Weekly semaglutide beat insulin glargine on A1c and weight in SUSTAIN 4

In a 30-week phase 3 trial, once-weekly semaglutide lowered A1c and body weight more than once-daily insulin glargine in adults with type 2 diabetes whose blood sugar was not controlled on metformin, with or without a sulfonylurea ^[1]. Fewer people taking semaglutide had episodes of low blood sugar ^[1].

The study, called SUSTAIN 4, enrolled 1,089 participants at 196 sites in 14 countries, randomizing them between Aug. 4, 2014, and Sept. 3, 2015 ^[1]. Everyone was an insulin-naive adult aged 18 or older with inadequate blood sugar control on metformin alone or metformin plus a sulfonylurea ^[1]. Participants were assigned in equal groups to once-weekly subcutaneous semaglutide at 0.5 mg or 1.0 mg, or to once-daily insulin glargine starting at 10 international units per day and adjusted weekly toward a pre-breakfast glucose target of 72 to 99 mg/dL ^[1]. Background metformin and sulfonylurea treatment continued in all groups ^[1]. The analysis included 362 people on 0.5 mg semaglutide, 360 on 1.0 mg semaglutide and 360 on insulin glargine ^[1].

A1c started at a mean of 8.17% ^[1]. At week 30, it fell 1.21 percentage points with 0.5 mg semaglutide and 1.64 points with 1.0 mg, compared with 0.83 points with insulin glargine ^[1]. The estimated differences versus insulin glargine were 0.38 points and 0.81 points, with both results reported as statistically significant (p<0.0001) ^[1]. The trial used a non-inferiority margin of 0.3 percentage points for the A1c comparison ^[1].

Mean body weight at the start was 93.45 kg ^[1]. By week 30, the 0.5 mg semaglutide group lost an average of 3.47 kg and the 1.0 mg group lost 5.17 kg, while the insulin glargine group gained 1.15 kg ^[1]. The estimated differences versus insulin glargine were 4.62 kg and 6.33 kg in favor of semaglutide (both p<0.0001) ^[1].

Severe or blood glucose-confirmed low blood sugar was reported by 16 participants (4%) on 0.5 mg semaglutide and 20 (6%) on 1.0 mg, compared with 38 (11%) on insulin glargine ^[1]. Severe low blood sugar occurred in two people (less than 1%) on 0.5 mg semaglutide, five (1%) on 1.0 mg and five (1%) on insulin glargine ^[1]. Nausea was the most common side effect with semaglutide, reported by 77 people (21%) on 0.5 mg and 80 (22%) on 1.0 mg; the most common side effect with insulin glargine was nasopharyngitis, reported by 44 people (12%) ^[1]. More participants stopped treatment early on semaglutide — 49 (14%) on 0.5 mg and 55 (15%) on 1.0 mg — than on insulin glargine, 26 (7%) ^[1]. Most of those discontinuations were due to side effects, mostly gastrointestinal with semaglutide and skin reactions such as rash, itching and hives with insulin glargine ^[1].

Six deaths were reported during the trial: four (1%) in the 0.5 mg semaglutide group, three from cardiovascular causes and one from pancreatic carcinoma that was assessed as possibly related to the study medication, and two (less than 1%) in the insulin glargine group, both cardiovascular ^[1]. The report does not establish what caused these deaths. The trial was funded by Novo Nordisk A/S and registered as NCT02128932 ^[1].

Why it matters for patients

For people with type 2 diabetes who cannot reach their blood sugar goals on metformin, insulin is one common next step. This trial put a once-weekly injectable non-insulin option, semaglutide — the molecule later sold in the US as Ozempic, Wegovy and Rybelsus — directly against daily insulin glargine in that exact situation ^[1]. The two treatments moved weight in opposite directions: semaglutide users lost weight on average, while insulin glargine users gained it ^[1]. Low blood sugar was less frequent with semaglutide, though it still occurred ^[1].

The practical trade-off showed up in side effects. Nausea was common with semaglutide, and more people on semaglutide stopped treatment early than on insulin glargine ^[1]. The trial was open-label, meaning participants and investigators knew which treatment was assigned ^[1], and it followed people for 30 weeks ^[1]. Whether the differences in A1c and weight hold up over years, whether they change the risk of diabetes complications, and how semaglutide compares with other non-insulin options have not been reported from this trial ^[1].