhibitors, compared with placebo, except for SAVOR-TIMI 53 trial, a finding largely unexplained (67). The predominately neutral trials of DPP-4 inhibitors do not come as a surprise since this class does not confer a large effect on appetite and gastric emptying.

C. Curbing Excess Fuel Supply by Promoting Glucose Efflux: Sodium/Glucose Cotransporter-2 Inhibitors

Sodium/glucose cotransporter-2 (SGLT2) inhibitors operate by blocking the reabsorption of glucose in the kidney, thereby eliminating excess fuel from the blood stream and averting intracellular influx and disrupted fuel homeostasis (Figure 1D). The beneficial effects of SGLT2 inhibitors extend beyond the modest effects on glycemia and arterial pressure. Shunting glucose in the urine, offers potent metabolic unloading in the setting of nutrient excess.

By now many studies have provided evidence supporting the beneficial cardiac effects of SGLT2 inhibitors in type 2 diabetes mellitus (68) (69). The EMPA-REG trial demonstrated modest reductions in blood pressure and weight, improvement in glucose control, but most importantly resulted in an impressive 38% reduction in cardiovascular mortality, a 35% relative risk reduction in hospitalization for heart failure and a 32% relative risk reduction in death from any cause. These evolutionary results cannot be simply attributed to minor improvements in metabolic parameters or the diuretic action of the medication. Ex juvantibus, the study provides evidence to support the fuel excess hypothesis as the SGLT2 inhibitors directly modulate substrate influx. Relative to other pharmacologic agents, SGLT2 inhibitors confer unprecedented metabolic unloading effects, only seen before in bariatric surgery. The favorable effects of SGLT2 inhibitors to reduce cardiovascular and heart failure events were confirmed in the CANVAS (Canagliflozin Cardiovascular Assessment Study) and DECLARE-TIMI 58 (Dapagliflozin Effect on Cardiovascular Events–Thrombolysis in Myocardial Infarction 58) studies (69,70). In a small study of individuals with type 2 diabetes, treatment with empagliflozin for 3 month was associated with a significant reduction in left ventricular mass index and improved diastolic function (71). The mechanisms through which empagliflozin improves diastolic dysfunction and regresses LV mass are not known. Several ongoing studies will provide further insight into the impact of SGLT2I in chronic heart failure patients.

Conclusions

We propose that excess nutrients and cellular fuel supply is harmful to the heart. Similar to the beneficial effect of hemodynamic and neurohormonal unloading, we propose that metabolic unloading is a sound principle for the prevention and treatment of heart failure in obesity and type 2 diabetes mellitus. Strategies of metabolic unloading of the heart include non-pharmacological and pharmacological interventions aimed at correcting the disrupted fuel hemostasis at its source.

Clinical Significance.

• In patients with obesity and diabetes, excess nutrient supply to the heart activates complex molecular mechanisms that lead to cellular metabolic distress and inflammation.

• Metabolic unloading of the heart, by means of restricting cellular fuel supply with nonpharmacologic and pharmacologic interventions, restores homeostasis and improves cardiac function.

Abbreviation list

AGEs

advanced glycation end products

AMPK

AMP-activated protein kinase

BMI

body mass index

FACOA

fatty acyl coenzyme A

FFA

free fatty acids

GIP

gastric inhibitory peptide

GLP

glucagon-like peptide

GLUT

glucose transporter protein

HOMA-IR

Homeostatic model assessment-insulin resistance

IRS

insulin receptor substrates

LVEF

left ventricular ejection fraction

NADPH

nicotinamide adenine dinucleotide phosphate, reduced

OXM

oxyntomodulin

PEPCK

Phosphoenolpyruvate carboxykinase

PKC

protein kinase C

PPAR

Peroxisome Proliferator-Activated Receptors

ROS

reactive oxygen species

RYGB

Roux-en-Y Gastric Bypass

SERCA

sarcoendoplasmic reticulum (SR) calcium transport ATPase

SGLT2I

sodium/glucose cotransporter-2 inhibitors

TZD

thiazolidinediones