Another Alzheimer's Setback

The late-stage clinical trial failure of Eli Lilly's Alzheimer's drug semagacestat does not necessarily mean the end of the line for the 'amyloid hypothesis', which holds that the best way of tackling Alzheimer's is by blocking the formation of the beta amyloid plaques that are pathological hallmarks of the disease. There remain far too many unanswered questions, both about the trial itself and about the wider drug mechanism, for any definitive conclusions to be drawn at this point.

But this latest setback - it follows several other late-stage recent failures, including Flurizan (flurbiprofen; Myriad Genetics) Alzhemed (tramiprosate; Bellus Health) and Dimebon (latrepirdine; Pfizer/Medivation) - further ups the ante on two antibody development programmes that also target beta amyloid, bapineuzumab (Elan, Johnson & Johnson, Pfizer) and solaneuzumab (Lilly). The list of failures is growing, while success remains frustratingly elusive.

Semagacestat works by inhibiting gamma secretase, which is responsible for processing amyloid precursor protein into beta amyloid, the abnormal deposition of which is thought by many to play a key role in the development of Alzheimer's. Bristol-Myers Squibb is also developing a gamma secretase inhibitor, BMS-708163, which is undergoing phase 2 trials.

Data from these studies could help to answer some questions about the Lilly trial. It is unclear whether a class effect is at work or whether semagacestat itself is the problem. Clinical evidence in support of the compound was thin on the ground before it entered phase 3 trials. Lilly relied on biomarker evidence rather than any real indication that it actually worked in patients.

A year ago, one analyst told this writer that the trial wasn't expected to demonstrate efficacy, as the dose was too low. Safety was the main concern. That the drug actually proved toxic for patients bears out this suspicion.

Inhibiting gamma secretase can interfere with notch signalling, which is centrally involved in many cellular communication processes. The increased risk of skin cancer, which Lilly observed in patients receiving the drug, may be linked to disruptions in the notch pathway. One recent review of the topic (Panelos J. & Massi D., Cancer Biol. Ther., 8, 1986-1993, 2009) is baldly titled Emerging role of Notch signaling in epidermal differentiation and skin cancer.

The over-riding problem with nearly every late-stage Alzheimer's drug trial underway at present - certainly those that target beta amyloid formation and deposition - is that blocking beta amyloid formation and/or deposition may be too late for the mild-to-moderate patients who are usually recruited onto drug trials. The damage may already be done by the time they have been diagnosed.

Earlier recruitment - assisted by better biomarkers - could help. But the failure of semagacestat and other drugs shows that our biological understanding of the condition remains hugely incomplete.