ATTAIN-OSA

Presented in LBS.05 at AHA 2026 (Monday 9 November) by Vaishnavi Kundel (Mount Sinai) and published simultaneously in JAMA (first author Atul Malhotra). ATTAIN-OSA comprises two 52-week, phase 3, double-blind, placebo-controlled trials at 66 sites in nine countries (November 2024 to July 2025, final follow-up July 2026) in adults with moderate to severe obstructive sleep apnoea and BMI 27 or higher, without diabetes: trial 1 (363 patients, no positive airway pressure; mean age 48.4, BMI 37.4, AHI 50.7, 31.4 percent women) and trial 2 (349 patients on PAP; mean age 51.1, BMI 38.6, AHI 57.6, 26.9 percent women). Randomised 1:1 to once-daily orforglipron (maximum tolerated dose 9, 14.5 or 17.2 mg) or placebo. Primary endpoint: change in AHI at week 52. Treatment-regimen estimand: trial 1 minus 17.8 versus minus 4.5 events per hour (difference minus 13.2, 95 percent CI minus 17.7 to minus 8.7, P less than 0.001); trial 2 minus 17.3 versus minus 5.3 (difference minus 12.0, minus 17.1 to minus 6.9, P less than 0.001). Six of seven key secondary endpoints significant. Per the AHA release: hypoxic burden minus 51 versus minus 18 percent (trial 1) and minus 44 versus minus 18 percent (trial 2); weight minus 10 to 11 versus minus 1 to 2 percent; hs-CRP minus 42 versus minus 4 and minus 38 versus minus 18 percent; systolic BP minus 7.5 versus minus 1.7 and minus 7.0 versus minus 2.2 mmHg (measured at week 48). Gastrointestinal adverse events 68.0 versus 37.4 percent and 70.6 versus 36.6 percent. The first large trials of an oral drug for OSA; not a replacement for CPAP, per the investigators.