Merck has stopped a phase 3 trial of BACE1 inhibitor verubecestat in prodromal Alzheimer’s patients following an interim review. News of the setback comes 12 months after Merck pulled the plug on a trial in patients with more advanced forms of the disease after seeing similarly lackluster data.
Work on the prodromal trial continued following the earlier flop in the hope patients who were yet to develop full-blown Alzheimer’s might respond better to the drug. Merck is yet to share data from the second phase 3 trial but the take-home message from the early stoppage is clear: Verubecestat is as ineffective in prodromal Alzheimer’s as it is in mild to moderate forms of the disease.
The data hammer another nail into the coffin into the amyloid hypothesis. Researchers identified the BACE1 cleaving enzyme as an Alzheimer’s target after linking it to the generation of the beta-amyloid peptide found in the brains of patients with the disease. Inhibiting BACE1 was supposed to suppress beta-amyloid production and thereby improve cognitive and functional performance.
Merck’s clinical development program linked verubecestat to lower levels of beta-amyloid. But in two phase 3 trials that enrolled thousands of patients it failed to show the drug also improves outcomes. The phase 3 failures sees verubecestat join the list of BACE drugs that have stumbled in the clinic. Eli Lilly and Roche both scrapped BACE1 inhibitors in 2013.
Other companies are still pursuing the idea, though. Lilly got back into BACE in 2014 by buying into AstraZeneca’s asset. That drug is now in phase 3, as is Biogen and Eisai’s E2609. The trials are assessing the inhibitors in patients with early forms of Alzheimer’s, a population in which other BACE drugs have already failed.
Amgen and Novartis have taken a different approach for their BACE collaboration by going back even further in the disease progression pathway. The partners are testing BACE inhibitor CNP520 in people who are healthy but at risk of Alzheimer’s based on their age, genotype and amyloid levels.