an uphill battle, because the FDA has been just as critical of the eteplirsen data.
Indeed, prior to today’s panel, analysts were highly skeptical of eteplirsen’s chances. In research notes, Jefferies analyst Gena Wang called approval a “low probability,” while RBC Capital Markets analyst Simos Simeonidis wrote that he expected a rejection. That’s because the agency has twice skewered Sarepta’s data, first in a set of briefing documents the FDA posted in January, then again last week in an update to the documents.
The reason for the skepticism is Sarepta’s data comes from a 12-patient, Phase 2b study. In that study, the boys were split into three equal-sized groups and given weekly infusions of a low or high dose of eteplirsen, or a placebo. After 24 weeks, the placebo patients were crossed over to one of the two eteplirsen doses. The goal was to show these boys were producing more dystrophin and walking farther than they would have without treatment.
Making the sample size even smaller, Sarepta excluded two of the 12 patients who lost their ability to walk early on. Sarepta argued that their disease may have already been too far along to benefit from any drug therapy.
Sarepta is making the case for eteplirsen with the remaining 10 patients who, after some four years on eteplirsen, are still able to walk, and that these patients are doing much better than they would otherwise, based on historical data the company has accrued. This is an important point: The gold standard of clinical trials is to build a study that compares patients on a drug with those on a parallel track without the drug. But Sarepta has chosen to compare eteplirsen patients on its drug with patients from a different time period who might or might not have resembled the eteplirsen patients. That decision was made for two reasons. First, Sarepta didn’t have the resources several years ago to start a placebo-controlled trial. Then, when it finally had the resources, ethical problems emerged. As Christine McSherry, founder of the nonprofit Jett Foundation, said at the hearing, “We saw signs the drug was working,” which would have made a trial in which half the kids take placebo an untenable situation.
Sarepta’s unusual trial design isn’t unprecedented for rare disease treatments. For instance, the company has pointed to the 2006 FDA approval of Genzyme’s alglucosidase alfa (Myozyme) for Pompe disease, a decision based on a study that compared the drug’s effects to a “historical control”—that is, a database of patients deemed to have similar characteristics to those who took the drug in Genzyme’s trial.
Still, the FDA has been highly skeptical of eteplirsen’s results, and that skepticism continued today. At the hearing, the FDA’s scientists called Sarepta’s trial an “apples to oranges” comparison, citing potentially critical differences between the eteplirsen patients and control group. In a hostile environment, the agency’s clinical review team—led by Ronald Farkash and Eric Bastings—told the crowd they “haven’t heard the whole story.”
“My role, regardless of the pressure that has been placed on my division—and on the eteplirsen review team—is to present for scientific review conclusions about [the data]. We are a science-based organization,” Bastings said. “Accelerated approval cannot be used to compensate for weak or inconsistent clinical findings.”
The majority of panelists sided with the FDA scientists. One panelist, University of Texas-Houston Medical Center neurology professor Nicole Gonzales, called the data “problematic.” Harvard Medical School professor Aaron Kesselheim said the study was “not adequate or well controlled.”
Chiadi Onyike, a psychiatry professor at Johns Hopkins University School of Medicine, said the data don’t draw a clear connection between the drug helping patients make more dystrophin and actually having a benefit. “I can’t get from dystrophin production, even if I accept it, to clinical effect,” Onyike said.
Sarepta brought the controversial data to the FDA for complex reasons. Under former CEO Chris Garabedian, Sarepta triggered the hopes of patients and advocates with promising Phase 2b data in 2012. With their encouragement, Garabedian felt Sarepta could ask for accelerated approval, then a new process, with a smaller data set.
Instead of a fast-track approval, Sarepta has since been on a regulatory roller-coaster. After much back-and-forth and delay, Garabedian resigned in April 2015 amidst a reportedly contentious relationship with the FDA. (Chief medical officer Ed Kaye is now interim CEO.)
Sarepta eventually started a larger late-stage trial in September, but that won’t read out for another few years. (Like the one under review today, it is not placebo-controlled.)
That’s a concern for another day, however. For now, it’s all about whether accelerated approval is still an option, and what it would mean for Duchenne patients and Sarepta one way or the other.
“To cure Duchenne, we will probably need a combination of therapies to treat the whole disease,” Miller said. “But we cannot test drug combinations until the first drugs are approved.”
