Effects of semaglutide with and without concomitant SGLT2 inhibitor use in participants with type 2 diabetes and chronic kidney disease in the FLOW trial

Abstract

People with type 2 diabetes and chronic kidney disease have a high risk for kidney failure and cardiovascular (CV) complications. Glucagon-like peptide-1 receptor agonists and sodium-glucose cotransporter-2 inhibitors (SGLT2i) independently reduce CV and kidney events. The effect of combining both is unclear. FLOW trial participants with type 2 diabetes and chronic kidney disease were stratified by baseline SGLT2i use (N = 550) or no use (N = 2,983) and randomized to semaglutide/placebo. The primary outcome was a composite of kidney failure, ≥50% estimated glomerular filtration rate reduction, kidney death or CV death. The risk of the primary outcome was 24% lower in all participants treated with semaglutide versus placebo (95% confidence interval: 34%, 12%). The primary outcome occurred in 41/277 (semaglutide) versus 38/273 (placebo) participants on SGLT2i at baseline (hazard ratio 1.07; 95% confidence interval: 0.69, 1.67; P = 0.755) and in 290/1,490 versus 372/1,493 participants not taking SGLT2i at baseline (hazard ratio 0.73; 0.63, 0.85; P < 0.001; P interaction 0.109). Three confirmatory secondary outcomes were predefined. Treatment differences favoring semaglutide for total estimated glomerular filtration rate slope (ml min^-1/1.73 m²/year) were 0.75 (-0.01, 1.5) in the SGLT2i subgroup and 1.25 (0.91, 1.58) in the non-SGLT2i subgroup, P interaction 0.237. Semaglutide benefits on major CV events and all-cause death were similar regardless of SGLT2i use (P interaction 0.741 and 0.901, respectively). The benefits of semaglutide in reducing kidney outcomes were consistent in participants with/without baseline SGLT2i use; power was limited to detect smaller but clinically relevant effects. ClinicalTrials.gov identifier: NCT03819153 .

Fig. 1. Outcomes for semaglutide 1.0 mg versus placebo in subgroups with/without SGLT2i use at baseline for the primary five-component outcome and for the kidney-specific, four-component outcome.

a, For the primary five-component outcome, cumulative incidence rates were calculated using the Aalen–Johansen method with non-CV and non-renal death as a competing risk. b, For the kidney-specific, four-component outcome (five-component outcome without CV death), cumulative incidence rates were calculated using the Aalen–Johansen method with all-cause death, excluding renal death, as a competing risk. A stratified Cox proportional hazards model was used (stratified by SGLT2i use at baseline (yes/no)), with treatment and subgroup as fixed factors and two-sided P values. In a, the P value for semaglutide versus placebo in participants on SGLT2i at baseline was 0.7546, and for those not on SGLT2i, it was <0.0001; the P interaction value was 0.1090.

Fig. 2. The composite primary outcome and its components, and MACE and all-cause death outcomes.

Data from the in-trial period (full analysis set). Data are presented as HR (blue symbol) and corresponding 95% CI (error bars). Time from randomization to relevant endpoint was analyzed using a stratified Cox proportional hazards model with treatment as a categorical fixed factor and two-sided P values. Participants without events of interest were censored at the end of their in-trial period. For the subgroup analyses, estimated HR and corresponding CI were calculated in a Cox proportional hazards model with interaction between treatment groups and subgroup as a fixed factor. P interaction values for the test of no interaction effect between SGLT2i use and treatment using a score test are shown. There was no renal death in the SGLT2i use subgroup, which is not displayed here. MI, myocardial infarction.

Fig. 3. eGFR over time with or without SGLT2i use at baseline, based on serum creatinine and cystatin C.

a–d, Observed data from the in-trial period. eGFR-creatinine in subgroup with SGLT2i use at baseline (a), eGFR-creatinine in subgroup without SGLT2i use at baseline (b), eGFR-cystatin C in subgroup with SGLT2i use at baseline (c) and eGFR-cystatin C in subgroup without SGLT2i use at baseline (d). Error bars are ± s.e.m. Numbers shown under the plots represent the number of participants contributing to the means. s.e.m., standard error of the mean.

Extended Data Fig. 1. Participant disposition and flow.

*Five participants were randomized more than once, and one participant was excluded due to good clinical practice issues at the site. ^†Fourteen additional withdrawals (two participants receiving semaglutide and twelve receiving placebo) were by sponsor decision, following closure of two sanctioned sites in Russia. ^‡Participants who attended the follow-up visit or died during the trial. FAS, full analysis set; SGLT2i, sodium-glucose co-transporter-2 inhibitors.

Extended Data Fig. 2. Compliance on randomized treatment over time by baseline SGLT2i use (participants planned to be on treatment).

The lines represent the proportion of participants compliant on treatment until the planned end-of-treatment visit. Compliance on-treatment period: all days from the date of first dose to the end of the in-trial period where a dose has been administered. The period could include non-consecutive time intervals, and the dose did not have to be the target dose. SGLT2i, sodium-glucose co-transporter-2 inhibitors.

Extended Data Fig. 3. Proportion of participants on SGLT2i at baseline who continued on SGLT2i during the trial.

Proportions are based upon the number of participants in the trial at a given time point. SGLT2i, sodium-glucose co-transporter-2 inhibitors.

Extended Data Fig. 4. Proportion of participants not on SGLT2i at baseline who started on SGLT2i during the trial (grey and blue areas).

Proportions are based upon the number of participants in the trial at a given time point. The red line indicates open-label use of GLP-1 RA. GLP-1 RA, glucagon-like peptide-1 receptor agonist; SGLT2i, sodium-glucose co-transporter-2 inhibitors.

Extended Data Fig. 5. Time to first MACE with semaglutide or placebo according to baseline use of SGLT2i.

Cumulative incidence was calculated using the Aalen–Johansen estimator for first MACE accounting for non-CV death as a competing event. CV, cardiovascular; MACE, major adverse CV event (non fatal myocardial infarction, non fatal stroke or CV death); SGLT2i, sodium-glucose co-transporter-2 inhibitors.

Extended Data Fig. 6. Time to all-cause death with semaglutide or placebo according to baseline use of SGLT2i.

Cumulative incidence was calculated using the Aalen–Johansen method for all-cause death. SGLT2i, sodium-glucose co-transporter-2 inhibitors.

Extended Data Fig. 7. UACR over time in participants with SGLT2i use at baseline (a) or without (b).

Observed data from the in-trial period. Error bars are +/– the standard error of the mean on the logarithmic scale back-transformed to natural scale with the exponential. Numbers shown in the lower panel represent the number of participants contributing to the means. SGLT2i, sodium-glucose co-transporter-2 inhibitors; UACR, urine albumin-to-creatinine ratio.

Extended Data Fig. 8. Most frequent serious adverse events by baseline SGLT2i use; by system organ class (a) and by preferred term (b).

Percentages are sorted in descending order by the frequency in the semaglutide 1.0 mg group. Most frequent was defined as the 10 most frequent terms by absolute count, where absolute count was >1 in the total population (up to 15 in case of a tie). N, number of participants in the semaglutide and placebo arms. COVID-19, coronavirus disease 2019; SGLT2i, sodium-glucose co-transporter-2 inhibitors. Note that the percentage range (abscissa) varies.