ACHIEVE-J was an open-label phase 3 trial at 40 centres in Japan. It randomly assigned 401 adults with type 2 diabetes, managed on diet and exercise alone or on one or two oral glucose-lowering drugs, to once-daily oral orforglipron at 3 mg, 12 mg or 36 mg. There was no placebo arm and no masking, and the primary endpoint was safety over 52 weeks. Of the 401, 352 (88%) completed treatment.
Orforglipron is a small-molecule GLP-1 (glucagon-like peptide-1) receptor agonist taken as a daily tablet rather than injected, which is why it matters in a setting where injection reluctance and cold-chain storage both limit GLP-1 use. The safety picture here is familiar rather than novel. In total 339 of 401 participants (85%, 95% CI 80.7 to 87.8) had at least one treatment-emergent adverse event, most of them mild (67%) or moderate (16%). Stopping treatment because of an adverse event tracked the dose closely: 5% at 3 mg (95% CI 2.6 to 10.5), 8% at 12 mg (4.6 to 14.0) and 14% at 36 mg (9.3 to 21.1), with gastrointestinal symptoms the commonest reason. Level 2 hypoglycaemia (blood glucose below 54 mg/dL) occurred in 2% of the 12 mg group and 2% of the 36 mg group, and there was no severe hypoglycaemia.
There is nothing to prescribe today, but there is something to plan. This trial was built to answer whether the drug is safe enough in an East Asian population, not how well it works, and the open-label design without a comparator means you cannot read efficacy from it. The finding to carry into clinic is the tolerability gradient: roughly one in seven stopped at 36 mg against one in twenty at 3 mg. That matters directly for Indian practice, where the population is younger, leaner at any given level of risk, and where an oral agent would reach many more people than an injectable one. When orforglipron arrives, the titration schedule and the counselling that goes with it will decide who stays on it.
- Ask about nausea, vomiting and early satiety at every visit during dose escalation, rather than waiting for the patient to raise it.
- Record the dose at which gastrointestinal symptoms first appear; that number predicts whether a patient stays on treatment.
- Do not treat the highest dose as the target dose - discontinuation was roughly threefold higher at 36 mg than at 3 mg.
- Background therapy here was diet and exercise or one to two oral agents, so the low hypoglycaemia rates say little about patients already on insulin or a sulphonylurea.
- Remember this is a daily tablet; adherence, not weekly convenience, is the likely failure mode.
The statistics, in plain English
This was a safety trial with three active doses and no placebo, so nothing here can be attributed to the drug with certainty; some of the 85% adverse event rate is simply what happens to 401 people over a year. The discontinuation confidence intervals overlap between 12 mg (4.6 to 14.0%) and 36 mg (9.3 to 21.1%), so the dose-response in tolerability is suggestive rather than proven, even though the direction is consistent. Open-label designs inflate the reporting of subjective symptoms such as nausea, because both patient and doctor know what has been given. And 401 people followed for 52 weeks is far too small to detect uncommon harms; a problem occurring in 1 in 1,000 users would very likely not appear once in this trial.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for diabetes & endocrinology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free