Twenty-four days have passed since Cochrane published its meta-analysis of anti-amyloid antibodies for Alzheimer's disease. The conclusion — that the drugs produce "absent or trivial" effects on cognitive decline — has hardened into a public methodology dispute with disclosed institutional conflicts on both sides. [1] What has not changed in those 24 days is the regulatory clock. NICE's final draft guidance, published in March 2026 and standing as the operative document, recommends that donanemab not be made available within its marketing authorisation on the NHS. [2] The April 28 closing date for comments has passed. The fourth committee meeting, scheduled for June 10, is the document that produces the answer. The paper's May 7 feature on the institutional-conflict architecture named the June meeting as the decision that counts. Today the meeting sits 33 days out.

The pushback this week ran along familiar institutional lines. Sir John Hardy of Alzheimer's Research UK and University College London — the originator of the amyloid hypothesis — has continued to call the Cochrane review "a silly paper which should not have been published," disclosing consulting relationships with Eli Lilly, Biogen, and Eisai. [3] Bart De Strooper of the UK Dementia Research Institute has held to the methodological critique that pooling 17 trials across seven anti-amyloid drugs — five of which failed and were discontinued — turns therapeutic progress into statistical noise. De Strooper has also disclosed consulting income from Eisai, Eli Lilly, and Johnson & Johnson, plus founder positions in two spin-off biotechs. [4] The Cochrane review's lead authors — Francesco Nonino at the IRCCS Institute of Neurological Sciences of Bologna and Edo Richard at Radboud University Medical Centre — have not held comparable industry consulting income.

What NICE has on the table is a regulator-versus-evidence axis split three ways. The FDA approved donanemab (Kisunla) for early Alzheimer's in July 2024 and lecanemab (Leqembi) earlier still. Health Canada approved donanemab in November 2024 with restrictions. The European Medicines Agency reversed an initial negative opinion and approved both drugs in 2025. The MHRA in the UK approved donanemab in October 2024 and lecanemab in August 2024. The Cochrane review is not a regulatory document; it is a meta-analysis that NICE will read alongside the manufacturer submissions, the Alzheimer's Research UK and Alzheimer's Society public-comment record, and its own committee's prior reasoning. NICE's final draft has already concluded the cost-effectiveness math does not work at the prices Eli Lilly and Eisai have submitted. [2]

The cost-effectiveness math is the artifact the June meeting produces. NICE concluded that the treatments slow progression from mild to moderate Alzheimer's by four to six months, but that the cost of purchasing and administering the drugs (regular infusions, MRI monitoring for amyloid-related imaging abnormalities, hospital infrastructure) far exceeds the £20,000-£30,000 willingness-to-pay threshold per quality-adjusted life year. [2] The committee's framing in the final draft was direct: "the small benefits of donanemab and lecanemab shown in the clinical trials and the lack of long-term evidence of effectiveness, together with the substantial resources the NHS would need to commit to the treatments means if they were approved they could displace other essential treatments and services that deliver significant benefits to patients." [2]

The pricing question is the only one that moves the math. Eli Lilly and Eisai have submitted confidential discount agreements through NICE's Patient Access Scheme; whether the discounts are deep enough to bring the cost-per-QALY below the threshold is the June 10 question. The Cochrane review's effect on the math is indirect but consequential: if the committee accepts the Cochrane reading that the absolute treatment effect is "absent or trivial" — falling below the minimum clinically important difference — then no price discount changes the conclusion, because the QALY denominator collapses. The Hardy-De Strooper reading preserves the four-to-six-month delay as a clinically meaningful effect; that reading is what keeps the price-discount path viable.

The lost-science register is what makes the day-24 read load-bearing. The 2025 Alzheimer's Disease Centers funding cut, the loss of Pacific Wildland Fire Sciences Lab equivalents in the dementia-research infrastructure, the decade of failed phase-3 trials Cochrane pooled into the meta-analysis: all of this sits behind the June 10 meeting as the structural backdrop. An FDA-approved drug, a Health Canada-approved drug, and a Cochrane meta-analysis are now splitting the regulator-evidence axis on the same week the U.S. paper carries the GLP-1 mental-health cohort onto the FDA desk. Both stories are versions of the same problem: regulators with approved drugs, observational data of varying quality, and price-and-coverage decisions that do not wait for randomized trials to settle the methodological dispute.

What this paper now tracks: whether the May Cochrane meta-analysis appears publicly before the June 10 meeting (versus existing only as the published April document), and whether any lecanemab- or donanemab-trial principal investigator files a comment for the closing record that surfaces previously unpublished long-term-extension data. The Alzheimer's Society's middle-position framing — "anti-amyloid drugs still to be a valid avenue to explore and optimise" but "not a silver bullet" — is the institutional voice NICE has weighed before. [5] The June 10 meeting is when the weighing produces an answer.

-- KENJI NAKAMURA, Tokyo