I have been following the story here of the newest Alzheimers drugs, the first to show that they can actually slow the progression of Alzheimer’s disease (AD). The benefits are modest, and come with the potential for serious side effects and a high price tag, but after decades of disappointment it was good to at least have a proof of concept that some application of our understanding of AD can help.
But now a Cochrane review casts doubt on the efficacy of these drugs. It is being widely reported as showing the drugs do not work, and is being used to justify not paying for treatments. The BBC, for example, reports, “Breakthrough £90,000 Alzheimer’s drugs unlikely to benefit patients, report suggests.”
The authors of the review looked at 17 studies involving over 20 thousand patients who were followed for at least 18 months. They concluded:
“The effect of amyloid‐beta‐targeting monoclonal antibodies on cognitive function and dementia severity at 18 months in people with mild cognitive impairment or mild dementia due to Alzheimer’s disease is trivial, while on functional ability, it is small at best. Amyloid‐beta‐targeting monoclonal antibodies increase the risk of amyloid‐related imaging abnormalities. Both desirable outcomes and adverse events were inconsistently reported in the studies included in the review.
Successful removal of amyloid from the brain does not seem to be associated with clinically meaningful effects in people with mild cognitive impairment or mild dementia due to Alzheimer’s disease. Future research on disease‐modifying treatments for Alzheimer’s disease should focus on other mechanisms of action.”
That is pretty devastating. But the fact that they looked at 17 studies was an immediate red flag for me – I didn’t think there were 17 studies of the currently two approved drugs. There isn’t. The authors looked at every monoclonal antibody that targets amyloid, including those that have previously been shown not to work. They included aducanumab, bapineuzumab, crenezumab, donanemab, gantenerumab, lecanemab, ponezumab, remternetug, and solanezumab (the “mab” stands for monoclonal antibody).
These are all drugs that are antibodies which target some form of amyloid protein, which is the protein that builds up in neuron in AD and is a significant marker for the disease. The longstanding question has been, however, is amyloid driving AD or is it just a downstream effect? If the latter than removing it might do nothing to alter the course of the disease or treat its symptoms.
That is really what is at the core of this evidence – is amyloid the target we should be focusing on? Or is AD more of an inflammatory disease? Inflammation is something else that can be primary (causing the disease) or secondary (just cleaning up the damage caused by the disease). There is also tau protein, which can also build up in some patients with AD. In fact the research is progressing to distinguishing patients in terms of how much amyloid vs tau they have. We can also distinguish patients who have familial forms of AD from those who don’t, so AD is really a category of multiple specific diseases.
It’s important to understand that the precise role of amyloid in AD vs other potential mechanisms is a decades-long heated debate among AD researchers. After decades of negative clinical trials, many researchers were pushing for the conclusion that amyloid is simply a dead end therapeutically. Then lecanemab and donanemab showed clear (but modest) clinical benefit, and it was the turn of the amyloid theory promoters to do their victory dance. Even though the clinical benefits might be small, it was taken as a proof of concept – the amyloid hypothesis must be onto something.
In this context, this latest review might be seen as a counterattack by the anti-amyloid brigade. That might explain why the researcher chose to include all anti-amyloid monoclonal antibodies, not just the ones that have been previously shown to work. It also explains their rather pointed conclusion – “Future research on disease‐modifying treatments for Alzheimer’s disease should focus on other mechanisms of action.” Should it, really?
Amyloid researchers are not taking this sitting down. They are quick to point out that this review does not clarify the data, it deliberately muddies it by mixing in older failed drugs with newer successful ones. In essence, critics claim, they have deliberately diluted out the evidence for efficacy then falsely claimed the entire class of drugs is a failure, and used that to deny the underlying mechanism.
I tend to be persuaded by this argument – I immediately found it extremely odd that they would do a review that combines drugs which don’t work with drugs that do and treat them as one class.
I think this 2025 review takes a more nuanced and appropriate clinical approach to this evidence. They found:
“Anti-amyloid mAbs consistently demonstrated robust amyloid clearance and modest slowing of clinical decline in early symptomatic AD. Differences emerged across agents in efficacy signals, safety profiles, and regulatory outcomes. Lecanemab and donanemab showed more consistent cognitive benefits, while aducanumab yielded mixed findings, leading to its withdrawal.”
and concluded…
“Anti-amyloid mAbs provide proof of concept for AD modification, with the greatest benefit in early disease stages and moderate tau burden. Optimal use requires biomarker confirmation of the amyloid, careful tau staging, and genetic risk assessment. While limitations remain, these therapies represent a pivotal step toward precision neurology and may serve as a foundation for multimodal strategies targeting tau, neuroinflammation, and vascular pathology.”
Exactly. I have heard several grand rounds by leading AD researchers, and this conclusion seems spot on to what they are saying. AD is a complex set of diseases, which is why it has proven so difficult to treat. No simple approach is likely going to have any significant impact. It took decades to develop treatments targeting amyloid that show clinical benefit, partly because monoclonal antibody technology has allowed more precise targeting and higher doses. They are just the beginning, however.
Going forward we need to do several things. First, continue to advance our screening efforts and ability to diagnose AD early, because the earlier the treatment the more effective it is, and this can be a massive factor. Second, we need to separate AD into various subtypes – as the above study concludes, we need to confirm the role of amyloid and determine how much of a role tau is likely playing also. We need to combine this with genetic data and family history. Together we can then target a suite of specific treatments towards individual patients. But of course, we have to develop the individual treatments first, and the anti-amyloid mabs are just the first step.
Finally, we need to continue to explore other mechanisms at work in AD and target them as well, not just amyloid and tau, but inflammation and vascular pathology. Perhaps in 20-30 years, or even sooner, we will have a cocktail of treatments, each of which individually has a modest benefit, but together have a significant benefit for slowing down AD. Since AD tends to present in older age, slowing down progression can mean that many patients die with AD rather than from AD, and while still highly functional.
In the end, I am not even sure what role this latest review plays. By mixing older drugs that don’t work with newer ones that do it seems designed to muddy the waters, and to have a negative outcome. This is likely just to confuse the public, and is being used by regulators to justify not paying for these newer treatments.
