Long-term effects, safety and tolerability of oral semaglutide (SEMA; a GLP-1 receptor agonist [GLP-1RA]) vs sitagliptin (SITA) as add-on to metformin ± sulfonylurea was investigated in patients (pts) with T2D in a 78-week, double-blind, double-dummy trial. Pts were randomized to once daily oral SEMA 3 mg (N=466), 7 mg (N=466) or 14 mg (N=465), or SITA 100 mg (N=467). Primary endpoint was change in HbA1c, confirmatory secondary endpoint was change in body weight, both from baseline to week 26. Two estimands were defined (‘treatment policy [TPol] estimand’: treatment effect regardless of trial product discontinuation and rescue medication use; ‘trial product estimand’: treatment effect assuming on trial product without rescue medication use) in all randomized pts. Confirmatory testing was based on the TPol estimand; within each dose level, non-inferiority for the primary endpoint (non-inferiority margin: 0.3%) had to be confirmed before testing superiority for the primary and confirmatory secondary endpoints. Estimated week 26 HbA1c reductions with 7 and 14 mg were –1.0% and –1.3%, vs –0.8% with SITA, and were superior (TPol estimand; estimated treatment difference vs SITA [ETD; 95% CI]: –0.3% [–0.4, –0.1]; –0.5% [–0.6, –0.4]; both P<0.001). Non-inferiority of 3 mg vs SITA was not confirmed (ETD: 0.2% [0.1, 0.3]; P=0.09); HbA1c reductions favored SITA (P=0.008). Similar results were obtained with the trial product estimand at week 26 (ETD: –0.3% [–0.4, –0.2]; –0.6% [–0.7, –0.5] for 7 and 14 mg; both P<0.001 favoring oral SEMA; 0.2% [0.1, 0.4] for 3 mg; P<0.001 favoring SITA). At week 78, HbA1c reductions with 14 mg were statistically significantly greater vs SITA for both estimands; there was no statistically significant difference with 3 mg (both estimands) or 7 mg (TPol estimand). All oral SEMA doses significantly lowered body weight vs SITA at week 26 for the TPol estimand (ETD: –0.6 kg [–1.1, –0.1]; –1.6 kg [–2.0, –1.1]; –2.5 kg [–3.0, –2.0] for 3, 7 and 14 mg; all P<0.02), confirming superiority of 7 and 14 mg (3 mg superiority not tested), and the trial product estimand (ETD: –0.5 kg [–1.0, –0.1]; –1.5 kg [–2.0, –1.1]; –2.6 kg [–3.1, –2.1] for 3, 7 and 14 mg; all P<0.03). Week 78 body weight reductions were also statistically significant, favoring oral SEMA (all doses, both estimands). Adverse events (AEs) occurred similarly across treatment arms. The most common AE with oral SEMA was transient mild/moderate nausea, affecting 7.5–15.7% of pts. Serious AEs were reported by 13.7%, 10.1% and 9.5% of pts for 3, 7 and 14 mg, and 12.4% for SITA. AEs led to premature trial product discontinuation in 5.6%, 5.8% and 11.6% of pts for 3, 7 and 14 mg, and 5.2% for SITA, mainly due to gastrointestinal AEs. In conclusion, oral semaglutide 7 and 14 mg provided statistically superior HbA1c and body weight reductions at week 26 compared with SITA 100 mg (TPol estimand). Safety and tolerability were consistent with other GLP-1RAs.
Read full abstract