Introduction

Heart failure (HF), chronic kidney disease (CKD), and type 2 diabetes mellitus (T2DM) are common and interrelated conditions, each conferring increased disease burden both individually and in combination [1]. Cardiovascular disease (CVD), particularly HF, and CKD each progress through a pathway of pathophysiologic steps, for which T2DM is a major risk factor [2]. Sodium-glucose cotransporter-2 inhibitors (SGLT2is) are used to treat hyperglycemia in patients with T2DM. Data from large, phase 3, randomized controlled trials (RCTs) have shown that SGLT2i therapy improved cardiovascular (CV) and kidney outcomes in patients with T2DM [3], and observed reductions in the risk of hospitalization for HF led to this drug class being evaluated in patients with HF, with or without T2DM. SGLT2is have been shown to reduce the development and progression of HF with reduced ejection fraction (HFrEF) [4, 5]; however, until recently, the effect of these agents in patients with HF with preserved ejection fraction (HFpEF) was not well studied. The aim of this review is to summarize key information regarding the interaction between HFpEF, CKD, and T2DM and discuss the role of SGLT2 inhibition in the management of patients with comorbid HFpEF and CKD, with or without T2DM.

Literature was retrieved from PubMed and Google Scholar databases using Boolean searches for terms related to SGLT2is, HFpEF, T2DM, and CKD (limits: English-language articles, humans). The reference lists from retrieved articles were also considered. Other relevant literature was obtained on the basis of the personal knowledge and experience of the authors. Additional data were obtained from the US National Institutes of Health website ClinicalTrials.gov and from websites pertaining to individual therapeutic agents of interest. The retrieved references were manually assessed by one reviewer and formed the basis for this narrative review.

Epidemiology of HF, CKD, and T2DM

HF epidemiology

HF is defined as a clinical syndrome with symptoms and/or signs caused by a structural and/or functional cardiac abnormality and corroborated by elevated natriuretic peptide levels and/or objective evidence of pulmonary or systemic congestion [6]. Left ventricular ejection fraction (LVEF) provides prognostic information for patients with HF and defines differing treatment groups [6,7,8]: HF with reduced EF (HFrEF; LVEF ≤ 40%), previously called “systolic HF”; HF with mildly reduced EF (HFmrEF; LVEF 41–49%); HF with preserved EF (HFpEF; LVEF ≥ 50%), previously called “diastolic HF”; and HF with improved EF (baseline LVEF ≤ 40%, followed by a ≥ 10-point increase from baseline and a second LVEF measurement > 40%) [6]. HF affects approximately 64 million adults globally (per 2017 data) [9]. This represents an almost doubling in HF cases over a 27-year period (1990–2017), with almost half of all cases coming from China and India [9]. In the United States (US), HF prevalence was 2.4% in 2012 and affected 5.7 million adults (aged ≥ 20 years) and is expected to rise to 3.0% by 2030, when it will affect > 8 million adults (aged ≥ 20 years) [10]. In the community setting, up to half of patients with HF have HFpEF [11,12,13], although this rate depends on diagnostic accuracy and an evolving clinical definition [12]. Factors contributing to the increased prevalence of HFpEF include an aging population and increased HFpEF-related risk factors (such as diabetes, hypertension, and obesity), as well as improved diagnosis and survival [12]. Trial-based analyses, mainly involving patients with chronic HF in ambulatory settings (i.e., outpatient care), report 1-year mortality rates for HFpEF of around 5%, whereas observational studies report rates of up to 30% using data primarily from inpatients with decompensated HFpEF [12]. The 5-year mortality rate in a community study of adults with HFpEF (aged > 45 years) was 10% for those with a mild degree of diastolic dysfunction, rising to 23% in moderate/severe disease [14]. A large meta-analysis of ambulatory patients with HF found that pooled survival rates were similar for HFpEF and HFrEF at 1 year (89% and 88%, respectively) and 5 years (70% and 63%, respectively) [15].

CKD epidemiology

CKD is defined as persistent albuminuria (albumin-creatinine ratio [ACR], ≥ 30 mg/g [≥ 3 mg/mmol]), persistently reduced estimated glomerular filtration rate (eGFR; < 60 mL/min per 1.73 m2), or both [16]. The Global Burden of Disease Study (2017 data) reported 698 million cases of all-stage CKD, giving a global prevalence of 9.1% [17], whereas an earlier systematic review and meta-analysis (100 studies; ~ 6,900,000 patients) reported a global prevalence of 13.4% [18]. In the US, the prevalence of CKD was 15% in 2021, equating to approximately 37 million American adults with CKD [19]. Changes in CKD prevalence over time have showed stabilization or even improvement (reviewed in [20]). The reasons for this are unclear, given the observed increases in CKD risk factors (such as T2DM and obesity), although hypertension prevalence has stabilized or decreased in many high- and middle-income countries due to improved detection and treatment [21]. The primary causes of CKD are T2DM (30–50% of cases), hypertension (~ 27%), and primary glomerulonephritis (~ 8%) [22]. The prognosis of CKD worsens with increasing Kidney Disease: Improving Global Outcomes (KDIGO) CKD category (based on GFR and albuminuria categories) [23, 24], but only a small proportion of individuals have severely decreased GFR (stage G4), kidney failure (G5), or severely increased albuminuria (A3) [23, 24]. In addition to the complications associated with CKD (such as anemia, mineral bone disease, end-stage kidney disease [ESKD], etc.), CKD is an important risk factor for CV morbidity and mortality, including coronary artery disease, HF, arrhythmias, and sudden CV death (reviewed in [25]). For individuals with CKD, the risk of develo** CVD is greater than that of develo** ESKD [25, 26].

T2DM epidemiology

T2DM accounts for up to 95% of all cases of diabetes, and is caused by a progressive loss of insulin secretion from pancreatic β-cells that becomes insufficient to compensate for insulin resistance, resulting in hyperglycemia [27]. T2DM can be diagnosed using various tests (e.g., fasting plasma glucose, glycated hemoglobin, etc.) and in a variety of clinical settings (e.g., incidental finding, asymptomatic/symptomatic screening, etc.) [27]. Traditional risk factors for T2DM include overweight and obesity, lack of physical activity, and unhealthy diet. Around 6% of the global population (~ 462 million people) are affected by T2DM, and T2DM accounted for more than 1 million deaths (Global Burden of Disease data from 2017) [28]. In the US, 35.4 million adults have T2DM (2019 data) [29]. In terms of disability-adjusted life years (DALYs), a measure of premature deaths and years lived with disability from a particular disease, T2DM causes the seventh highest burden of disease [28]. Global prevalence of T2DM has increased over the last 30 years and is forecasted to rise (cases per 100,000 people) from 6059 in 2017 to 7079 by 2030 [28]; furthermore, global trends modeled to 2025 show continued increases in T2DM incidence, age-standardized rates, deaths, and DALYs [30]. It is well established that diabetes-induced hyperglycemia is a causative factor in the development of microvascular disease (retinopathy, nephropathy, and neuropathy) and macrovascular disease (peripheral artery disease, coronary artery disease, and stroke) [31,32,33,34], via the activation of pathways that trigger cellular oxidative stress, release of inflammatory mediators, mitochondrial dysfunction, and the development of atherosclerosis (reviewed in [35]). Atherosclerotic CVD (ASCVD) is the leading cause of morbidity and mortality in patients with T2DM, and conditions that commonly coexist with T2DM (such as hypertension and dyslipidemia) are risk factors for ASCVD, as is T2DM itself [36].

The relationship between CKD and HF: cardiorenal syndrome

HF and CKD have a bidirectional relationship [36], and the presence of either condition is associated with a worse prognosis in the other. Patients with CKD have a three-fold increased risk of incident HF than those without CKD [37]. The presence of HF in patients with CKD is associated with increased risk of death, more frequent hospitalizations, and reduced health-related quality of life [37,38,39]. The coexistence of HF (with a reduced or preserved ejection fraction [EF]) and CKD can manifest as cardiorenal syndrome (CRS), although comparatively less is known about CRS in HFpEF than in HFrEF [84] and presented in Fig. 3A and B) [84]. Some of these groups may be expected to be more responsive to SGLT2i treatment than others [81], such as patient phenotypes related to obesity and metabolism/inflammation [85, 86]. In addition, other less common cardiac disorders may present with a HFpEF phenotype. These are broadly divided into conditions affecting the myocardium (such as inherited or acquired infiltrative, restrictive, inflammatory, or genetic cardiomyopathies) and those altering cardiac loading conditions (such as hypertension, congenital or acquired valvular and structural defects, rhythm abnormalities, etc.) [87]. Some of these presentations may not be responsive to SGLT2i treatment, particularly those caused by infiltrative diseases, such as cardiac amyloidosis [88]. Cardiac amyloidosis causes restrictive cardiomyopathy, of which the major clinical presentation is HFpEF, with symptoms caused by raised LV filling pressure secondary to increased stiffness and reduced elasticity of the heart tissue [89]. Kidney involvement may occur as part of the primary condition (e.g., systemic amyloidosis is a cause of type 5 CRS) or may be secondary to the ensuing cardiac disease. The importance of recognizing patients with previously undiagnosed cardiac amyloidosis in clinical trials for HFpEF was described recently, and this condition may contribute further to the heterogeneity of HFpEF populations and the failure to obtain positive results in some studies [90].

Fig. 3
figure 3

Reproduced from Heart 2022, volume 108, pages 1342–1350, Gevaert AB, Kataria R, Zannad F, Sauer AJ, Damman K, Sharma K, et al. Heart failure with preserved ejection fraction: recent concepts in diagnosis, mechanisms and management, Copyright 2022, with permission from BMJ Publishing Group Ltd

Evolution of pathophysiological understanding of HFpEF [84]. A Prevailing concept of HFpEF and HFrEF as separate diseases, HFpEF is caused by microvascular inflammation and HFrEF is caused by cardiomyocyte loss. B Emerging concept of heart failure as phenotypes overlap** across the spectrum of LV systolic function. There is a gradual change in underlying pathophysiology, mode of death, and response to HF therapies across the LVEF spectrum, with influences from genetics, sex, comorbidities, and lifestyle. C Personalized treatment of HFpEF. Different phenotypes (based on clinical, imaging, biomarker and/or transcriptomic data) are represented by red, green and blue colors. Personalized treatment: considering the phenotype-specific response to medical therapy, a targeted approach using specific drugs in specific phenotypes could lead to net clinical benefit for all patients. CV cardiovascular, GLS global longitudinal strain, HF heart failure, HFpEF heart failure with preserved ejection fraction, HFrEF heart failure with reduced ejection fraction, LV left ventricle, LVEF left ventricular ejection fraction, NO nitric oxide, ROS reactive oxygen species, SV stroke volume.

Full size image

Factors related to the design of EMPEROR-Preserved may also be relevant to the observed effect of LVEF. For example, the inclusion of patients with LVEF > 40%, which encompasses HFmrEF (LVEF 41–49%), and the use of a variety of imaging techniques to measure LVEF up to 6 months before study entry, which increases variability and may underestimate potential changes over time [81]. However, the DELIVER trial design may refute this, as it also enrolled patients with LVEF > 40% and documented the EF over a longer period prior to trial enrollment (≤ 12 months), with assessment only via echocardiography or cardiac magnetic resonance imaging [58].

Secondary kidney outcomes in HFpEF vs. HFrEF

A further question is why the secondary kidney outcomes with HFpEF look different when compared with HFrEF. In a comparison of EMPEROR-Reduced and DAPA-HF [91], in which patients with HFrEF were enrolled, both trials reported similar and significant effects of SGLT2is in reducing the decline in the eGFR slope (mean slope of eGFR change vs. placebo: 1.73 and 1.78 mL/min/1.73 m2 per year for empagliflozin and dapagliflozin, respectively) [91]. However, the composite kidney outcome showed a statistical decrease in EMPEROR-Reduced (0.50; 95% CI 0.32–0.77) but not in DAPA-HF (0.71, 95% CI 0.44–1.16), possibly due to fewer kidney events in the latter because of a higher eGFR entry criterion (≥ 20 vs. ≥ 30 mL/min/1.73 m2, respectively) and differences in the eGFR decline defined in the composite kidney outcome (≥ 40% for EMPEROR-Reduced vs. ≥ 50% decline in DAPA-HF) [91]. A prespecified analysis of EMPEROR-Reduced, in which patients were categorized by the presence or absence of CKD at baseline (defined as eGFR < 60 mL/min/1.73 m2 or UACR > 300 mg/g), investigated the direct impact on kidney events via a prespecified composite kidney outcome (defined as a sustained profound decline in eGFR, chronic dialysis, or transplant) [92]. Empagliflozin reduced the slope of eGFR decline in patients with CKD (1.11 [95% CI 0.23–1.98] mL/min/1.73 m2 per year) and without CKD (2.41 [95% CI 1.49–3.32] mL/min/1.73 m2 per year), and the risk of the composite kidney outcome was similarly reduced in patients with and without CKD (HR: 0.53, 95% CI 0.31–0.91 vs. HR: 0.46, 95% CI 0.22–0.99, respectively) [92].

Despite these analyses, the reason why the kidney outcomes appear to be less impressive in patients with HFpEF largely remains unknown. The comparison of EMPEROR-Reduced and DAPA-HF presents relevant points concerning the level of eGFR at trial entry, number of kidney events, and differences in definitions, but it is equally possible that the difference in kidney outcomes is simply due to chance. Furthermore, it may be erroneous to postulate that HFrEF and HFpEF consistently behave differently with respect to kidney outcomes. Data from a post hoc analysis of renal outcomes in the Prospective Comparison of angiotensin receptor-neprilysin inhibitor (ARNI) with angiotensin receptor blocker (ARB) Global Outcomes in HF with Preserved Ejection Fraction trial (PARAGON-HF; NCT01920711) demonstrated a significant reduction in the prespecified kidney composite outcome (time to first occurrence of either ≥ 50% reduction in eGFR, ESKD, or death from renal causes) [93], even though the primary outcome (composite of total HF hospitalizations and CV death) was not achieved [43]. Conversely, the United Kingdom Heart and Renal Protection-III trial (UK HARP-III; ISRCTN11958993) showed no benefits in the kidney with ARNI use in patients with CKD only [94], although the results may have been affected by the trial design (patient characteristics, heterogeneous CKD etiologies, small study size, short follow-up duration, etc. [93]). These data highlight the need for obtaining a more accurate phenotype for patients with HF and CKD (carried out using methods other than cut-offs for eGFR and LVEF) to better determine which patients will respond to different guideline-directed medical therapies (as depicted in Fig. 3B).

Emerging therapies for HFpEF and CKD: finerenone

Although SGLT2is are undoubtedly valuable in the management of HF and CKD, other drugs are also important. Finerenone is one of the standards of care in patients with DKD [95, 96]. It is a selective, non-steroidal mineralocorticoid receptor (MR) antagonist (MRA) that blocks MR-mediated sodium reabsorption and MR overactivation and has demonstrated anti-inflammatory and anti-fibrotic effects in preclinical models of kidney and CV disease [97]. The complementary phase 3 RCTs, Finerenone in Reducing Kidney Failure and Disease Progression in Diabetic Kidney Disease (FIDELIO-DKD; NCT02540993 [98]) and Finerenone in Reducing Cardiovascular Mortality and Morbidity in Diabetic Kidney Disease (FIGARO-DKD; NCT02545049 [99]) comprise the largest cardiorenal outcomes program in CKD in T2DM to date [100]. Patients with chronic symptomatic HFrEF were excluded from the FIDELIO and FIGARO trials. FIDELITY was a prespecified pooled analysis of efficacy and safety data from FIDELIO and FIGARO and allowed for evaluation across the range of CKD severity [96] (N = 13,026; broad spectrum of CKD and T2DM; all patients were treated with an optimized dose of angiotensin-converting enzyme inhibitor or ARB) [100]. Approximately 8% of all trial participants were noted to have HF at baseline. FIDELITY provided evidence of both CV and renal protection with finerenone compared with placebo. The analysis showed a 14% risk reduction in the composite CV outcome (consisting of CV death, non-fatal myocardial infarction [MI], non-fatal stroke, or hospitalization for HF) for finerenone vs. placebo (12.7% vs. 14.4%, respectively; HR: 0.86 [95% CI 0.78–0.95]; p = 0.0018) and 23% reduction in risk of the composite kidney outcome (consisting of sustained ≥ 57% decrease in eGFR from baseline over ≥ 4 weeks or renal death) for finerenone vs. placebo (5.5% vs. 7.1%; HR: 0.77 [95% CI 0.67–0.88]; p = 0.0002) [96, 100]. Hospitalization for HF was the primary contributor to the CV benefit observed in the FIDELITY analysis, with a 22% risk reduction (HR: 0.78, 95% CI 0.66–0.92; p = 0.0030) [100]. Per the US Food and Drug Administration, finerenone is now indicated to reduce the risk of sustained eGFR decline, ESKD, CV death, non-fatal MI, and hospitalization for HF in adults with CKD associated with T2DM [101].

Recommendations for management of HFpEF

Joint guidelines from the American Heart Association/American College of Cardiology/Heart Failure Society of America (AHA/ACC/HFSA 2022) now include the use of SGLT2is for patients with HFpEF (Class of Recommendation 2a, evidence moderate; benefit >  > risk) [8] due to their benefits in decreasing HF hospitalizations and CV mortality [61] (presented in Fig. 4). However, these recommendations were issued before the results of DELIVER were published, and will likely be updated when the new data are taken into consideration. MRAs and ARNIs may also be considered for decreasing hospitalizations in selected patients with HFpEF, particularly those at the lower end of the LVEF spectrum, per the AHA/ACC/HFSA guidelines (Class of Recommendation 2b, evidence weak; benefit ≥ risk) [8]. European Society of Cardiology (ESC) guidelines were published in September 2021 prior to the availability of data from recently completed trials with SGLT2is in HFpEF; thus, no recommendations regarding disease-modifying therapies are provided. However, the use of SGLT2is (dapagliflozin and empagliflozin) is recommended by the ESC in patients with HFrEF to reduce the risk of hospitalization for HF and death [7].

Fig. 4
figure 4

Reproduced from J Am Coll Cardiol. 2022, volume 79(17), pages e263–e421, Heidenreich PA, Bozkurt B, Aguilar D, et al. 2022 AHA/ACC/HFSA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines, Copyright (2022), with permission from The American Heart Association, Inc., The American College of Cardiology Foundation, and The Heart Failure Society of America

Recommendations for patients with preserved LVEF (≥ 50%), per AHA/ACC/HFSA 2022 [8]. Medication recommendations for HFpEF are displayed. ARB angiotensin receptor blocker, ARNI angiotensin receptor-neprilysin inhibitor, HF heart failure, HFpEF heart failure with preserved ejection fraction, LVEF left ventricular ejection fraction, MRA mineralocorticoid receptor antagonist, SGLT2i sodium-glucose cotransporter-2 inhibitor. *Greater benefit in patients with LVEF closer to 50%.

Full size image

Limitations

There are several limitations associated with this work. Only two databases were used in the search strategy. As the retrieved references from the searches were only assessed by one reviewer, there is a possibility of selection bias.

Conclusions

SGLT2is have demonstrated efficacy and safety in treating HFpEF in patients with comorbid CKD, with and without T2DM. The efficacy of SGLT2is appears to be a class effect. Data from some clinical trials have led clinicians to question whether SGLT2is are effective across the spectrum of EFs in HF, and whether there may be a difference in kidney outcomes between patients with HFpEF vs. HFrEF. Further analysis of the individual trial designs and participant characteristics reveal potentially mitigating factors that may explain the relevant sets of ostensibly neutral results and highlights the need to obtain more accurate phenotypes for patients with HF and CKD (using more nuanced methods than cut-off values for eGFR and LVEF) to better determine which patients might respond to different guideline-directed medical therapies. Furthermore, due to their high risk of develo** HFpEF, patients with CKD may benefit from therapies such as SGLT2is, ARNis, ARBs, and MRAs even if they have not yet been diagnosed with HF.