Alzheimer's disease immunotherapy and the amyloid hypothesis: when aggregation obscures interpretation.

# Alzheimer's disease immunotherapy and the amyloid hypothesis: when aggregation obscures interpretation *The Lancet 2026* Comment Alzheimer’s disease immunotherapy and the amyloid hypothesis: when aggregation obscures interpretation A widely publicised Cochrane review published on clearance and a broadly similar slowing of clinical April 16, 2026, set out to synthesise evidence on the progression in its 2023 phase 3 randomised trial.8 efficacy and safety of amyloid β-targeting monoclonal Examination of side-effect rates, and ARIA-E in antibodies for Alzheimer’s disease.1 The authors particular, provides further evidence for the flaws in aggregated data from 17 randomised trials of multiple the conclusions of the Cochrane review. Although the amyloid immunotherapy agents involving more than authors find overall that there was a small increase in 20 000 participants, concluding that these therapies the risk of ARIA-E rates, this concatenates the extremely probably result in little to no difference in cognitive low rates seen in drugs that had little or no impact on function or dementia severity; small improvements in amyloid burden (eg, solanezumab and crenezumab had functional ability; and small increases in side-effects, no more ARIA-E than placebo9) with the higher, albeit notably amyloid-related imaging abnormalities usually manageable and largely asymptomatic, rates indicating oedema or effusions (ARIA-E), with little or (12·6% and 24%) seen in the phase 3 trials of lecanemab no difference in serious adverse events or death. These and donanemab.7,8 Interpreting these pooled estimates results and their overarching conclusion that “successful as representative of a uniform class effect risks obscuring removal of amyloid from the brain does not seem to be important differences between individual therapies. associated with clinically meaningful effects in people For drugs where statistically significant slowing of with mild cognitive impairment or mild dementia due cognitive or functional decline is determined, it is, of to Alzheimer’s disease”1 are, however, fundamentally course, important to consider whether any effects are undermined by pooling agents with markedly different, clinically meaningful. Although methodologies can be and in some cases negligible, target engagement. applied to try to quantify clinical meaningfulness, none The therapeutic premise—that removal of is perfect.10,11 Importantly, a sustained reduction in the amyloid β will modify disease trajectory—has driven the rate of decline, even if moderate, could delay key clinical development of successive generations of monoclonal transitions: loss of independence, escalation of support, antibody therapies over the last two decades. Although or need for institutional care. Some trials included in the these treatments all target amyloid β, they have Cochrane review1 report small but consistent benefits fundamental differences in their pharmacokinetics, in functional measures, particularly those reflecting dosing, and specific amyloid β target—which ranges more complex daily activities. The extent to which these from soluble monomeric forms through to fibrillar differences accumulate over time to produce ongoing plaques. These distinctions are not only pharmacological benefit remains uncertain. But for patients and caregivers, but are also reflected by, and predictive of, the biological disease slowing might be meaningful even if they fall and clinical effect seen with individual drugs.2,3 Early short of predefined, and inevitably arbitrary, thresholds. agents such as bapineuzumab,4 solanezumab,5 and This leads to the central point. Patients and families are crenezumab6 neither achieved any meaningful amyloid not interested in class effects. They are interested in the removal nor resulted in clinical benefits. By contrast, potential benefits, risks, burden, and, in some instances, in a 2023 pivotal phase 3 study, lecanemab, which costs of the specific therapies available. Patients, and preferentially targets amyloid protofibrils, reduced the the public, rate delaying progression and maintaining mean amyloid burden in treated patients to below the functional independence very highly and perceive threshold for being amyloid-positive on PET and slowed drastic reductions in quality of life as Alzheimer’s disease clinical decline by about 27% compared with placebo progresses from mild dementia to its most advanced (adjusted mean change from baseline of 1·21 vs 1·66 stages.12 Licensing authorities make judgements on on the Clinical Dementia Rating—Sum of Boxes scale) the benefits and risks of a specific drug on the basis of by 18 months.7 And donanemab, which targets fibrillar available evidence for that drug: the fact that lecanemab amyloid plaques, achieved similar or greater amyloid and donanemab have gained approval in multiple 1664 segamI ytteG aiv segamI tniopflaH Published Online April 23, 2026 https://doi.org/10.1016/ S0140-6736(26)00789-0 Comment countries and jurisdictions across the world10,13 reflects that and is a trustee of UK Dementia Research Institute and Worldwide Cancer these drugs are efficacious. The different views of payers Research, a scientific advisory board member of Dementia Discovery Fund, and an advisory board member of Alzheimer’s Disease Data Initiative. point to different evaluations of value—clinical, economic, *Nick C Fox, Susan Kohlhaas, Jonathan M Schott and societal—as well as different models of funding and n.fox@ucl.ac.uk systems readiness.9 Dementia Research Centre, UCL Queen Square Institute of Neurology, London Pooling the effects of agents with differing mechanisms WC1N 3BG, UK (NCF, JMS); UK Dementia Research Institute at UCL, London, UK or target engagement—many unsuccessful, some (NCF, JMS); Alzheimer’s Research UK, Granta Park, Cambridge, UK (SK) biologically inactive, and most abandoned—risks diluting 1 Nonino F, Minozzi S, Sambati L, et al. Amyloid-beta-targeting monoclonal antibodies for people with mild cognitive impairment or mild dementia both any clinical effects and any safety signals. The due to Alzheimer’s disease. Cochrane Database Syst Rev 2026; 4: CD016297. 2 Hardy J, Mummery C. An anti-amyloid therapy works for Alzheimer’s disease: resulting summary1 assumes a consistent class effect that why has it taken so long and what is next? Brain 2023; 146: 1240–42. is not borne out by the evidence, and in doing so risks 3 Karran E, De Strooper B. The amyloid hypothesis in Alzheimer disease: new insights from new therapeutics. Nat Rev Drug Discov 2022; 21: 306–18. obscuring real signals emerging from specific drugs. This 4 Salloway S, Sperling R, Fox NC, et al, and the Bapineuzumab 301 and Cochrane review negates neither the amyloid hypothesis 302 Clinical Trial Investigators. Two phase 3 trials of bapineuzumab in mild- to-moderate Alzheimer’s disease. N Engl J Med 2014; 370: 322–33. nor amyloid-targeting therapy as a whole: the question 5 Honig LS, Vellas B, Woodward M, et al. Trial of solanezumab for mild is no longer whether drugs can remove amyloid and slow dementia due to Alzheimer’s disease. N Engl J Med 2018; 378: 321–30. 6 Ostrowitzki S, Bittner T, Sink KM, et al. Evaluating the safety and efficacy of decline, but which individual drugs do so safely, cost- crenezumab vs placebo in adults with early Alzheimer disease: two phase 3 randomised placebo-controlled trials. JAMA Neurol 2022; 79: 1113–21. effectively, and in ways that are meaningful for patients. 7 van Dyck CH, Swanson CJ, Aisen P, et al. Lecanemab in early Alzheimer’s NCF reports consulting and speaker fees paid to his institution from Eisai, disease. N Engl J Med 2023; 388: 9–21. F. Hoffmann-La Roche, and Eli Lilly; advisory board fees paid to his institution 8 Sims JR, Zimmer JA, Evans CD, et al. Donanemab in early symptomatic from Biogen and F. Hoffmann-La Roche and donated to charity from AbbVie; a Alzheimer’s disease. JAMA 2023; 330: 512–27. speaker fee from Eisai donated to charity; and is a Member of the Research 9 Belder CRS, Schott JM, Fox NC. Preparing for disease-modifying therapies Strategy Council for Alzheimer’s Society (UK). JMS is the Chief Medical Officer for in Alzheimer’s disease. Lancet Neurol 2023; 22: 782–83. Alzheimer’s Research UK, and reports institutional grants from LifeArc 10 Fox N, Belder C, Ballard C, et al. Treatment for Alzheimer’s disease. Lancet Foundation, British Heart Foundation, Alzheimer’s Society, Alzheimer’s 2025; 406: 1408–23. Association, Alzheimer’s Research UK, and Medical Research Council; royalties 11 Petersen RC, Aisen PS, Andrews JS, et al. Expectations and clinical from Oxford University Press and Henry Stewart Talks; consulting, speaker, and meaningfulness of randomized controlled trials. Alzheimers Dement 2023; advisory board fees from Eli Lilly, Roche Pharmaceuticals, Alamar Biosciences, 19: 2730–36. and Receptive Bio; travel support from Alzheimer’s Association, Alzheimer’s 12 Zhang X, Aylwin M, Thomas D, Stott J, Crutch S, Fox NC. The quality of life Research UK, and Eli Lilly; and receives PET amyloid tracer from Alliance Medical of patients with dementia and their caregivers: important and yet for research. NCF and JMS acknowledge the support of the National Institute for inadequately assessed. Lancet Neurol 2025; 24: 906–07. Health and Care Research (NIHR) UCLH Biomedical Research Centre and the UK 13 Guizzaro L, Bałkowiec-Iskra E, Haberkamp M, et al. Balancing benefit and risk in Dementia Research Institute; JMS is an NIHR Senior Investigator. These funders early Alzheimer’s disease: the European Medicines Agency (EMA) assessment of lecanemab and donanemab. Lancet Reg Health Eur 2026; 63: 101644. did not have any role in the writing of this Comment or the decision to submit it for publication. SK reports institutional funding from Alzheimer’s Research UK Targeted advertising in generative artificial intelligence chatbots: a new public health risk OpenAI has announced plans to introduce advertising extends advertising infrastructures that have long Published Online within free and low-cost versions of ChatGPT, shaped social norms and consumption of health-harming April 2, 2026 https://doi.org/10.1016/ alongside voluntary safeguards including separation of products into conversational systems increasingly S0140-6736(26)00464-2 advertisements from responses, privacy protections, relied upon for therapeutic support, companionship, exclusion of users younger than 18 years, and limits on and sensitive advice.3 Extensive evidence shows that advertising around sensitive topics such as health.1,2 marketing of tobacco, alcohol, gambling, and unhealthy This shift was predictable, given the substantial losses foods increases consumption, contributes to obesity associated with capital-intensive artificial intelligence (AI) and non-communicable disease, and exacerbates models and the proven profitability of targeted advertising mental health vulnerabilities.4–6 Similarly, marketing of on digital platforms, and could signal a broader industry breastmilk substitutes undermines breastfeeding and pivot, with other providers likely to follow.2 maternal and child health outcomes.7 From a commercial determinants of health perspective, The Lancet Series on commercial determinants of this development warrants urgent scrutiny. The change health describes how expanding corporate power --- [PDF原文](https://sci-net.xyz/storage/7932541/cf36857e048884a20d6dc99e187dcf831097fee3e81f20d03387e248b7e81380/Alzheimer-s-disease-immunotherapy-and-the-amyloid-hypothesis-when-aggregation-obscures-interpretation.pdf) DOI: 10.1016/S0140-6736(26)00789-0