EXOME

all the information, none of the junk | biotech • healthcare • life sciences

Alkermes Eyes FDA Nod After Retooled Study Yields Desired Results

Xconomy Boston — 

Alkermes believes it has done it: showed that patients who took its treatment for depression fared better than people who didn’t.

Based in Dublin and Waltham, MA, Alkermes (NASDAQ: ALKS) said Thursday afternoon that its once-a-day pill, ALKS 5461, met the primary aim of reducing symptoms of major depressive disorder in comparison to a placebo during the third portion of a Phase 3 trial. That’s a turnaround from results released in January, when the company said it failed to meet its primary goals in the first two parts of the Phase 3 trial.

The market responded positively, sending shares of Alkermes up 46.2 percent in after-hours trading to $64.35 apiece as of 4:30 p.m. in New York.

During a call with investors and analysts, Alkermes attributed the reversal to a change in its statistical analysis plan.

During the first two studies, which took place over a few weeks, Alkermes recorded the drug’s impact using a standardized test that measures symptoms of depression. It then picked results from a single week to measure against a baseline established at the start of the trials, the company said. However, in the third study, Alkermes took the results from multiple weeks and averaged them, using that average to compare to the baseline.

“The continuous measurement of efficacy across multiple time points is a more powerful and precise depiction across the drug’s efficacy profile,” Alkermes chief medical officer Elliot Ehrich said during the call.

The news means Alkermes plans to land a meeting with the FDA, in hopes of developing a plan to get ALKS 5461 on the road to approval. The data that Alkermes submits to the FDA from the three studies, which were dubbed the FORWARD program, will be key.

During the call, Ehrich said Alkermes determined that the second of the three studies—FORWARD-4— failed because the company examined data from a single week, rather than the average of multiple weeks, like it did in FORWARD-5. But Ehrich said FORWARD-4 still had “overall” strong supportive evidence of ALKS 5461 at a larger dose.

Alkermes will be able to pool data from those two studies, Ehrich said, which was part of its previously stated statistical plan. The studies, along with other data from the 1,500 patients it has worked with, provide “a robust body of evidence for data submission,” Ehrich said. (The other Phase 3 study wouldn’t be used in part because it had a high placebo response, he said.)

Alkermes had a unique design for the Phase 3 trial, which tested its experimental drug in a way that specifically set aside patients who benefit from the placebo effect, something that can cloud clinical trial data when studying depression. As Xconomy previously reported, that’s one reason depression is a notoriously difficult disease to test new drugs against. In these trials, some patients show improvement even when they’re not taking the drug being tested.

But in the first two parts of the Phase 3 trial, patients taking the drug didn’t fare better than the other patients, as determined by a standardized test that measures symptoms of depression. The placebo effect Alkermes hoped to stifle—using what’s known as sequential parallel comparison design, or SPCD—still cropped up, as Xconomy reported in January.

In this third part, the drug was delivered in two different doses: one high-dose regimen and one low-dose regimen. The high dosing of the drug had “significant reductions” of signs of depression compared to the placebo, while the low dose showed improvement but did not separate itself significantly, Alkermes said.

You can read more on the history of the trial’s design, which was originally devised more than a decade ago, in Xconomy’s story from December 2015.

ALKS 5461 is for people who don’t respond to initial depression treatment, such as selective serotonin reuptake inhibitors, and before they move on to more invasive therapies like deep brain stimulation.

Ben Fidler and Alex Lash contributed to this report.