A weekly insulin shot passes the test it was set

In Short. Eli Lilly's newly approved weekly insulin for type 2 diabetes controls average blood sugar about as well as the usual daily shot, which means far fewer injections but not, so far, a better or safer drug.

TYPE 2 DIABETES is a slow failure of the body's sugar control. Cells stop responding well to insulin, the hormone that moves sugar out of the blood, and the pancreas cannot make enough extra to compensate, so blood sugar runs high. When pills and other medicines are not enough, doctors add basal insulin, a slow, steady background dose that covers the body's needs between meals and overnight (mealtime insulin handles the rise after food). For most, that means a daily shot.

The new weekly kind is an insulin molecule fused to a piece of an antibody. Its label says its size slows clearance by the kidneys and the antibody piece slows its breakdown, so it lingers. The drug is Onswik (insulin efsitora alfa-gobe), approved by America's Food and Drug Administration on September 24th for adults with type 2 diabetes. The approval rests on four phase 3 trials, called QWINT, in more than 3,400 adults.

In each, chance alone decided who got weekly or daily insulin, so the groups started out alike. That makes a trial randomized. Patients and doctors knew which insulin they were on, a design called open-label. This matters less for a lab blood test than for a symptom score, but can still shape how doses are adjusted and how often low-sugar episodes get reported. The weekly shot was compared not with a dummy injection (a placebo) but with a daily insulin people already take, glargine or degludec. That is an active comparator. Each trial was judged on one result chosen in advance, its primary endpoint. Here it was the change in HbA1c (A1C for short), a blood test that reflects average blood sugar over roughly the past three months. The weekly shot only had to show it was not worse than the daily one by more than a pre-set amount, a non-inferiority design. That pre-set amount, the margin, was 0.4 percentage points of A1C.

QWINT-2 gave 928 adults who had never used insulin either weekly Onswik or daily degludec for 52 weeks. A1C started at about 8.2%. In the published paper's main analysis, which counts everyone however their treatment went, it fell by 1.26 points on the weekly shot and 1.17 on the daily one. Lilly's figures of 1.34 and 1.26, reported by BioSpace, assume everyone stayed on the assigned drug. Either way the gap was about a tenth of a point. Even the least flattering end of its statistical range left weekly insulin 0.04 points behind, well inside the non-inferiority margin.

QWINT-1 (795 adults new to insulin, against glargine), QWINT-3 (986 already on basal insulin, against degludec) and QWINT-4 (730 also taking mealtime insulin, against glargine) asked the same question of different people. All four met their primary endpoint, and Lilly calls overall safety similar to the daily insulins.

Safety is where this design says least. Insulin's main hazard is hypoglycemia, blood sugar low enough to cause shaking, confusion or worse. In QWINT-2, clinically significant or severe episodes came at 0.58 per patient per year on weekly insulin and 0.45 on daily, a gap the trial could not tell apart from chance. People with type 1 diabetes fared worse. In a year-long trial, 10% of them had a severe episode on Onswik against 3% on degludec, so the label advises against its use in type 1. It is unproven in children.

What the trials settle is narrow but useful. For adults with type 2 diabetes, a weekly shot controls average blood sugar about as well as a daily one, with about 52 injections a year instead of 365 (Lilly stresses that count is not an efficacy or safety claim). They were not built to show that weekly insulin is better, protects the heart or kidneys, or suits type 1. For people who struggle with daily injections, that is a real option. Whether fewer shots mean better health over decades is something these trials never asked.