Depression Trials Fail, Alkermes Sees Silver Lining. Investors Don’t.

Xconomy Boston — 

Depression is a notoriously difficult disease to test new drugs against for a multitude of reasons. One reason is that a pervasive placebo effect tends to cloud the clinical trial data.

Alkermes hoped it had the answer. It has been testing an experimental drug in an unusual kind of trial designed specifically to set aside patients who benefit from the placebo effect; that is, patients in a trial who show improvement even when they’re not taking the drug being tested.

The first results came in from that test today, and they weren’t good.

Based in Dublin and Waltham, MA, Alkermes (NASDAQ: ALKS) said this morning that its once-a-day pill, ALKS-5461, failed to meet its primary goals in the first two of three Phase 3 studies for major depressive disorder. Patients taking the drug didn’t fare better than placebo on 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.

Shares of Alkermes were down about 35 percent, to $39.79 apiece, in pre-market trading on Thursday.

Still, Alkermes believes it has a shot. The company said there was a “clear trend toward efficacy” in one of the trials. After a look at the data, Alkermes believes there is evidence that the drug is working.

That type of silver-lining analysis usually receive sideways looks from investors, and today is no exception. Occasionally a drug maker will forge ahead with new or revised studies based on these fainter signals. For example, Eli Lilly (NYSE: LLY) saw its anti-Alzheimer’s treatment solanezumab fail a major 2,000-person study in 2012. But the company said the drug’s effect was stronger in the subset of people whose disease was not as far along, and it used those data as the basis of a third trial, now ongoing. Independent investigators have also chosen solanezumab for trials in patients whose Alzheimer’s is in the early stages.

Alkermes hopes to learn and benefit from its setbacks as well. Based on insights from the two failed trials, Alkermes said it plans to enroll more patients in the third of the three Phase 3 trials—dubbed the FORWARD program—and change its statistical analysis plan. If that study generates a “clear positive outcome,” Alkermes would take the data to regulators to discuss a path forward, the company said in a statement.

“Clinical trials of new medicines for the treatment of major depressive disorder are complicated by significant placebo response,” said chief medical officer Elliot Ehrich, in a statement. “We designed the… pivotal program to include three efficacy studies as we recognize that this is a challenging disease state where multiple clinical studies and expansive analyses are generally necessary to confirm the efficacy of a new medicine.”

The SPCD design was supposed to wash out the placebo responders and give a clearer picture of a drug’s effect. As Xconomy wrote last month, the design was devised over a decade ago by Maurizio Fava and David Schoenfeld of Massachusetts General Hospital in Boston. The idea is to randomize and split patients into a drug and placebo arm, and then take the placebo patients who don’t respond and randomize them a second time into either a placebo or treatment arm. It’s a method that’s only been used in about 25 psychiatric trials since 2003. The Alkermes study was its most advanced test.

ALKS-5461 is meant to be an option for folks who don’t respond to initial depression treatment, typically selective serotonin reuptake inhibitors (SSRIs), and before they move on to more invasive therapies like deep brain stimulation. It works differently than SSRIs, balancing opioid receptors in the brain, rather than changing the balance of serotonin. The drug is of major importance to Alkermes, which has been transitioning from its roots as a drug delivery specialist to a developer of its own in-house medicines over the past several years. I spoke with CEO Richard Pops in 2014 about that evolution.

—Alex Lash contributed to this report