Another Alzheimer’s drug has failed a major test. Merck (NYSE: MRK) reported late Tuesday that its pill verubecestat was not helping people with significant Alzheimer’s symptoms. The drug maker announced that it halted its EPOCH study early because an oversight group said there was “virtually no chance of finding a positive clinical effect.”
Verubecestat is the latest in a long string of experimental Alzheimer’s drugs that have failed to meet study goals and help people whose symptoms are fairly advanced—known as “mild to moderate” Alzheimer’s. Most of these drugs have attacked the underlying disease by breaking up accumulations, or plaques, of the protein amyloid-beta, or by preventing the plaques from forming. Verubecestat was meant to be a preventative measure. Every failure of an amyloid-targeting drug adds fuel to skepticism about the prevailing theory of treatment: that attacking the protein, in whichever form it takes, will help Alzheimer’s patients.
“Expectations for this trial were low,” wrote ISI Evercore biopharma analyst Mark Schoenebaum in a research note Tuesday evening. He cited two factors: First, Merck did not test patients for amyloid-beta before enrolling them. That means people with other forms of dementia were likely included in the 2,200-person study, and verubecestat was not designed to help them. Second, the study aimed to treat people well into the course of the disease.
Stung by other failures in mild-to-moderate Alzheimer’s patients, drug makers and researchers have shifted hopes to patients who have amyloid in their brains, or high risk of Alzheimer’s based on their genetics, but whose symptoms are not as far along. Biogen (NASDAQ: BIIB) is making a huge bet that its drug aducanumab, now in Phase 3 studies due to provide data in 2019, will help people with early disease. But Eli Lilly (NYSE: LLY) said last fall that its drug solanezumab failed to deliver enough impact for patients with mild Alzheimer’s, and two months later it pulled the plug on a related study with even earlier-stage patients.
Merck said it would continue to test verubecestat in a separate study of prodromal Alzheimer’s patients, who have memory loss but can still live relatively normal lives. The prodromal study is dubbed APECS.
Because Alzheimer’s progresses so slowly, researchers are keen to find an underlying biological signal that could stand in for health improvements to show a drug is working. In heart disease, that surrogate signal is lower cholesterol; in HIV, it is viral load. But so far the ties between amyloid levels and changes in health have proven elusive. “A third of people with amyloid in their brains don’t get Alzheimer’s,” said USAgainstAlzheimer’s chairman George Vradenburg in an interview with Xconomy last week. “You can demonstrate a positive change in amyloid, but it’s not necessarily established that you’d have a positive clinical benefit.”
More major Alzheimer’s data are due later this year from Axovant Sciences (NASDAQ: AXON), which bought a discontinued drug from GlaxoSmithKline (NYSE: GSK) for $5 million with hopes it could help mild-to-moderate Alzheimer’s patients stave off cognitive decline. The drug, intepirdine, is being tested in a big Phase 3 study in combination with donepezil, one of the few cognitive boosters approved for Alzheimer’s. It works by interacting with the brain chemical serotonin, not by disrupting amyloid.
(Axovant’s trial and Merck’s EPOCH were two of 14 key trials to watch in 2017 that Xconomy flagged late last year.)
Verubecestat blocks a protein called beta-site amyloid precursor protein cleaving enzyme 1, or BACE1. When left unchecked, BACE acts like tiny scissors to create the amyloid-beta fragments that clump up into plaques. Earlier BACE inhibitors failed because of safety concerns. Verubecestat is part of a new generation that their makers hope have fewer side effects and change the course of the disease. The outside advisors monitoring EPOCH said that side effects were not part of the decision to stop the study, according to Merck, and recommended that APECS continue without changes.