Semaglutide beats dulaglutide on blood sugar and weight in head-to-head trial

A head-to-head trial published on Feb. 1, 2018 found that once-weekly semaglutide lowered blood sugar and body weight by more than once-weekly dulaglutide in adults with type 2 diabetes that was not controlled on metformin alone ^[1][2]. The results appeared in The Lancet Diabetes & Endocrinology ^[1][2].

SUSTAIN 7 was an open-label, phase 3b trial run at 194 hospitals, clinics and private practices in 16 countries ^[2]. It enrolled 1,201 adults aged 18 and older whose HbA1c — a lab measure of average blood sugar over roughly the previous three months — was between 7.0% and 10.5% while they took metformin by itself ^[2]. Participants were randomly assigned to one of four weekly injections: semaglutide 0.5 mg, semaglutide 1.0 mg, dulaglutide 0.75 mg or dulaglutide 1.5 mg ^[2]. The trial was open-label, meaning patients and investigators knew which drug each person was getting; there was no placebo or blinding ^[2]. The main goal was change in HbA1c at 40 weeks, with weight change as the confirmatory secondary goal ^[2]. Enrollment ran from Jan. 6, 2016 to June 22, 2016 ^[2].

At the higher doses, HbA1c fell by 1.8 percentage points with semaglutide 1.0 mg compared with 1.4 percentage points with dulaglutide 1.5 mg, an estimated treatment difference of −0.41 percentage points (95% CI −0.57 to −0.25; p<0.0001) ^[2]. At the lower doses, semaglutide 0.5 mg reduced HbA1c by 1.5 percentage points versus 1.1 with dulaglutide 0.75 mg ^[2].

Weight fell by 6.5 kg with semaglutide 1.0 mg compared with 3.0 kg with dulaglutide 1.5 mg — roughly 14 pounds versus 7 pounds — an estimated difference of −3.55 kg (95% CI −4.32 to −2.78; p<0.0001) ^[2]. At the lower doses, the figures were 4.6 kg with semaglutide 0.5 mg versus 2.3 kg with dulaglutide 0.75 mg ^[2].

Gastrointestinal problems were the most frequently reported side effect and the most common reason people stopped treatment ^[2]. They were reported by 133 of 300 patients (44%) on semaglutide 1.0 mg and 143 of 299 (48%) on dulaglutide 1.5 mg, and by 129 of 301 (43%) on semaglutide 0.5 mg and 100 of 299 (33%) on dulaglutide 0.75 mg ^[2]. In total, 72 patients (6%) withdrew from the trial ^[2]. Six deaths occurred: one in each semaglutide group and two in each dulaglutide group ^[2]. The causes of those deaths have not been reported ^[2].

The authors concluded that at both low and high doses, semaglutide was better than dulaglutide at improving blood sugar control and reducing weight, with a similar safety profile ^[2]. The trial was funded by Novo Nordisk ^[2], and it drew an accompanying commentary in the same journal titled "Once-weekly GLP-1R agonists: moving the goal posts" ^[2].

Why it matters for patients

Before this trial, comparisons between GLP-1 drugs mostly came from separate studies with different patient groups, which makes them hard to line up. SUSTAIN 7 put two of these weekly drugs in the same trial, in the same patients, at the same time — the kind of comparison that is usually more reliable for weighing one option against another ^[2].

The gaps were real but not enormous: about half a percentage point of HbA1c and about 3.5 kg (roughly 7 to 8 pounds) of weight separated the higher doses over 40 weeks ^[2]. Side effects were common in every group, and stomach and gut complaints were the leading reason people quit ^[2].

Two limits are worth knowing. The trial enrolled only people with type 2 diabetes taking metformin alone, so the averages may not carry over to people without diabetes, or to those taking insulin or other drugs ^[2]. And because the trial was open-label, patients knew which medicine they received, which can affect how they report symptoms and how they stick with treatment ^[2].

What the trial did not show

The study ran for 40 weeks ^[2]. It did not measure heart attacks, strokes, hospitalizations or other long-term health outcomes, and it did not test either drug against other classes of diabetes medicine ^[2]. Average results also hide wide differences between individuals; a person's own response can be larger or smaller than the group average reported here ^[2].