Oral semaglutide matches injected liraglutide for A1C and beats it on weight in 711-patient trial

In July 2019, The Lancet published a 711-patient trial showing that oral semaglutide, a once-daily GLP-1 pill now sold as Rybelsus, matched an injected GLP-1 drug on blood sugar and beat it on weight loss. The study, called PIONEER 4, tested oral semaglutide 14 mg against subcutaneous liraglutide 1.8 mg and placebo in adults with type 2 diabetes ^[1].

PIONEER 4 was a randomized, double-blind, double-dummy phase 3a trial run at 100 sites in 12 countries. Researchers screened 950 people between Aug 10, 2016, and Feb 7, 2017, and randomly assigned 711 of them: 285 to oral semaglutide, 284 to injected liraglutide and 142 to placebo. Participants were 18 or older, had an A1C between 7.0% and 9.5%, and were on a stable dose of metformin of at least 1,500 mg (or the most they could tolerate), with or without an SGLT-2 inhibitor. About 48% were female and the average age was 56. Doses were escalated over time, and treatment ran for 52 weeks, with the main measurement taken at week 26 ^[1].

At week 26, A1C fell by 1.2 percentage points with oral semaglutide, 1.1 points with liraglutide and 0.2 points with placebo. The trial was built to test whether oral semaglutide was at least as good as liraglutide, using a margin of 0.4 percentage points. The estimated difference between the two active drugs was 0.1 points (95% CI 0.3 to 0.0), which met that bar. Against placebo, the difference was 1.1 points (95% CI 1.2 to 0.9) ^[1].

Weight fell by 4.4 kg with oral semaglutide, 3.1 kg with liraglutide and 0.5 kg with placebo. The gap between the two active drugs was 1.2 kg (95% CI 1.9 to 0.6; p=0.0003) in favor of the pill, and 3.8 kg against placebo ^[1].

Those results depend in part on how the researchers counted people who stopped treatment or needed rescue medication. Under one method, oral semaglutide was non-inferior to liraglutide on A1C; under the other, it was significantly better on both A1C (a 0.2-point difference, p=0.0056) and weight (1.5 kg, p<0.0001) ^[1].

Adverse events were more common with the active drugs: 80% of people on oral semaglutide reported at least one, compared with 74% on liraglutide and 67% on placebo. Most participants finished the 52-week trial: 97% in the oral semaglutide group, 96% in the liraglutide group and 94% in the placebo group ^[1]. The authors wrote that the safety and tolerability of oral semaglutide were similar to subcutaneous liraglutide, and that the pill could potentially lead to earlier initiation of GLP-1 receptor agonist therapy. The study was funded by Novo Nordisk A/S, and a correction notice for the paper was published in The Lancet on the same date ^[1].

Why it matters for patients

The trial was run in people with type 2 diabetes, not in people seeking weight loss, and its main questions were about blood sugar and weight over 26 weeks. It did not test whether a GLP-1 pill reduces heart attacks, strokes, kidney problems or deaths, and those outcomes are not reported from this study ^[1].

Both active drugs lowered A1C by roughly a point in this group, and the difference between them was small. That means the practical choice between a daily pill and an injection may turn on other things a patient and clinician weigh, including tolerability, price and insurance coverage. None of those details appear in this publication ^[1].

Side effects were common in every group, including placebo, but the report does not spell out which side effects occurred or how many people stopped treatment because of them ^[1]. The comparison here was against liraglutide and placebo; how the pill stacks up against other diabetes medicines is not answered by this trial ^[1].

What happens next

Any regulatory filing, approval or launch date for oral semaglutide has not been reported in the trial publication ^[1]. Whether the pill changes long-term outcomes such as heart attacks or kidney disease also has not been reported from this study, which followed participants for 52 weeks ^[1]. A correction notice and two commentaries on the trial were published alongside it ^[1].