Personalizing kidney-protecting treatment for obese patients with chronic kidney disease

The nephroprotective effects of GLP-1RAs in CKD were initially observed in cardiovascular outcomes trials (CVOTs) of GLP-1RAs in patients with T2DM and high cardiovascular risk (Table 1) [11–17]. The signal for nephroprotection, particularly with respect to reducing new-onset macroalbuminuria, accumulated from these studies, resulted in the design of the FLOW trial (NCT03819153↗), the first clinical trial of a GLP-1RA (semaglutide 1 mg weekly) with a primary kidney endpoint [18]. Additionally, semaglutide 2.4 mg weekly was found to be nephroprotective in patients with overweight/obesity and at high cardiovascular risk, but without T2DM, in the SELECT trial (NCT03574597↗) [19]. Similarly, in the SMART trial (NCT04889183↗), semaglutide 2.4 mg weekly lowered albuminuria by ≈50% from baseline over a 24-week treatment period in patients with overweight/obesity and CKD, but without T2DM [20]. The impact of GLP-1RAs compared with placebo on kidney disease endpoints including albuminuria, eGFR decline, incidence of ESKD and composite kidney/cardiovascular endpoints in these CVOTs and in primary kidney outcome studies is summarized in Table 1.

It is notable from Table 1 that GLP-1RAs primarily reduced incident macroalbuminuria in secondary outcome and post hoc analyses of CVOTs, while the effect on the estimated glomerular filtration rate (eGFR) slope was less consistent. Nevertheless, a trend towards slower eGFR decline in patients with T2DM was observed with dulaglutide in the REWIND trial (NCT01394952↗) and with efpeglenatide in the AMPLITUDE-O trial (NCT03496298↗) [16, 17]. Similarly, a slower rate of annual eGFR decline in patients with overweight/obesity was observed with semaglutide in the SELECT trial [19]. GLP-1RAs also reduced the incidence of composite kidney endpoints in these CVOTs, in large part driven by the reduced incidence of macroalbuminuria. The CVOTs of GLP-1RAs outlined in Table 1 were conducted in populations at high cardiovascular risk but not at high risk of developing clinically relevant kidney disease endpoints. These study populations primarily consisted of patients without CKD or with early stage CKD. As such, power to detect a meaningful impact of GLP-1RAs on the rate of eGFR decline, CKD progression and incidence of ESKD was limited in these studies as the event rates were low.

The FLOW trial randomized patients with T2DM at high risk of CKD progression to semaglutide 1 mg weekly or placebo to determine the impact of semaglutide on clinically relevant kidney disease endpoints [18]. Patients at high risk of CKD progression were defined as having an eGFR of 50–75 ml/min/BSA and a urine albumin:creatinine ratio (UACR) of 300–5000 mg/g or an eGFR of 25–50 ml/min/BSA and a UACR of 100–5000 mg/g. The proportion of patients with an eGFR >60 ml/min/BSA was capped at 20% of the study population. The mean eGFR was 47 ± 15 ml/min/BSA and the median UACR was 568 mg/g at enrolment. More than 95% and 15% of the study population were on a RAASi and SGLT2i at enrolment, respectively. No data on nsMRAs were provided; it is assumed that no patients were taking finerenone at enrolment. No BMI eligibility criterion was imposed, but the mean BMI of the study population was in the World Health Organization class 1 obesity category at 32 ± 6 kg/m². The incidence of the primary kidney outcome (a composite of ≥50% eGFR decline, ESKD or death from kidney or cardiovascular causes) was reduced by 24% for semaglutide relative to placebo. After excluding cardiovascular death from the composite endpoint, the effect estimate was similar, with a 21% reduction observed in the semaglutide arm. Semaglutide also lowered the annual rate of eGFR decline (total slope) by >1 ml/min/BSA/year compared with placebo.

With the promising results of primary kidney outcome studies of GLP-1RAs (the FLOW and SMART trials) and continued secondary outcome analyses of GLP-1RA CVOTs, the role of GLP-1RAs in improving cardiovascular and kidney outcomes in patients with overweight/obesity or T2DM and CKD appears to be expanding rapidly. However, many priority research questions remain. The impact of GLP-1RAs on kidney endpoints in type 1 DM, non-albuminuric type 2 DKD, non-diabetic CKD and in CKD arising in the context of normal body weight remains to be fully defined. The results of the SMART trial are promising with respect to the application of GLP-1RAs to non-diabetic CKD [20]. However, this was a small trial (101 patients randomized to semaglutide or placebo) with a short duration of follow-up (24 weeks) that demonstrated a positive impact of semaglutide on albuminuria. Larger and longer-term studies are needed to test the impact of GLP-1RAs on the GFR slope and other more clinically relevant kidney endpoints in non-diabetic CKD. The impact of GLP-1RAs in underrepresented populations at high risk of CKD progression, such as Black patients, needs to be studied further. The timing and sequence of how to combine GLP-1RAs with other goal-directed therapies for CKD also needs further study, as does how they might synergize with metabolic/obesity surgery to provide durable improvements in cardiovascular and kidney outcomes.

Patients with advanced CKD (eGFR <30 ml/min/BSA) and ESKD have largely been excluded from CVOTs of GLP-1RAs. For example, participants with an eGFR <30 ml/min/BSA constituted <0.5% of the SELECT trial population [21]. In the first clinical trial with a primary kidney outcome (FLOW), participants could be enrolled down to an eGFR of 25 ml/min/BSA and participants with an eGFR <30 ml/min/BSA constituted 11% of the study population [18]. As such, the impact of GLP-1RAs on cardiovascular and kidney outcomes is limited in the setting of advanced CKD and ESKD, and this should be a priority for future studies.

Separate from their end-organ benefits on the heart and kidney, GLP-1RAs may also play an important role in weight management in patients with ESKD. Although higher BMI is associated with lower mortality in people with ESKD on maintenance dialysis [22], obesity is a major barrier to kidney transplantation in people with ESKD. Most kidney transplant centres consider BMIs >40 kg/m² and >35 kg/m² absolute and relative contraindications, respectively, to waitlisting for kidney transplantation [23]. GLP-1RAs may improve rates of waitlisting for kidney transplantation by effecting intentional weight loss in patients with ESKD.

Existing data on the use of GLP-1RAs in patients with ESKD are derived from small randomized controlled trials (RCTs) and retrospective cohort studies, most of which used liraglutide [24]. The largest randomized study treated 20 people with type 2 diabetes and ESKD and 20 control patients with type 2 diabetes not complicated by kidney disease with liraglutide or placebo [25]. Over a 12-week period, liraglutide improved glycaemic control and reduced body weight in people with and without ESKD. The magnitude of weight loss was greater in patients without kidney disease than in those with ESKD, with body weight decreasing by 2.4 ± 0.8 kg in those with ESKD (P = .22). Plasma trough liraglutide concentrations were higher and gastrointestinal (GI) side effects occurred more frequently in the ESKD group.

A retrospective observational cohort study between 2018 and 2023 evaluated the use of oral (18%) and injectable (82%) semaglutide in patients with CKD stage 4 (n = 63), CKD stage 5 (n = 1) and ESKD on maintenance dialysis (n = 12; n = 11 HD, n = 1 peritoneal dialysis) [26]. The median duration of semaglutide therapy was 17.4 months (interquartile range 0.4–48.8). Body weight decreased by an average of 5.1 kg, corresponding to a 4.6% loss of baseline body weight. Of the 96% of the study population who had T2DM, the average haemoglobin A1c (HbA1c) reduction was 0.9%. Hypoglycaemia occurred in 20 (26%) patients. Eight patients discontinued insulin after starting semaglutide. Of the 27 patients who discontinued semaglutide, 10 (37%) discontinued due to GI side effects.

GI side effects, including constipation, are commonly encountered with GLP-1RAs, particularly during the dose titration phase [27]. From the limited available evidence, these side effects may be more common in patients with ESKD and the tolerability of GLP-1RAs will be an important determinant of their clinical utility in this population [24]. Slower dose titration may be preferable. GLP-1RA-induced constipation may be particularly problematic in patients on peritoneal dialysis, and a proactive bowel regimen should be instituted to prevent catheter-related complications [28]. The risk of hypoglycaemia is likely to be increased in patients with ESKD; other glucose-lowering therapies, particularly insulin, should be down-titrated or discontinued as GLP-1RAs are started [26]. Frequent snacking to counter hypoglycaemia may counter weight loss.

Unlike a patient with CKD, a kidney transplant patient has a single, denervated kidney, in most cases being exposed to nephrotoxic drugs, e.g. calcineurin inhibitors. Accordingly, data from CKD studies cannot be extrapolated to renal transplant recipients.

It is not known if hyperfiltration occurs at the single-nephron level in a healthy transplanted kidney with presumably intact autoregulation. The current belief is that autoregulation on the glomerular level is independent of external kidney innervation. In fact, in a small study of kidney transplant recipients treated with the SGLT2 inhibitor dapagliflozin, an early decrease in GFR, as measured by iohexol clearance, was detected in these patients [29]. This would indicate that haemodynamic responses are intact in a denervated kidney.

Nephroprotection with GLP1-RAs could take place both through reduced hydrogen–sodium antiporter activity, which increases natriuresis [30], and also through a reduction in total body mass, ectopic fat and blood pressure (BP), which might alleviate filtration pressure in glomerular filtration tufts. However, to date, we have neither experimental nor clinical data to prove that this is the case. No RCTs have been performed with GLP1-RAs in kidney transplant recipients, although studies are on their way. However, we do have available small observational studies of limited observation time (<12 months), as reviewed by Valencia-Morales et al. in 2023 [31]. Twelve observational studies in solid organ transplant patients (of which eight studies were performed in kidney transplant recipients) all reported that the use of GLP1-RAs was primarily associated with body weight loss (2–5 kg) and a corresponding decrease in HbA1c. On the other hand, marginal and inconsistent effects on eGFR (not measured) were reported, but trough levels of tacrolimus were in general not affected. Side effects were identical and just as frequent as those reported in non-transplanted patients [31]. Two small observational studies [32, 33] found a reduction in UACR with GLP1-RAs in kidney transplant recipients, but this has not been measured or confirmed in other studies.

In a recent observational study on 318 solid organ transplant recipients (of which 272 patients received a kidney transplant), the use of GLP1-RAs was associated with a reduced occurrence of MACE and all-cause mortality [34]. However, this finding has to be verified by RCTs.

Tirzepatide is a dual GLP-1 and glucose-dependent insulinotropic polypeptide (GIP) receptor agonist. It is a 39 amino acid chain connected to a C20 fatty acid chain in order to prolong its half-life [54]. The polypeptide is primarily a GIP receptor ligand but modified also to stimulate the GLP-1 receptor [55]. An additive effect with GIP receptor activation is to increase insulin sensitivity in white fat tissue, which in turn lowers plasma fatty acids as well as triglycerides. This improves insulin sensitivity in addition to the effects of GLP-1R activation [56]. This in turn increases the loss of body weight and glucose lowering. However, GIP receptors have not been located in the kidneys, so any extra effect of tirzepatide on the kidneys must be due to desirable extrarenal effects, including a reduction in body weight, reduced glycaemia in diabetes and prediabetes, lowered BP and reversal of insulin resistance.

No primary studies with tirzepatide in CKD are available. However, in the SURPASS program, which addresses the glucose- and weight-lowering properties of tirzepatide in patients with T2DM, the UACR was found to be lowered compared with comparators such as placebo or insulin. In fact, a post hoc analysis of the phase 3 SURPASS-4 study (NCT03730662↗) was clearly associated with beneficial kidney outcomes [57]. In this study, 2002 eligible patients with T2DM that was inadequately controlled with any combination of sulfonylurea, metformin or SGLT2 inhibitor were randomized in a 1:1:1:3 ratio to either treatment with three different weekly doses of tirzepatide (5 mg, 10 mg or 15 mg) or daily insulin glargine. At baseline, the median eGFR was 81 ml/min/BSA (18% had an eGFR <60 ml/min/BSA), median UACR was 15 mg/g and average BMI was 32.6 kg/m². Furthermore, baseline HbA1c was 8.5% and 25% of the patients used an SGLT2 inhibitor. In the combined tirzepatide groups (all doses taken together), the UACR did not increase over the median treatment duration of 85 weeks while it increased by 37% from baseline over the same period in those treated with insulin glargine. The eGFR slope was −1.4 ml/min/BSA/year in the combined tirzepatide groups compared with −3.6 ml/min/BSA/year in the insulin glargine group. Furthermore, patients treated with tirzepatide were less likely to develop the composite kidney endpoint (eGFR decline of ≥40%, new-onset macroalbuminuria, ESKD or renal death), an effect that was mainly driven by the effect on macroalbuminuria.

There are other nutrient-stimulated hormone-based anti-obesity medications in development. These new pharmaceuticals are based on stimulating (or inhibiting) the action of one or more gut hormones, including GLP-1, GIP, glucagon and amylin.

Glucagon promotes lipolysis in the liver and may increase energy expenditure. There is emerging evidence of the efficacy on weight reduction of dual GLP-1R/glucagon receptor (GCGR) agonists (cotadutide, mazdutide, survodutide) [58–60] and triple GLP-1R/GIP receptor/GCGR agonists (retatrutide) in overweight/obese individuals with or without diabetes. Of note, the weight reduction with the triple agonist retatrutide has been shown to be almost as high as with surgical interventions (24.2% after 48 weeks) [61].

There is an abundance of GCGR in the kidney. The expression of these receptors in the kidney is second only to the liver. Activation of GCGR by analogues can directly and indirectly influence kidney function and lead to increased GFR, natriuresis and diuresis [62]. A combination of GCGR agonism and incretin receptor agonism therefore looks promising for a beneficial impact on kidney function, such as albuminuria reduction with cotadutide [63]. Kidney benefits of the triple agonist–based anti-obesity management strategy is an active area of investigation as well. A phase 2 clinical trial is in progress to explore the effect of retatrutide on kidney outcomes in participants with overweight/obesity and CKD with or without diabetes [64].

Amylin, which is released by the beta cells of the pancreas at the same time as insulin, improves sensitivity to leptin and has a direct activation in the brain to promote the feeling of satiety after meals. The second-generation amylin analogue cagrilintide in combination with semaglutide was well tolerated, safe and very effective (15.6% weight loss with combination compared with 5.1% and 8.1% with semaglutide and cagrilintide alone, respectively) in a phase 2 trial [65]. Another amylin analogue, petrelintide, is under investigation. It demonstrated improved GI tolerance after repeated intake in humans [66]. Amycretin, an amylin analogue and GLP-1R agonist, caused significant (13.1%) weight loss after 3 months in a recent phase 1 study [67]. However, the impact of these new compounds on kidney outcomes has not yet been reported.

Recently, GLP-1RAs have emerged as key players across the CKM spectrum: overweight/obesity, T2DM and associated CKD and cardiovascular disease [8]. However, as in other areas of medicine, when new drugs appear that modify the natural course of the disease, many gaps in knowledge emerge. Examples include: do GLP-1RAs provide additional nephroprotection in patients with T2DM and CKD on RAAS blockade and SGLT2 inhibitors?; what is the optimal duration of treatment with GLP-1RAs?; is there a BMI threshold below which GLP-1RAs are ineffective?; could GLP-1RAs have a role in kidney transplantation?; what about the nephroprotective effects of GLP-1RAs in non-diabetic kidney diseases?; and what other pleiotropic effects do GLP-1RAs have?

Dialysis patients are at high cardiovascular risk and the burden of CKM syndrome is critical. There is little information in the literature on the safety and efficacy of these drugs in dialysis patients [68]. Recent observational studies suggest that in the small percentage of patients who receive them, GLP-1RAs were associated with lower all-cause mortality [69]. Future studies in this vulnerable population are needed to determine the cardiovascular benefits of this treatment.

A promising and optimistic world is opening up for the application of GLP-1RAs in renal disease as well as in other pathological conditions (metabolic dysfunction-associated steatotic liver disease, neurodegenerative disorders, obstructive sleep apnoea, polycystic ovaries, addictions etc.) [70].

RAASis and nsMRAs are not primarily accepted and classified as anti-obesity drugs. However, both groups of drugs may indirectly address obesity through their obesity-related pathways, which affect adipose tissue function, inflammation and metabolic abnormality.

Overactivation of the RAAS in obesity increases angiotensin II and aldosterone levels, which may contribute to adipogenesis, inflammation and impaired insulin sensitivity. RAASis (ACEis and ARBs) inhibit the effects of angiotensin II and aldosterone, which may reduce adipogenesis and fat accumulation [71–73]. Angiotensin II impairs insulin sensitivity by increasing oxidative stress and inflammation in metabolic tissues. RAASis can improve insulin sensitivity and glucose uptake, which could indirectly lead to weight loss and fat reduction. Furthermore, angiotensin II may influence central pathways in the hypothalamus that regulate appetite and satiety. RAASis could modulate these pathways and potentially affect food intake.

The potential of nsMRA (finerenone) as an anti-obesity drug is not well established but may be expected through some known mechanisms. Mineralocorticoid receptors in adipose tissue are involved in processes contributing to adipogenesis, inflammation and metabolic abnormalities. Finerenone affects signalling pathways of inflammation. Furthermore, finerenone may help regulate adipose tissue function by antagonising mineralocorticoid receptors. However, there is no substantial evidence from preclinical and clinical trials to directly support the use of finerenone as an anti-obesity drug. Finerenone does not directly affect weight loss, fat reduction, appetite regulation or energy consumption but may indirectly support weight management through metabolic benefits [74].

For more than 4 decades, strong evidence has confirmed that RAASis, including ACEis and ARBs, improve renal outcomes in patients with kidney disease due to diabetes and other causes by reducing proteinuria and progression towards ESKD [36]. Said protective kidney effects were obtained along with a significant reduction in cardiovascular risk [75]. Regrettably, a small number of human studies with contradicting results have evaluated the specific effects of RAASis in slowing the progression of CKD in patients with obesity [76, 77]. In a real-world primary care database, Cohen et al. [78] found no difference in renal and mortality outcomes in patients with obesity and hypertension but without diabetes treated with ACEis and ARBs when compared with other antihypertensive agents. Thus, despite the known increased RAAS activity in patients with obesity and CKD, treatment with RAASis may not confer additional nephroprotection over other antihypertensives.

In the following years, several trials tested the effects of dual RAAS blockade (combined therapy with ACEi + ARB) in patients with and without DKD, which failed to enhance CKD protection and caused increased adverse events [79, 80]. However, stagnation regarding additional CKD treatments beyond RAASis ended with the advent of SGLT2is [81]. As with ACEis and ARBs, SGLT2is improve kidney and cardiovascular outcomes in patients with CKD with and without diabetes [82]. Several studies have consistently shown that patients taking SGLT2is experience modest weight loss, which can be clinically meaningful when compared with placebo or other glucose-lowering therapies [83]. Network meta-analyses reveal an average weight reduction of ≈1.5–2 kg with SGLT2is compared with placebo, and this can be sustained for up to 4 years [83].

In recent years, GLP-1RAs have demonstrated therapeutic applications beyond traditional metabolic disorders such as obesity and T2DM. These agents lower cardiovascular death, with a favourable benefit–risk profile in patients with and without diabetes [84]. Weight loss in the FLOW trial was also 4.1 kg greater with semaglutide compared with placebo over a median 2.4-year follow-up in a population with a baseline BMI of 31 kg/m² [18]. The majority of patients enrolled in phase 3 RCTs of GLP-1RAs for cardiovascular and kidney endpoints were either overweight or obese [8]. Thus semaglutide and other GLP-1RAs are promising therapies for the treatment of patients with obesity and CKD.

Given the complementary mechanisms of action regarding glycaemic control and the strong evidence supporting the cardiovascular and kidney benefits of SGLT2is and GLP-1RAs in patients with T2DM and established cardiovascular disease, it is speculated that an approach using both glucose-lowering agents may have additive cardiovascular and kidney benefits [85]. This was tested to some degree in the FLOW trial (15.6% baseline SGLT2i use) [18], and post hoc analyses in patients with and without SGLT2is have been published [86]. However, the FLOW trial was not designed to test the combination of SGLT2i and GLP-1RA therapy in CKD; prospective RCTs are needed to investigate this further.

It is important to mention that almost all patients included in SGLT2i and GLP-1RA trials were on background ACEis or ARBs that are considered the standard of care for the treatment of patients with diabetes and CKD. Regrettably, real-life studies have shown that treatments that reduce cardiovascular and kidney outcomes in patients with diabetes and CKD have historically been underused, particularly in advanced disease [87, 88]. This condition is probably increased with real-life use of SGLT2is and GLP-1RAs, treatments that retain a higher level of discontinuation compared with RAASis [89].

Although with a reduced magnitude of benefit, the nsMRA finerenone improved kidney and cardiovascular outcomes in patients with T2DM and CKD on a background of RAASi therapy [90, 91], similar to SGLT2is and GLP-1RAs. A post hoc analysis of the FIDELIO-DKD (NCT02540993↗) and FIGARO-DKD (NCT02545049↗) trials, the pooled FIDELITY analysis, showed that in >7000 patients with CKD due to T2DM, the reduction in kidney and cardiovascular events by finerenone was not affected by varying severities of obesity [92]. Finerenone is currently recommended by guidelines as an additional drug treatment with specific kidney and cardiovascular protection in patients with T2DM and CKD [35]. The current real-life use of finerenone is probably lower than other standards of care for CKD, but its use in combination with RAASi, SGLT2i and GLP-1RA deserves attention in future studies.

To conclude, a multitarget approach for the treatment of CKD in patients with obesity and with or without diabetes is of potential additive effect in terms and cardiovascular and kidney protection. Future clinical trials should test the effects of combining these therapies in patients with obesity, T2DM and CKD who have high cardiovascular and kidney risk.

All data presented in this review are published and available in the referenced sources. No additional source data are required.