Choosing oral antihyperglycaemic drugs in people living with Type 2 diabetes and severe chronic kidney disease

Abstract

Managing hyperglycaemia in individuals with Type 2 diabetes (T2D) and advanced chronic kidney disease (CKD) involves several important considerations [1]. For instance, metformin is renally excreted, and a reduced estimated glomerular filtration rate (eGFR) increases the risk of drug accumulation, potentially leading to serious adverse events such as lactic acidosis. Although sulphonylureas and pioglitazone are primarily metabolized in the liver, their active metabolites are renally excreted, which—particularly in the context of impaired kidney function—may increase the risk of hypoglycaemia or fluid retention, respectively. SGLT2 inhibitors have limited glucose-lowering efficacy in patients with eGFR values below 20–25 mL/min/1.73 m². Additionally, GLP-1 receptor agonists promote weight loss, which may be undesirable in normal-weight patients, and insulin therapy, though often necessary, is associated with a heightened risk of hypoglycaemia and can be challenging to optimize in this vulnerable group. In contrast, the meglitinide repaglinide has been considered a particularly suitable oral agent under these circumstances. It undergoes hepatic metabolism, and only its inactive metabolites are excreted renally. Moreover, repaglinide's short duration of action and meal-time dosing provide safety and additional flexibility in glucose control.

Dipeptidyl peptidase-4 inhibitors (DPP-4i), such as sitagliptin and linagliptin, have been in clinical use for nearly two decades and are generally regarded as safe, with a favourable side-effect profile. Notably, their risk of inducing hypoglycaemia is significantly lower compared to sulphonylureas and meglitinides. This is attributed to their glucose-dependent mechanism of action, with insulinotropic effects markedly reduced at plasma glucose concentrations below approximately 4.5 mmol/L. Furthermore, DPP-4i are weight-neutral, which may be advantageous for patients with T2D and CKD in the normal or lower body mass index range. The pharmacokinetics of agents within this class vary. Thus, while sitagliptin is primarily excreted unchanged by the kidneys, linagliptin undergoes hepatic metabolism [2]. Consequently, although sitagliptin requires dose adjustment in the context of reduced eGFR, linagliptin does not. Importantly, sitagliptin itself is not considered intrinsically nephrotoxic, even in the setting of advanced CKD. However, despite these favourable characteristics, clinical evidence on the safety and efficacy of DPP-4i in patients with T2D and Stage 5 CKD (eGFR <15 mL/min/1.73 m²) remains limited.

It is therefore timely to highlight the work by Hung et al., ‘Use of dipeptidyl peptidase-4 inhibitors is associated with lower risk of severe renal outcomes in pre-dialysis patients with Type 2 diabetes: A cohort study’ in the Journal of Internal Medicine [3]. In this study, the authors investigated renal outcomes in adults with T2D treated with either DPP-4i or meglitinides. Data spanning from 2012 to 2020 were sourced from Taiwan's comprehensive nationwide healthcare claims database, which covers approximately 99% of the population. The study focused on patients with Stage 5 CKD, operationally defined as those receiving erythropoiesis-stimulating agents (ESAs), a treatment which in Taiwan is reimbursed for individuals with eGFR below 15 mL/min/1.73 m². From this cohort, patients initiating treatment with a DPP-4i (n = 5028) or a meglitinide (n = 2243) after their first ESA prescription were identified. Propensity score matching was employed to minimize confounding.

The primary composite outcome included time to renal replacement therapy, renal death, and kidney-related hospitalization. Patients treated with DPP-4i had a 14% reduced risk of experiencing the composite outcome compared with those on meglitinides (hazard ratio [HR] 0.86; 95% confidence interval [CI] 0.81–0.92), driven primarily by a lower incidence of renal replacement therapy. Additionally, the risk of severe hypoglycaemia was 41% lower in the DPP-4i group. These findings not only reinforce the renal safety of DPP-4i in patients with T2D and Stage 5 CKD but also suggest possible therapeutic advantages over meglitinides. Given the high morbidity, economic burden, and quality-of-life impact associated with the initiation of renal replacement therapy, even modest delays in disease progression may have important clinical implications.

As with all non-randomized observational studies, causality cannot be definitively established. Nevertheless, the findings from Hung et al. provide meaningful evidence supporting the safe use of DPP-4i in advanced CKD. This is consistent with recommendations from the KDIGO 2022 guidelines, which support the use of DPP-4i as safe alternatives in patients with advanced renal impairment [1]. Although several meta-analyses [4-6] have confirmed the safety of DPP-4i in people living with T2D and CKD, direct comparative studies with meglitinides remain scarce. The latter are still widely used due to their affordability and long-standing clinical availability. However, DPP-4i may be less accessible in lower resource settings due to cost considerations, which may influence treatment selection and equity in care.

Randomized controlled trials of DPP-4i, including linagliptin (CARMELINA trial) [7] and saxagliptin (SAVOR-TIMI 53 trial) [8], excluded patients with Stage 5 CKD. Both trials demonstrated reductions in albuminuria—potentially independent of glycaemic control—but failed to show improvements in hard renal endpoints. In contrast, another large population-based study from Taiwan, also using the National Health Insurance Research Database, compared DPP-4i to sulphonylureas in patients with CKD Stages 3b–5 [9]. That study found no significant differences in renal outcomes but did report reductions in hypoglycaemia-related hospitalizations (HR 0.53; 95% CI 0.43–0.64) and all-cause mortality (HR 0.71; 95% CI 0.53–0.96). Notably, patients receiving ESA therapy were excluded in that analysis, underscoring the importance of cohort selection when interpreting comparative effectiveness research.

A key methodological limitation of the Hung et al. study is its reliance on ESA prescription as a proxy for CKD Stage 5, rather than direct measurement of GFR. Moreover, the study lacked data on important clinical variables such as glycaemic control, lipid levels, blood pressure, and albuminuria. Although propensity score matching was employed, residual confounding remains possible, particularly because matching was based on medication use rather than (continuous) clinical risk factors. Differences in unmeasured baseline characteristics could still influence outcomes. Another important consideration is the study population itself. The cohort consisted almost exclusively of Taiwanese patients, approximately 96% of whom are of Han Chinese descent. This ethnic homogeneity limits the generalizability of the findings to broader, more diverse populations. Consequently, these results should be extrapolated with caution to populations with different ethnic backgrounds and healthcare infrastructures.

The mechanisms underlying the observed renal benefits of DPP-4i remain incompletely understood [10]. Although reductions in albuminuria have been documented, their impact on long-term GFR preservation appears limited. Compared to GLP-1 receptor agonists, which exert more pronounced renoprotective effects, DPP-4i are considered less potent in this regard [10]. The observed benefits may be mediated through GLP-1 receptor–dependent pathways, such as reductions in oxidative stress, or via alternative mechanisms, including antifibrotic effects [2]. Further elucidation of these pathways will be essential to fully understand the therapeutic potential of DPP-4i in advanced CKD.

It is also important to consider patient-centred factors such as treatment burden, tolerability, and adherence. In patients with advanced CKD, who often face complex medication regimens and comorbidities, agents like DPP-4i—offering a favourable safety profile and requiring minimal (or no) dose adjustment—may contribute to improved adherence.

In summary, the study by Hung et al. adds valuable real-world evidence to the growing body of literature supporting DPP-4i use in patients with T2D and severe CKD. Although limitations in design and generalizability should be acknowledged, the findings support a favourable risk–benefit profile in a population with limited therapeutic options and high clinical need. Future randomized controlled trials including diverse ethnic populations, as well as mechanistic studies exploring the pathways of renal benefit, are warranted to further clarify the role of DPP-4i in advanced CKD.

Mikael Rydén has received lecture fees from Amgen Inc., AstraZeneca, Bayer, Boehringer Ingelheim, Eli Lilly & Company, Johnson & Johnson, MSD, Novartis, Novo Nordisk A/S, and Sanofi. He has served on advisory boards for Altimmune, Atrogi, AstraZeneca, Eli Lilly & Company, Marea Therapeutics, MSD, Novo Nordisk A/S, Sanofi, and SIGRID Therapeutics AB. In addition, he has received a research grant from Novo Nordisk A/S for studies on white adipose tissue.