Type 2 diabetes (T2D) and chronic kidney disease (CKD) significantly increase the risk of kidney failure, cardiovascular (CV) complications, and mortality. Both glucagon-like peptide-1 receptor agonists (GLP-1RAs) and sodium-glucose cotransporter-2 inhibitors (SGLT2i) have demonstrated independent benefits in reducing CV and kidney events in patients with T2D and high CV risk, primarily demonstrated in trials powered for CV outcomes. However, the effects of combining these two classes of medications remain unclear. Previous trials with SGLT2i, some with kidney outcomes as the primary endpoint, showed substantial renal benefits. The FLOW trial was designed to specifically examine the impact of the GLP-1RA semaglutide on kidney outcomes in individuals with T2D and CKD. The study aimed to address the knowledge gap regarding the combined use of GLP-1RAs and SGLT2i, especially considering their potentially complementary mechanisms of action on the CV system and kidneys. This study performs a prespecified analysis to investigate the potential benefits and safety of semaglutide in reducing kidney and CV events, stratified by baseline SGLT2i use and also analyzing the effects of SGLT2i use initiated after randomization. The potential for additive benefits from this combination therapy is of significant clinical interest for improving patient outcomes in this high-risk population.

Prior research has established the individual benefits of GLP-1RAs and SGLT2i in patients with T2D and high cardiovascular risk. Multiple trials have shown that GLP-1RAs and SGLT2i independently reduce cardiovascular and kidney events, although some of the earlier trials were primarily powered for cardiovascular outcomes. More recent trials focused specifically on kidney outcomes using SGLT2i, providing stronger evidence of their renal benefits. However, the combined use of GLP-1RAs and SGLT2i has been less studied, particularly in the context of chronic kidney disease, leading to this study's focus on investigating their combined effects. Observational studies and meta-analyses have suggested potential additive benefits, but a dedicated, randomized controlled trial was needed to confirm these findings in a well-defined population. The lack of large-scale, randomized controlled trials examining the combined use of these agents necessitated the present investigation.

The FLOW trial (ClinicalTrials.gov identifier: NCT03819153) was a randomized, double-blind, placebo-controlled, international, multicenter trial conducted between June 2019 and May 2021. A total of 3,533 participants with T2D and CKD were randomized 1:1 to receive either semaglutide 1.0 mg once weekly subcutaneously or a matching placebo. Participants were stratified by baseline SGLT2i use (yes/no). Inclusion criteria included an eGFR of 25–75 ml/min/1.73 m² and a UACR within specified ranges, and the use of maximal tolerated doses of RASi. Exclusion criteria included various factors that could affect patient safety or trial compliance. The primary outcome was a five-component composite of: 1) onset of kidney failure (defined as commencement of chronic dialysis, kidney transplantation, or a sustained reduction in eGFR to <15 ml/min/1.73 m²), 2) a sustained (≥28 days) ≥50% reduction in eGFR from baseline, 3) death due to kidney causes, 4) death due to cardiovascular causes, 5) all-cause death. Three key secondary outcomes were: 1) total eGFR slope, 2) time to first major adverse cardiovascular event (MACE), and 3) all-cause mortality. Other outcomes were assessed as exploratory. The trial was event-driven and powered to detect a 20% relative risk reduction in the primary outcome with semaglutide versus placebo. Statistical analysis was performed using intention-to-treat principles, with time-to-event endpoints analyzed using stratified Cox proportional hazards models. Subgroup analyses were conducted, stratified by baseline SGLT2i use, to assess heterogeneity of treatment effects. Continuous outcomes were analyzed using analysis of covariance (ANCOVA), with multiple imputation used for missing data. Missing eGFR data were not imputed. An interim analysis was conducted which triggered early completion of the study for efficacy. Post-hoc analysis was performed for the primary outcome based on changes in eGFR using cystatin C. Time-dependent Cox models investigated the effect of SGLT2i initiation during the study, treating SGLT2i use as both a time-dependent variable and a time-dependent covariate.

During a median follow-up of 3.4 years, semaglutide reduced the risk of the primary composite kidney outcome by 24% (95% CI: 34%, 12%) compared to placebo. However, this benefit was primarily driven by the subgroup of participants not on SGLT2i at baseline. In the subgroup without baseline SGLT2i use, semaglutide reduced the risk of the primary composite kidney outcome to 0.73 (95% CI: 0.63, 0.85; P<0.001). In contrast, the subgroup with SGLT2i use at baseline showed no significant difference between semaglutide and placebo (HR 1.07; 95% CI: 0.69, 1.67; P=0.755). A similar pattern was observed for the kidney-specific four-component outcome (excluding CV death). Semaglutide consistently showed benefits on reducing the rate of eGFR decline, regardless of baseline SGLT2i use (0.75 (95% CI: -0.01, 1.50) in SGLT2i subgroup, 1.25 (95% CI: 0.91, 1.58) in non-SGLT2i subgroup). The treatment effects of semaglutide on major adverse cardiovascular events (MACE) and all-cause mortality were similar in both subgroups (Pinteraction 0.741 and 0.901, respectively), indicating no significant interaction between SGLT2i use and semaglutide's effects on these outcomes. Semaglutide also lowered UACR and HbA1c compared to placebo and caused more significant body weight reduction, with no significant differences in these effects between the two subgroups. The analysis also assessed the impact of new SGLT2i initiations during the trial. In participants who initiated SGLT2i, the HR for the primary outcome with semaglutide versus placebo was 0.88, while in those who never used SGLT2i, it was 0.70. When SGLT2i use was treated as a time-dependent covariate, the HR for the primary outcome was 0.75. Post-hoc analyses using cystatin C for eGFR measurements showed consistent results with no statistical interaction for SGLT2i use.

The FLOW trial demonstrates that semaglutide provides substantial benefits in reducing kidney and CV outcomes and mortality in patients with T2D and CKD, irrespective of concomitant SGLT2i use. The observed lack of interaction between semaglutide and SGLT2i for CV and mortality outcomes suggests that the benefits of semaglutide on these endpoints may be independent and potentially additive to the effects of SGLT2i. However, the relatively small number of participants using SGLT2i at baseline limits the study's power to definitively conclude on interactions. The findings suggest that semaglutide's benefit on eGFR decline is consistent across subgroups regardless of SGLT2i use, although the trial duration may have limited the ability to detect effects beyond changes in eGFR and UACR for the combined therapy. These results support the potential for a complementary therapeutic approach, using semaglutide along with SGLT2i, in managing T2D and CKD. While post-hoc analysis including individuals initiating SGLT2i during the trial showed results consistent with independent actions of both drugs, further research is needed to fully characterize this effect. Future large-scale outcome trials directly comparing the combined use of semaglutide and SGLT2i with either agent alone are essential to definitively assess their combined therapeutic potential.

Semaglutide effectively reduces the risk of kidney, cardiovascular, and all-cause mortality in patients with T2D and CKD, with no evidence of significant interaction effects related to concomitant SGLT2i use. This suggests a potentially additive benefit of combining semaglutide with SGLT2i for managing this high-risk patient population. Future research should include dedicated, large-scale trials designed to assess the benefits of combined therapy directly.

The primary limitation is the relatively small number of participants using SGLT2i at baseline (15.6%), which reduced the study's power to detect smaller, yet potentially clinically relevant interactions between semaglutide and SGLT2i. The uncontrolled initiation of SGLT2i treatment during the study also introduces potential bias. The relatively short duration of the trial (median follow-up 3.4 years) might have limited the ability to fully assess the long-term effects of combined semaglutide and SGLT2i therapy on kidney outcomes beyond changes in eGFR and UACR. Furthermore, the study was not designed to evaluate interactions based on specific subgroups, therefore the interpretation regarding interactions requires caution.