The Cochrane review that found anti-amyloid Alzheimer's drugs make little to no difference pooled seventeen trials, and fifteen tested drugs that are not the two now prescribed. The average buries the two survivors.
Alone, those two do something small, real, and costly. Lecanemab slowed decline by 0.45 of a point on an eighteen-point scale over eighteen months, about six months deferred, with brain swelling in about 126 of every 1,000 treated and microbleeds in about 173. Donanemab slowed decline by 3.25 points on a 144-point scale, with three deaths in its trial tied to the swelling-and-bleeding effect.
Britain licensed both in 2024, but NICE will not fund them, at five to six times its usual price per quality-adjusted life year.
Half a point is not nothing. It is not much either. It buys months, not years, with a scan schedule, a gene test, and a bill.
