Alzheimer’s disease affects around 55 million people worldwide. It has been studied intensively for decades. And yet, until 2021, not a single drug had been approved that slowed its progression rather than just managing symptoms.
Understanding why is a study in the complexity of the brain and the limits of the disease models we’ve been working with.
The amyloid hypothesis
For most of the last 30 years, the dominant theory was the amyloid cascade hypothesis. It proposed that Alzheimer’s is caused by the accumulation of amyloid-beta plaques in the brain — abnormal protein fragments that aggregate between neurons, triggering a chain of events leading to neuronal death. The logical conclusion: remove the plaques, stop the disease.
Drug after drug targeted amyloid. Phase III trials of antibodies that cleared plaques effectively from the brain produced disappointing results — in many cases, clearing amyloid didn’t produce meaningful cognitive benefit. Patients were still declining. The plaques were gone, but something else was driving the damage.
What this revealed
Several explanations have been proposed. One is that by the time patients enter trials — with measurable cognitive symptoms — the disease has already caused irreversible neuronal loss. Amyloid may be necessary but not sufficient; downstream processes involving tau protein tangles, neuroinflammation, and synaptic loss may be doing more damage once the cascade has started.
Another possibility is that the hypothesis itself is incomplete. Amyloid-beta may be a marker of something upstream rather than the primary cause. The role of microglia (the brain’s immune cells), mitochondrial dysfunction, and vascular factors in Alzheimer’s pathology is increasingly recognised and historically underweighted.
Recent developments
Lecanemab, approved by the FDA in 2023 and subsequently by the MHRA in the UK, is an anti-amyloid antibody that showed modest but statistically significant slowing of cognitive decline in early-stage disease. The effect size is small — roughly a 27% slowing over 18 months — and the drug carries a risk of brain swelling and microbleeds. Whether the benefit justifies the risk and cost in a broad patient population remains debated.
It is, nonetheless, the first drug to slow Alzheimer’s progression. After decades of failure, even a modest effect represents a real shift.
Why this matters
The Alzheimer’s story is instructive for anyone thinking about medicine as a career. It demonstrates that a plausible, well-supported biological hypothesis can be wrong, or at least incomplete. It shows how difficult it is to develop drugs for the brain, where the blood-brain barrier limits what compounds can be used and where the only reliable outcome measure is cognitive function, which is hard to assess and slow to change. And it reveals the real cost of incomplete models — not just in financial terms, but in the years of patients’ lives spent in trials that didn’t help them.