STREAM

Presented in LBS.03 at AHA 2026 (Saturday 7 November) and published simultaneously in NEJM. STREAM randomised 1,880 adults aged 70 or over (mean 77.5, 46.9 percent women, 32.1 percent diabetes, 53.5 percent recruited by general practitioners) who took a statin for primary prevention for at least a year, had at least two chronic conditions and no atherosclerotic cardiovascular disease, at 24 sites in Switzerland, the Netherlands (LUMC, Leiden) and France between November 2021 and April 2025, to discontinuing (938) or continuing (942) lipid-lowering therapy. Open label, blinded adjudication. Primary outcome: non-fatal MI, non-fatal ischaemic stroke or death from any cause, analysed as restricted mean event-free survival time at 2 years: 96.5 versus 98.5 weeks, difference minus 2.0 weeks (95 percent CI minus 3.7 to minus 0.3), within the prespecified margin of minus 5.2 weeks, P less than 0.001 for noninferiority. Over the whole follow-up (median 103 weeks, up to 4 years) the primary outcome occurred in 12.7 percent (119) after discontinuation versus 9.2 percent (87) with continuation, HR 1.38 (1.05 to 1.81): MI 1.4 versus 0.6 percent, ischaemic stroke 1.9 versus 0.8 percent, death 8.2 versus 6.9 percent; one extra event per 40 patients who stopped over 2 years. LDL cholesterol at 12 months in a subsample of 235: 176.0 versus 99.4 mg/dL. No differences in quality of life, pain, falls or sarcopenia; serious adverse events 43.0 versus 39.8 percent. Crossover 131 versus 112 patients. After the trial the data and safety monitoring board advised the discontinuation group to discuss restarting. The noninferiority claim rests on a statistic that the hazard ratio contradicts; STAREE in the same age group showed benefit of continuing. Shared decision making, not a general stop.