The amyloid hypothesis, set out by Hardy and Higgins in 1992, holds that clearing amyloid-beta from the brain would slow Alzheimer’s disease, and many drug programs have tested it. The record since 2014 is mixed. Two antibodies, bapineuzumab and solanezumab, failed in large Phase 3 trials. Biogen’s aducanumab won an accelerated approval in 2021 and was discontinued in 2023. Then lecanemab and donanemab each slowed decline modestly in Phase 3 trials.
So the record shown here does not settle whether the amyloid idea is right or wrong. In these papers the answer varied with the form of amyloid a drug targeted, the patients, and the stage of disease. This page sets out what the papers and Biogen’s filings say. Every figure was read from the cited source. No dollar spend is stated, because none of the sources read gives one.
| 1984 | Glenner and Wong report purifying a novel cerebrovascular amyloid protein from Alzheimer’s disease, and suggest it may help in understanding the disease’s pathogenesis. | Glenner & Wong, BBRC 1984, PMID 6375662 |
| 1992 | Hardy and Higgins publish a two-page article titled “Alzheimer’s disease: the amyloid cascade hypothesis.” | Hardy & Higgins, Science 1992, PMID 1566067 |
| 2004 | Klunk and colleagues report imaging brain amyloid in people with Alzheimer’s disease using Pittsburgh Compound-B. | Klunk et al., Ann Neurol 2004, PMID 14991808 |
| 2014 | Two Phase 3 trials of bapineuzumab, an anti-amyloid antibody, in mild-to-moderate Alzheimer’s (1,121 APOE ε4 carriers and 1,331 non-carriers). The authors conclude it did not improve clinical outcomes, despite biomarker differences in carriers. Brain swelling (amyloid-related imaging abnormalities) increased with dose. Funded by Janssen Alzheimer Immunotherapy and Pfizer. | Bapineuzumab Phase 3, NEJM 2014, PMID 24450891 |
| 2014 | Two Phase 3 trials of solanezumab (EXPEDITION 1 and 2; 1,012 and 1,040 patients) in mild-to-moderate Alzheimer’s. Solanezumab preferentially binds soluble forms of amyloid. It failed to improve cognition or functional ability. Funded by Eli Lilly. | Doody et al., NEJM 2014, PMID 24450890 |
| 2018 | EXPEDITION 3 tests solanezumab in 2,129 people with mild dementia and confirmed amyloid. The cognitive score difference was −0.80 points (95% CI −1.73 to 0.14; P = 0.10), not significant. Funded by Eli Lilly. | Honig et al., NEJM 2018, PMID 29365294 |
| Jun 2021 | The FDA grants accelerated approval to Biogen and Eisai’s ADUHELM (aducanumab) based on reduction in amyloid beta plaques, with a confirmatory trial required to verify clinical benefit. In December 2021 the European regulator’s committee adopts a negative opinion. | Biogen 10-K, fiscal 2021 |
| Nov 2022 | Lecanemab, an antibody that binds soluble amyloid protofibrils, is tested in 1,795 people with early Alzheimer’s (Clarity AD, 18 months). The disease-severity score (CDR-SB) worsened by 1.21 on lecanemab versus 1.66 on placebo, a difference of −0.45 (95% CI −0.67 to −0.23). The authors describe moderately less decline, with adverse events, and call for longer trials. Funded by Eisai and Biogen. | van Dyck et al., NEJM 2023, PMID 36449413 |
| Jul 2023 | The A4 trial tests solanezumab in people with preclinical Alzheimer’s (240 weeks). The cognitive composite difference was −0.30 (95% CI −0.82 to 0.22; P = 0.26). Brain amyloid rose by 11.6 centiloids on solanezumab and 19.3 on placebo. The authors conclude it did not slow cognitive decline. | Sperling et al., NEJM 2023, PMID 37458272 |
| Aug 2023 | Donanemab, an antibody designed to clear amyloid plaque, is tested in 1,736 people with early symptomatic Alzheimer’s (TRAILBLAZER-ALZ 2, 76 weeks). Of 24 gated outcomes, 23 were statistically significant; the iADRS difference was 3.25 points (95% CI 1.88 to 4.62) versus placebo. The authors conclude it slowed clinical progression in people with low or medium tau and in the combined group. | Sims et al., JAMA 2023, PMID 37459141 |
| Sep–Nov 2023 | Biogen discontinues the EMBARK study of aducanumab in September 2023, and in November 2023 notifies Neurimmune of its decision to discontinue development and commercialization of ADUHELM. | Biogen 10-K, fiscal 2023 |
The counts cover only the programs named on this page. Other Alzheimer’s programs exist and were not reviewed.
None of these is a finding that a company’s decision was wrong. The method behind them, tracing each load-bearing claim to its primary source and tagging each link supported, partial, contradicted or missing, is described on the main page.
Hindsight is cheap. What an audit can do is show, before the money moves, which claims rest on a marker, which on a trial, and which on a drug-specific result.