While there is still more to be learned, amyloid-β (Aβ) appears to play a role in subsequent tau tangle formation, neuronal loss and dysfunction, and eventual cognitive decline in Alzheimer’s disease. This has led to the development of treatments for the disease that directly target Aβ and associated proteins.

The development of the Aβ protein is the earliest seminal event that leads to the pathology of Alzheimer’s dementia. Interestingly, the amount of Aβ in a person’s brain does not correlate with clinical progression very well, but it turns out that the spread of tau tangles does. This was established by Braak and colleagues in their 1991 paper on the neuropathological staging of Alzheimer’s disease, showing that the spread of tau correlates very well with clinical progression. Some people even speculate that the reason we see clinical efficacy signals with some Aβ-targeting therapies is due to a secondary effect of lowering tau.

As the formation of Aβ is thought to be the primary step in the development of Alzheimer’s disease, the development of amyloid-targeting therapies has been investigated for many years as potential treatments. We have learned a lot of things as we have developed these Aβ-targeting monoclonal antibodies. In some of the first-generation trials, patients were enrolled based on a clinical diagnosis of mild cognitive impairment or mild Alzheimer’s dementia. It turns out that some of these patients did not actually have amyloid in their brain, which, of course, raises concerns that you are giving an anti-amyloid treatment to a patient who does not even have amyloid in their brain. So, now the new trials screen patients for the presence of biomarkers, Aβ on positron emission tomography (PET) or cerebrospinal fluid, or an APOE genotype that increases the risk for Alzheimer’s disease.

We have also learned that we should not be afraid of amyloid-related imaging abnormalities and that we should not treat them with therapies at doses that may be too low, as we were doing in earlier trials. Instead, we should go higher on the dose to remove amyloids more aggressively. Targeting some specific amyloid species may not make much sense. Solanezumab and crenezumab targeted the monomeric and oligomeric Aβ species, which was probably not their best approach. It turns out that more advanced species such as protofibrils should be the preferred targets.

You also have to remove amyloids quickly and aggressively, as we have seen with lecanemab. Clinical trials with lecanemab demonstrate a significant removal of Aβ within 12 to 18 months. The titration schedules for aducanumab and gantenerumab were very slow, which is probably why we did not see efficacy. Finally, we have also learned that you have to reduce amyloid levels on PET to less than 25 CL (quantitative measure of amyloid PET), which is what lecanemab demonstrated in its phase 3 trial. The third-generation investigational drugs may remove amyloids even faster, more aggressively, and to lower levels.

It has been nearly 18 years since the first amyloid-targeting therapies were developed as treatments for Alzheimer’s disease, and, as stated, we have learned a lot since then. The third-generation drugs will benefit from everything we have learned, so the modern trials are very different than trials from the early 2000s.