Updated evidence on heart and kidney effects of GLP-1 drugs alone and combined with SGLT2 inhibitors and finerenone

Diabetes is one of the most important risk factors for cardiovascular disease (CVD) and chronic kidney disease (CKD). Compared with patients without diabetes, patients with diabetes have a 2–fourfold higher risk of atherosclerotic cardiovascular diseases (ASCVD), including coronary artery disease, stroke, and peripheral artery disease [1]. Moreover, fasting hyperglycaemia was responsible for 2.3 million cardiovascular deaths worldwide in 2021 [2]. Additionally, as the first CVD to develop in patients with diabetes is often heart failure rather than myocardial infarction [1], it is becoming clear that the pathology of diabetes itself causes heart failure through direct effects on cardiomyocytes and other myocardial cell types. CKD is a principal microvascular complication of diabetes and increases the risk of CVD and death [3]. About 40% of patients with diabetes could develop CKD [4], and it could be already present at diagnosis especially in patients with type 2 diabetes [5]. Furthermore, CKD in patients with diabetes is globally the leading cause of end-stage kidney disease and leading to increased health care cost due to renal replacement therapy [4].

Whether hyperglycaemia itself directly affects the pathogenesis of atherosclerosis and heart failure remains inconclusive; this is consistent with the fact that previous large clinical trials have not reached a conclusion on whether cardiovascular disease can be suppressed by strict glycaemic control [6]. Effect of glycaemic control itself on preventing CKD is relatively small, compared with multifactorial therapy in patients with diabetes [7, 8]. Lowering blood glucose levels is the basis of diabetes treatment, not only in terms of preventing microvascular complications, including CKD, but also in preventing cardiovascular events. In contrast, it is difficult to completely prevent the development of cardiovascular events and CKD by hypoglycaemic therapy alone. Thus, in addition to a comprehensive approach encompassing multidisciplinary risk factors (dyslipidaemia, hypertension, obesity, smoking, etc.) [8, 9], considering "what drugs to use to lower blood glucose" by selecting hypoglycaemic agents with evidence-based cardiorenal protective effects, according to patient background and risk status is increasingly important [10].

Glucagon-like peptide-1 receptor agonist (GLP-1RA) and sodium-glucose cotransporter 2 inhibitor (SGLT2i) are currently recommended as first-line hypoglycaemic agents for patients with diabetes with established cardiovascular disease or multiple cardiovascular and renal risks [10]. Particularly, GLP-1RAs have greater hypoglycaemic and weight-loss effects and are attracting attention as a means of intervening in obesity, which is the key basis of type 2 diabetes pathogenesis. The cardiorenal protective effects of finerenone, a novel nonsteroidal mineralocorticoid receptor antagonist (MRA), in patients with diabetes and CKD have also been revealed in recent years [11]. The strategies for preventing cardiovascular and renal events in patients with diabetes are becoming increasingly diverse. However, the results of clinical trials suggest that the characteristics of the cardiorenal protective effects achieved by these drugs differ slightly, and consensus for their use and evidence for their optimal use and combined benefits are still lacking. In this article, we primarily summarize the clinical effects of GLP-1RAs on the risk of cardiovascular and renal events and discuss the emerging role of their combined therapy with SGLT2is and finerenone.

Vascular damages of ASCVD in patients with diabetes are characterized by endothelial and vascular smooth muscle cell dysfunction [12]. Increased oxidative stress, inflammation and lipotoxicity associated with hyperglycaemia and insulin resistance potentially impair vascular endothelial function and increase endothelial permeability, leading to the penetration of glycated and oxidized low-density lipoprotein (LDL) cholesterol into the endothelium, i.e., initiation of atherosclerosis. Subsequently, macrophages transform into foam cells in the LDL cholesterol-deposited vascular endothelium. Hyperglycaemia has been suggested to cause increased expression of inflammation-related genes in macrophages. In vascular smooth muscle cells, hyperinsulinemia also promotes cell proliferation, contributing to the development of atherosclerosis and vascular remodelling. Independent of ASCVD and other risk factors, such as hypertension, diabetes also increases the risk of heart failure [13]. Myocardial remodelling characterized by fibrosis and hypertrophy, impaired microcirculation, abnormal protein function caused by advanced glycation end products, changes in myocardial metabolism and abnormal mitochondrial function have been noted as associated factors of diabetic cardiomyopathy [14].

The process of CKD progression in diabetes is also complicated [15]. Glomerular hyperfiltration due to hyperglycemia is one of the important features of CKD in diabetes and lead to diabetic glomerulopathy in combination with advanced glycation end products and inflammation. Inflammation and fibrosis can affect tubulo-interstitial lesions, and both glomerular and interstitial injuries will development of CKD. Comorbidities of diabetes (i.e., hypertension, dyslipidemia, obesity) also have crucial roles in pathophysiology of CKD in diabetes.

The results of SELECT, which examined the effect of subcutaneous semaglutide injection for the prevention of cardiovascular events in 17,604 patients with obesity (body mass index [BMI] ≥ 27 kg/m²) and established cardiovascular disease but without diabetes, were presented in 2023 [25]. During the mean observation period of 3.3 years, semaglutide at 2.4 mg per week resulted in a mean weight loss of − 9.39% from baseline and a 20% reduction in the risk of MACE (HR 0.80, 95% CI 0.72–0.90), especially a 28% reduction in the risk of myocardial infarction (HR 0.72, 95% CI 0.61–0.85), compared with that in the placebo group. GLP-1RAs not only have glucose-lowering but also greater weight-loss effects as discussed in detail in a previous literature [26], the results of SELECT also suggest the weight-loss effect as an important factor explaining the cardiovascular protective effects of GLP-1RAs regardless of diabetes. However, in Harmony Outcomes, albiglutide reduced the risk of MACE by 22%, with 0.52% and 0.83 kg reductions in HbA1c and weight, respectively, compared with those in the placebo group [18]. Factors independent of glycaemic and weight loss effects have also been suggested to affect the cardiovascular protective effects of GLP-1RAs.

Although a meta-analysis of large-scale clinical trials suggested that GLP-1RAs reduce heart failure hospitalizations (HHF) by approximately 10% in patients with diabetes at high cardiovascular risk [22, 23], the treatment effects on the risk of HHF in individual trials other than AMPLITUDE-O were not significant (Tables 1, 2). Thus, unlike for MACE, the effect of GLP-1RAs in reducing the risk of HHF has not yet been consistently demonstrated in individual trials. FIGHT and LIVE were clinical trials involving the administration of liraglutide in patients with heart failure with reduced ejection fraction (HFrEF). Three hundred patients with HFrEF hospitalized for acute heart failure within 14 days and taking diuretics equivalent to at least 40 mg furosemide were randomized to receive 1.8 mg per day of liraglutide or placebo in FIGHT [27]. Their median left ventricular ejection fraction (LVEF) was 25%, 82% of the patients had ischaemic heart disease, and 59% had type 2 diabetes. At 180 days follow-up, death (HR 1.10, 95% CI 0.57–2.14) and heart failure rehospitalization (HR 1.30, 95% CI 0.89–1.88) did not differ compared with those in the placebo group. In addition, the results were similar for subgroup analysis by diabetes status. In LIVE, 241 clinically stable patients with HFrEF were randomized to receive liraglutide at 1.8 mg per day or placebo [28]. The mean LVEF was 33.7% and 35.4% in the liraglutide and placebo groups, respectively, and 60% of patients had ischaemic heart disease. During the 24-week observation, liraglutide did not improve LVEF but rather caused more death, ventricular arrhythmia, and atrial fibrillation, compared with the placebo. Thus, no evidence currently indicates that patients with HFrEF benefit from the administration of GLP-1RAs, but rather, the increased arrhythmias and death are concerning. One factor that may have contributed to such a result in patients with HFrEF is the increased heart rate caused by GLP-1RAs [21].

On the other hand, it is becoming clear that obesity and diabetes themselves affect the pathogenesis, progression, and prognosis of heart failure with preserved ejection fraction (HFpEF). The effect of semaglutide subcutaneous injection at 2.4 mg per week in patients with HFpEF and obesity (BMI ≥ 30 kg/m²) but without diabetes was tested in STEP-HFpEF [29]. The change in body weight from baseline was − 13.3% and − 2.6% in the semaglutide and placebo groups, respectively. The semaglutide group showed a significantly improved Kansas City Cardiomyopathy Questionnaire clinical summary score and a longer 6-min walking distance from baseline (21.5 m and 1.2 m in the semaglutide and placebo groups, respectively, intergroup difference 20.3 m, 95% CI 8.6–32.1 m), compared with that in the placebo group despite the relatively short observation period of 52 weeks. Similar results were reported in STEP-HFpEF DM, which included patients with diabetes excluded in STEP-HFpEF [30]. Pooled analysis including these two studies in addition to SELECT and FLOW [31] reported that semaglutide reduced the risk of hospitalization or urgent visits due to heart failure by 41% (HR 0.59, 95% CI 0.41–0.82) in HFpEF patients compared with placebo [32]. In contrast, the effect on cardiovascular death did not reach at the statistical significance (HR 0.82, 95% CI 0.57–1.16), possibly due to lower event rate. Nearly all patients were overweight or obesity (BMI ≥ 27 kg/m²) in this analysis, and reduction of worsening heart failure event was more compelling in patients with higher BMI (HR 0.49, 95% CI 0.33–0.70 in BMI ≥ 35 kg/m², HR 0.96 95% CI 0.67–1.38 in BMI < 35 kg/m²). Weight loss effect of semaglutide could be one of important factors leading to improved quality of life in patients with HFpEF and overweight. However, the effect beyond weight loss was also indicated by decreasing trend of NT-proBNP or C-reactive protein levels independent of weight loss in STEP-HFpEF [33, 34].

Collectively, semaglutide improved the heart failure-related symptoms and reduce the incidence of worsening heart failure events especially in HFpEF patients with higher BMI, suggesting that it could be a promising therapeutic approach for this patient population. However, whether this could be a class effect of GLP-1RA remains uncertain, and the evidence on HFrEF is scarce.

In individual clinical trials of GLP-1RAs, the therapy improved the kidney outcomes compared with those of placebo treatment, although most of these results emanated from secondary outcomes or exploratory analysis. Zelniker et al. [35] defined kidney outcomes as broad (new onset of macroalbuminuria, worsening estimated glomerular filtration rate [eGFR], end-stage kidney disease, or death attributable to renal causes) and narrow composite (excluding macroalbuminuria from a broad composite) and performed a meta-analysis of ELIXA, LEADER, SUSTAIN-6, and EXSCEL [36]. Broad composite outcomes were significantly reduced compared with those following placebo treatment (HR 0.82, 95% CI 0.75–0.89), but no statistically significant reduction in narrow composite outcomes, except for macroalbuminuria, was observed (HR 0.92, 95% CI 0.80–1.06). In a meta-analysis by Sattar et al., composite kidney outcomes, including macroalbuminuria, were reduced by 21% compared with those in the placebo group (HR 0.79, 95% CI 0.73–0.87). Although the incidence of worsening kidney function outcome based predominantly on eGFR change was nominally reduced after removal of ELIXA (0.82, HR 95% CI 0.69–0.98), composite outcomes excluding macroalbuminuria, mainly declining eGFR and doubling creatinine, were not significant (HR 0.86, 95% CI 0.72–1.02) [23]. Therefore, most of renoprotective effects of GLP-1RAs were likely to be due to the inhibition of progression to macroalbuminuria [35], and the effects on other clinically important renal endpoints, including eGFR decline, initiation of renal replacement therapy and renal death, was unclear [5].

The result of FLOW, a study to assess the effect of subcutaneous semaglutide injection on CKD in patients with type2 diabetes, was reported recently [31]. In this study, 3533 patients with CKD (eGFR of 50–75 ml/min/1.73 m² and a UACR of > 300 and < 5000 mg/gCr or an eGFR of 25 to < 50 ml/min/1.73 m² and a UACR of > 100 and < 5000 mg/gCr) and diabetes were enrolled and assigned to 1 mg per week subcutaneous semaglutide injection or a placebo group. The primary outcome defined as time to first kidney failure (persistent eGFR < 15 mL/min/1.73 m² or initiation of chronic renal replacement therapy), ≥ 50% eGFR decline from baseline, or death from renal or cardiovascular causes was lower in semaglutide group (HR 0.76, 95% CI 0.66–0.88). In the component of primary outcome, persistent ≥ 50% eGFR decline from baseline was significantly reduced in semaglutide group (HR 0.73, 95% CI 0.59–0.89). Initiation of renal replacement therapy tended to decrease, but did not reach to statistical significance (HR 0.84, 95% CI 0.63–1.12). Thus, this study firstly added an insight that semaglutide can favourably affect more clinically important renal endpoints in patients with type 2 diabetes and CKD.

GLP-1 receptors are thought to be expressed on atrial and ventricular cardiomyocytes, endothelial cells in the human heart, and endothelial and smooth muscle cells in blood vessels [21]. However, humans and mice display large differences. Furthermore, information on GLP-1 receptor expression in the heart is still lacking, and whether GLP-1RAs act directly or indirectly on the cardiovascular system remains inconclusive. Nevertheless, in vitro studies on blood vessels have shown that GLP-1RAs inhibit vascular smooth muscle cell proliferation, reduce reactive oxygen species, and increase nitric oxide in vascular endothelial cells [37].

In animal models of heart failure induced by high-frequency pacing, ischaemia, and obesity, GLP-1RAs inhibit left ventricular remodelling and improve left ventricular contractility and diastolic function through anti-inflammatory effects, cardiomyocyte apoptosis inhibition, and glucose metabolism enhancement in cardiomyocytes [38 –40]. However, in a LIVE sub-study, myocardial glucose utilization did not improve in the liraglutide group as examined by F-fluorodeoxyglucose positron emission tomography [41], similar to a study using albiglutide [42]. This discrepancy between results in animal models and clinical studies may partly be due to myocardial insulin resistance in severely failing hearts and the use of beta-blockers [21].

Also in kidney, localization of GLP-1 receptor is under investigation, but it is thought to be expressed in vascular smooth muscle cells and immune cells in kidney [43]. Although GLP-1RAs have inhibitory effect on sodium hydrogen exchanger-3 in tubules, it is unclear that renoprotective effect of GLP-1RAs is attributable to hemodynamic effect [44]. Preclinical studies suggested that GLP-1RAs reduced immune cell activity and oxidative stress in kidney and led to decrease in fibrosis [15, 45]. Immunomodulation effects are currently thought to have principal roles in renoprotective effect of GLP-1RAs.

In a meta-analysis comparing the effects of GLP-1RAs with SGLT2is, five GLP-1RA (ELIXA, LEADER, SUSTAIN-6, EXSCEL, and Harmony Outcomes) and three SGLT2i trials (EMPA-REG OUTCOME, CANVAS Program, and DECLARE-TIMI 58) with a total of 77,242 patients were include [35]. The risk of MACE was reduced by 12% (HR 0.88, 95% CI 0.84–0.94) with GLP-1RAs and by 11% (HR 0.89, 95% CI 0.83–0.96) with SGLT2is compared with that in placebo groups, suggesting that GLP-1RAs and SGLT2is reduce the risk of MACE to a similar extent. Furthermore, the risks of myocardial infarction (HR 0.91, 95% CI 0.84–0.98 with GLP-1RAs and HR 0.89, 95% CI 0.80–0.98 with SGLT2is) and cardiovascular death (HR 0.88, 95% CI 0.80–0.96 with GLP-1RAs and HR 0.84, 95% CI 0.75–0.94 with SGLT2is) were similarly reduced in both GLP-1RA and SGLT2i groups, while the risk of stroke was reduced only with GLP-1RAs (HR 0.86, 95% CI 0.77–0.97 with GLP-1RAs and HR 0.97, 95% CI 0.86–1.10 with SGLT2is). On the other hand, the risks of HHF (HR 0.93, 95% CI 0.83–1.04 with GLP-1RAs and HR 0.69, 95% CI 0.61–0.79 with SGLT2is) and composite kidney outcomes excluding macroalbuminuria (HR 0.92, 95% CI 0.80–1.06 with GLP-1RAs and HR 0.55, 95% CI 0.48–0.64 with SGLT2is) significantly reduced with SGLT2is only. Particularly for heart failure, several clinical trials and meta-analyses have already confirmed that SGLT2is reduce HHF by approximately 30% in both HFrEF and HFpEF, with or without diabetes [46]. Thus, the clinical benefits for the risk of heart failure and kidney outcomes, except albuminuria, are likely to be a strength of SGLT2is, whereas those for the risk of stroke are a strength of GLP-1RAs (Graphical Abstract).

Recently, in FIDELIO-DKD [52], FIGARO-DKD [53] and their pooled meta-analysis FIDELITY [11], finerenone has also been shown to reduce cardiovascular and renal events in CKD patients with diabetes. The cardiovascular benefit of finerenone in CKD patients with diabetes was primarily attributable to the reduction in the risk of HHF and was similar in FINEARTS-HF for patients with HFpEF [54]. In the FINE-HEART pooled analysis of these three large clinical trials, finerenone also reduced the risks of HHF, composite kidney outcome without macroalbuminuria, MACE, and all cause death [55]. Among participants, background use of GLP-1RA and SGLT2i was 5.8% and 8.9%, respectively, and treatment effect of finerenone on cardiovascular death was consistent regardless of those medication usages. The clinical benefits of finerenone on cardiovascular and kidney events are also unlikely affected by combination therapy with GLP-1RA or SGLT2i, however, clinical evidence on additive effects of finerenone in combination with GLP-1RA or SGLT2i are still limited [56, 57]. The CONFIDENCE is currently examining the effect of empagliflozin and finerenone combination therapy on kidney outcomes in CKD patients with diabetes [58].

Neuen et al. performed a cross-trial analysis using data from clinical trials of GLP-1RAs (ELIXA, LEADER, SUSTAIN-6, EXSCEL, HARMONY, REWIND, PIONEER-6 [59], and AMPLITUDE-O), SGLT2is (CANVAS and CREDENCE), and finerenone (FIDELIO-DKD and FIGARO-DKD) to estimate the effects of the triple combination therapy of GLP-1RA, SGLT2i, and finerenone using the actuarial method [60]. In patients with diabetes and albuminuria, the HRs for MACE, HHF, and CKD progression with triple combination therapy were estimated to be 0.65 (95% CI 0.55–0.76), 0.45 (95% CI 0.34–0.58), and 0.42 (95% CI 0.31–0.56), respectively, compared with those for conventional therapy alone (renin-angiotensin system inhibitors and classical risk factor management). The mechanisms of action of GLP-1RAs, SGLT2is, and nonsteroidal MRA are generally considered complementary and independent but may overlap in part, and the effects of each drug may not be completely additive when the three drugs are combined. Therefore, when estimated assuming 50% additivity, the HRs were 0.73 (95% CI 0.62–0.87) for MACE, 0.53 (95% CI, 0.41–0.70) for HHF, and 0.51 (95% CI, 0.38–0.68) for CKD progression. Some attenuation in HR was observed, but the impact of triple combination therapy on MACE, HHF, and CKD progression was still consistent, assuming 50% additivity. When patients aged 50 years old start triple therapy, event-free survival was estimated to increase by 3.2 years (95% CI 2.1–4.3 years) for MACE, 3.2 years (95% CI 2.4–4.0 years) for HHF, and 5.5 years (95% CI 4.0–6.7 years) for CKD progression compared with that for conventional therapy alone. Results of this integrated analysis suggest that the triple combination of GLP-1RA, SGLT2i, and finerenone provides stronger cardiorenal protective benefits than either conventional therapy or each alone. Given existing evidence, triple combination therapy can be expected to provide additional cardiorenal protection, although further randomized controlled trials are needed to determine whether triple combination therapy actually has additive or synergetic effects compared with mono or dual therapy.

Thus, we have currently obtained multiple therapeutic tools that reduce the risks of cardiovascular and renal events in patients with diabetes. Moreover, their combined use has the potential to further enhance treatment effects. Appropriate drug choice according to the patient clinical situation will lead to the precision medicine in the field of cardiovascular-kidney-metabolic syndrome.

To further implement GLP-1RA-centric combination therapy, we need to consider some evidence gaps (Table 4). First, we should identify populations that are suitable or unsuitable for GLP-1RA-centric combination therapy. In recent GLP-1RA trials focusing on patients with obesity, favourable effect on worsening heart failure in HFpEF patients was indicated to be modulated by baseline BMI [32], although the underlying mechanisms that explain the cardiovascular and renal benefits may not necessarily be due to weight loss alone. Since background BMI can be largely different between races, caution is needed when interpreting and applying the results, including combination use with SGLT2is and finerenone. In addition, few studies currently support the use of GLP-1RAs in patients with HFrEF. Second, determining the ideal timing and priority to intensify therapy is also urgently required, since present evidence are not enough to cover this evidence gap. In particular, treatment effect size may differ among each combination and be affected by patient clinical situations and targeting outcomes. Third, long-term adherence and safety of the combination therapy need to be further addressed. Finally, cost-effectiveness is an important issue that is often addressed in multidrug regimen research. Given the tremendous costs of treating cardiovascular disease and continuing renal replacement therapy, the suspected total benefits obtained from therapy should not be overlooked. Therefore, continuous assessment is also required to determine the cost-effectiveness of the combination therapy.