Conference Coverage

Unblinded data show extent of verubecestat’s failure in mild-moderate Alzheimer’s



– The Alzheimer’s research community absorbed yet another downer recently, when Merck scientists revealed the unblinded efficacy and safety data of EPOCH, the company’s failed verubecestat trial: The BACE inhibitor didn’t score in any endpoint, no matter how the data were sliced and diced.

Merck halted the study last February, when an interim analysis determined there was no chance of a positive outcome. No safety data played into the decision, officials said. At the time of discontinuation, Merck had not yet examined the unblinded data, which were released to a packed audience at the Clinical Trials on Alzheimer’s Disease conference in Boston.

Compared with placebo, the nonselective beta-secretase (BACE) inhibitor conferred no cognitive or functional benefit upon patients with mild-moderate Alzheimer’s disease, either in the overall analysis or in any age, disease stage, or genetic subgroup, Michael Egan, MD, said during a panel discussion. And although there was plenty of biomarker evidence that the drug did block beta amyloid production, there was also a plethora of concerning adverse events.

Dr. Michael Egan

Dr. Michael Egan

The complete lack of response at both doses tested (12 mg and 40 mg) is a very strong signal that BACE inhibition in patients with mild-moderate disease is a dead end. “Turning off Abeta [amyloid beta] production with a BACE inhibitor in patients at this stage of the disease is not helpful,” Dr. Egan said.

However, the failure of yet another antiamyloid drug doesn’t mean that researchers should abandon amyloid as a therapeutic target, said Dr. Egan, Merck’s associate vice president of clinical neuroscience. Verubecestat is still being investigated in the APECS study of patients with mild cognitive impairment, and a number of antiamyloid antibodies are still going forward in patients whose disease stages run from preclinical to moderate.

“It’s still possible that we may see a clinical benefit in some of these studies, so we have to keep an open mind,” Dr. Egan said.

EPOCH, a pivotal phase 2/3 trial, randomized about 1,200 patients with mild-moderate Alzheimer’s to either placebo or verubecestat 12 mg or 40 mg daily for 18 months. None of the patients had amyloid PET imaging, but in subsets of patients who had the imaging or lumbar puncture for Abeta levels, 90% were amyloid-positive. The primary efficacy outcomes were the change from baseline in the Alzheimer’s Disease Assessment Scale Cognitive Subscale (ADAS-Cog) score and the change from baseline in the Alzheimer’s Disease Cooperative Study–Activities of Daily Living (ADCS-ADL) score.

There were a number of secondary endpoints, including the Clinical Dementia Rating–Sum of Boxes (CDR-sb); total hippocampal volume; cerebrospinal fluid total and phosphorylated tau; changes on the Neuropsychiatric Index and Mini Mental State Exam; and brain amyloid burden.

In a nutshell, Dr. Egan said, there was virtually no efficacy signal on any of the primary or secondary endpoints. On the ADAS-Cog, all patients, regardless of treatment group, lost 7-8 points over the trial. The same was true for the ADCS-ADL, with everyone declining about 8 points over time. On the CDR-sb, all groups declined about 2 points.

“We also looked at a number of subgroups: younger subjects, those with earlier disease, and those who we knew were amyloid-positive, including all of our ApoE4 carriers. We found no evidence of efficacy in any subgroup.”

It is “worth noting,” Dr. Egan said, that in the amyloid PET substudy, there were treatment-related reductions in plaque burden. While those taking placebo showed no changes in brain amyloid, the 12-mg group experienced a mean 2% reduction in amyloid, and the 40-mg group, a 4% reduction. “While this is modest, it does show target engagement,” Dr. Egan said – an important and positive finding in light of the ongoing APECS study.

The study of cerebrospinal fluid showed no effect on either tau protein, but marked, dose-related reductions in Abeta1-40 and soluble amyloid precursor protein – both products of BACE cleavage of the amyloid precursor protein. “We saw a 70% reduction in the 12-mg group and more than an 80% reduction in the 40-mg group, but no change in the placebo group. This is an important finding, demonstrating that the drug got into the brain and turned off production of Abeta. EPOCH is the first phase 3 study of an antiamyloid agent where target engagement of this magnitude has been demonstrated.”

Verubecestat also had its share of adverse events, Dr. Egan said. The most common was rash, which developed in about 10%; 20% of those who developed rash discontinued treatment for that reason. More concerning were falls and injuries; diarrhea and weight loss; and a variety of neuropsychiatric events, including insomnia and sleep disorders, anxiety, depression, and suicidal ideation.*

The drug was also associated with more loss of hippocampal volume, compared with placebo (5.7% vs. 5%), Dr. Egan said in an interview. The etiology isn’t clear; he suggested that it could be related to amyloid plaque removal, resolution of neuroinflammation, or an actual worsening of neurodegeneration. “That is a concerning possibility, although if that were the case we would expect to see worsening cognition, which we did not.”

Falls and injuries occurred in 15% of the placebo group and 20%-23% of the active groups. A detailed analysis didn’t turn up any specific risk factors, though. The episodes of suicidal ideation were passive and more common in the first 6 months of treatment and among patients who had a history of depression or prior suicidal ideation. Four patients discontinued due to that side effect.

Verubecestat is a nonselective inhibitor of both BACE1 and BACE2, and it’s not clear if that wide-ranging inhibition increased the likelihood of adverse events over what might be seen with a more selective compound. “It’s difficult to attribute them to BACE2 over BACE1,” Dr. Egan said. “Any BACE inhibitor could potentially have similar side effects.”

Only time will provide those answers; BACE inhibition is an area of active investigation among several large companies. The newly announced Generation studies will test a selective BACE1 inhibitor called CNP520.

Eli Lilly is recruiting for a phase 2 study of its BACE1 inhibitor, dubbed LY3202626. AstraZeneca is also looking at BACE1 inhibition with its candidate, lanabecestat.

Dr. Egan remains hopeful, though, and said that Merck retains its commitment to bringing an effective Alzheimer’s treatment to market.

“It’s natural to get discouraged with negative trials, and there certainly have been a lot of them. But I think we have to continue to work very hard to try and find something to help patients, and we have more and more knowledge every year about how to do that. I believe BACE inhibition continues to offer the possibility that if we treat earlier that there could be benefit, but for those with dementia, BACE inhibition is just too late.”

Dr. Egan is employed by Merck Sharp & Dohme, which sponsored EPOCH.

This article was updated 11/16/17.

Correction, 11/20/17: An earlier version of this article misstated the percentage of patients who experienced rash.

[email protected]

On Twitter @Alz_Gal

Next Article:

Amyloid imaging changed management for 80% of patients with uncertain dementia diagnosis

Related Articles