A lipid guideline moves to PREVENT and Lp(a), and thalassaemia gets its first oral disease-modifying drug

The edition in brief

Five findings for clinicians outside the specialties they come from. The 2025 VA/DoD lipid guideline, summarised in Annals, replaces pooled cohort equations with the PREVENT calculator for primary prevention, adds one-time lipoprotein(a) testing and selective coronary artery calcium scoring, recommends moderate-intensity statins for adults living with HIV even at low calculated ten-year risk, and shifts secondary prevention toward statin plus ezetimibe or a PCSK9 inhibitor in very high-risk disease. ENERGIZE-T, a phase 3 trial across 19 countries, randomised 258 adults with transfusion-dependent alpha or beta thalassaemia to oral mitapivat or placebo: 30% versus 13% achieved a transfusion reduction response, an adjusted difference of 18 percentage points (95% CI 8–27, p=0.0003), with 6% discontinuing for adverse events. CREST-H, a substudy of CREST-2, tested whether revascularising asymptomatic carotid stenosis improves cognition in patients with impaired cerebral haemodynamics: it does not, with an adjusted cognitive z-score difference of −0.15 SD (95% CI −0.54 to 0.24) and a non-significant interaction. A Ugandan phase 4 trial in 1,784 children aged 9–23 months found one-fifth and one-half doses of 17DD yellow fever vaccine non-inferior to the full dose for seroconversion at four weeks. An updated Annals systematic review of 38 randomised trials in 25,816 adults without diabetes reports placebo-subtracted weight loss of −14.8% for subcutaneous semaglutide, −19.0% for tirzepatide and −12.4% for oral orforglipron, with gastrointestinal adverse events in 76% versus 40% and no new safety signals.