Oral semaglutide 7 mg and 14 mg beat sitagliptin on blood sugar and weight in PIONEER 3 trial

Adults with type 2 diabetes who took oral semaglutide at 7 mg or 14 mg once daily had larger reductions in blood sugar and body weight than those who took sitagliptin 100 mg, results of the PIONEER 3 trial published in JAMA on April 16, 2019 show. The 3 mg starting dose did not meet the trial's standard for being at least as good as sitagliptin on blood sugar ^[1].

The phase 3a trial enrolled 1,864 adults whose type 2 diabetes was not controlled with metformin, with or without a sulfonylurea. It ran at 206 sites in 14 countries over 78 weeks, from February 2016 to March 2018. Participants were randomly assigned to once-daily oral semaglutide at 3 mg (466 people), 7 mg (466) or 14 mg (465), or to sitagliptin 100 mg (467). Everyone began semaglutide at 3 mg and increased every four weeks until reaching their assigned dose ^[1].

From the start of the trial to week 26, the 7 mg dose lowered HbA1c — a measure of average blood sugar over roughly three months — by an additional 0.3 percentage points compared with sitagliptin (95% CI, −0.4 to −0.1), and the 14 mg dose by an additional 0.5 points (95% CI, −0.6 to −0.4); P<.001 for both. Body weight fell by an additional 1.6 kg with 7 mg (95% CI, −2.0 to −1.1) and 2.5 kg with 14 mg (95% CI, −3.0 to −2.0) versus sitagliptin, also P<.001 ^[1].

The trial set a non-inferiority margin of 0.3 percentage points for HbA1c, meaning a dose had to come within that margin of sitagliptin before superiority could be tested. The 3 mg dose did not clear that bar, and the authors concluded there was no significant benefit with 3 mg. At week 78, the 14 mg dose still showed significantly greater reductions than sitagliptin in both HbA1c and weight; the abstract does not give the size of those week-78 differences ^[1].

Of the 1,864 participants, 1,758 (94.3%) completed the trial. Treatment was stopped early by 16.7% of the 3 mg group, 15.0% of the 7 mg group, 19.1% of the 14 mg group and 13.1% of the sitagliptin group ^[1]. The published abstract does not list rates of specific side effects, such as nausea or vomiting.

Participants averaged 58 years old (SD, 10), with a baseline HbA1c of 8.3% (SD, 0.9%) and a body mass index of 32.5 (SD, 6.4); 879 (47.2%) were women ^[1]. The comparison was against sitagliptin, an oral DPP-4 inhibitor, not against injectable GLP-1 drugs or other oral GLP-1 treatments. The authors wrote that further research is needed to assess effectiveness in a clinical setting ^[1].

The disclosure statement lists three authors as Novo Nordisk employees, and several other authors report consulting fees or research support from Novo Nordisk, Eli Lilly, Sanofi, Boehringer Ingelheim and other companies ^[1]. A funder for the trial is not named in this record. The trial is registered as NCT02607865 ^[1].

Why it matters for patients

This was a phase 3 test of an oral GLP-1 drug against another class of oral diabetes medicine, a comparison the authors say had not been made before ^[1]. For people who would rather take a daily pill than give themselves an injection, the results describe how the two higher doses performed against sitagliptin over 26 weeks, and the 14 mg dose continued to show greater reductions at 78 weeks ^[1].

Two limits are worth noting. First, the 3 mg dose was used as a step-up dose in this trial, so its results do not describe what the 7 mg or 14 mg doses did ^[1]. Second, the trial did not compare oral semaglutide with injectable GLP-1 medicines, so it does not answer how the pill stacks up against those drugs, and cardiovascular outcomes are not reported in the abstract ^[1].

The results are averages across a trial population and do not show how any one person responded. Findings were also measured in adults with a baseline HbA1c of 8.3% taking metformin with or without a sulfonylurea, and have not been shown to apply to other groups ^[1]. Cost, insurance coverage and side effects were not part of this trial's endpoints.