Alkermes plots course to the FDA after its depression drug scores success in last-stand PhIII
Just after the start of this year, Alkermes was blasted by the news that the first two of three late-stage studies for its depression drug ALKS-5461 had failed. Analysts hit the red-alert button and the biotech’s shares $ALKS skidded down 44%, wiping out $4 billion of market cap as execs immediately pivoted and researchers scrambled to salvage the third Phase III by revamping the trial.
Today, Alkermes says that third trial scored a success. And even though it doesn’t measure up to standard FDA guidelines, which look for a positive readout from at least two of the three studies, CEO
Richard Pops plans to follow through with his earlier announced strategy of sounding out the agency to see if this drug warrants a green light now with the data in hand.
Said Elliot Ehrich, M.D., Chief Medical Officer of Alkermes:
“With these data now in hand, we will move forward rapidly to meet with the FDA to determine the appropriate next steps toward a regulatory submission for ALKS 5461, with a goal of bringing this important new medication to patients with MDD.”
Shares immediately rocketed up 49% on the news.
Alkermes posted top line data reflecting a success for the key dose in the last-stand study.
In the study, ALKS 5461 2mg/2mg met the prespecified primary endpoint of significantly reducing depression scores compared to placebo, as measured by 6-item Montgomery–Åsberg Depression Rating Scale (MADRS-6) scores (p=0.018). ALKS 5461 2mg/2mg also demonstrated statistically significant reductions in 10-item MADRS (MADRS-10) scores compared to placebo (p=0.026). The 1mg/1mg dose of ALKS 5461 showed improvement in depressive symptoms in the study, but did not separate significantly from placebo.
Pops never lowballed expectations on this drug, which is used as an adjunctive therapy for treatment resistant patients — a huge potential market. Just before the first round of bad news hit, he called the upcoming results in January a “seismic” event in a chat we had at JP Morgan. He wasn’t looking for the tremors, though, to bring down his stock price.
Some analysts, including Paul Matteis from Leerink, wrote the whole thing off in the wake of the January debacle. And while Alkermes managed to regain some of what it had lost that day, its shares are still far off their 52-week high.
Pops, though, started to rebuild the argument for this drug before the dustup had quieted down. The company cited a trend toward statistical significance and noted that a post hoc analysis ascertained that the entire 2mg/2mg dose group in one study—FORWARD-4—achieved the critical endpoint on the Montgomery–Åsberg Depression Rating Scale (MADRS-6) scores. Researchers went back to the drawing board, adding new patients and rejigging its statistical analysis plan for FORWARD-5. And now the plan is to combine what they gleaned from FORWARD-4 and 5 to make its case.
FORWARD-3 simply failed, says Alkermes, due to a high placebo response. Unlike Forward-4 and Forward-5, though, Forward-3 did not use what’s called a sequential parallel comparison design, or SPCD. In an SPCD study, the first round of placebo patients who don’t respond to the drug are re-randomized between the drug arm and the sugar pill, in order to quell the high placebo responses that have scuttled numerous other trials for depression.
Alkermes outlined its argument for an approval of their drug in a call with analysts on Thursday evening. The key to the turnabout started with their analysis of FORWARD-4, their second Phase III study, which failed. The reason it failed, they said, is that the study hinged on the depression scores at a specific time. To improve their odds, investigators shifted from a specific deadline to coming up with an average score from responses tracked over weeks for FORWARD-5.
Significantly, Alkermes made the change without getting any feedback from the FDA. Under careful questioning by analysts, they also noted that MDRS -6 scores have not been used as a primary endpoint for a registration study before, but felt that it reflects a response to the core symptoms of depression.
In a followup, Pops told me:
The point we were making was that MADRS-6, which is 6 of the 10 MADRS questions, is focused on the core symptoms of depression. This endpoint had not been used as a primary endpoint in a registration program. But it really doesn’t matter, as the study achieved statistical significance on BOTH MADRS-6 and the full MADRS-10.
Data will be released at an upcoming scientific conference.
“This is literally hot off the presses,” said Pops.
An approval is far from a slam dunk. A retrospective analysis can be hard to sell. But it’s a lot better than being forced to explain going 0 for 3. As of today, Alkermes is still in the fight. If it goes down now, it will go down swinging.
— Maxim Jacobs, CFA (@MaxJacobsEdison) October 20, 2016