The recent meeting of an FDA Advisory Committee charged with making a recommendation about an innovative gene therapy for the treatment of Duchenne Muscular Dystrophy was surprisingly dramatic. It was, in fact, so highly scripted that could be watched as a telenovela. Anyone who wants a glimpse at the problems we will soon face in evaluating hundreds of new gene therapies should watch this amazing movie.
The drama came from the fact that the raw data amassed to support the approval of the new drug were woefully inadequate. The drug manufacturer, Sarepta, pooled data from three different studies with three different study designs. Only one of the studies was blinded and placebo controlled. That study showed no statistically significant difference in the primary outcome between the treatment group and the placebo controls. But a subgroup of the youngest patients in that study seemed to do better. Another study showed that the drug led to improvements in a surrogate endpoint – the production of dystrophin. But an FDA analysis showed that this surrogate endpoint did not reliably predict clinical improvement. Most confusingly, the company changed its manufacturing process halfway through the studies, so that it wasn’t clear whether the pooled data was even applicable to the product that the FDA was being asked to approve. Looking only at the objective data, the drug didn’t stand a chance of approval. In fact, an internal group at the FDA had recommended that the company’s request for accelerated approval be rejected. They were overruled by higher-ups who decided to convene this Advisory Committee. The drama, then, was in the ways that Sarepta made a remarkably compelling case for approval.
Some of the drama seemed inadvertent. For example, the Chair of the Advisory Committee, Dr. Taby Ahsan, was from City of Hope Medical Center. Her live screen shots say “City of Hope” in the upper left-hand corner, as if it were a tag line for the entire meeting.
The order of presentations was key to the dramatic effect. The FDA went first and explained the guiding regulations, particularly the requirements for approval under the FDAs accelerated approval (AA) program. That program has been controversial. It grants accelerated approval to drugs that don’t meet usual FDA criteria with the understanding that confirmatory trials will be done quickly. Serepta had three other drugs approved under the AA program. To date, they haven’t completed a single confirmatory trial.
Not surprisingly, then, Serepta started by assuring the FDA that their confirmatory trial, was under way, fully enrolled, and likely to be completed by September 2023. They claimed that accelerated approval, now, would not jeopardize that study. Sarepta experts then explained the studies already completed, their primary endpoint, the animal data that supported clinical trials. They glibly explained away many of the most problematic aspects of studies.
After the drug company presentations, the meeting opened the floor to members of the public. At such meetings, we now know, many members of the public are supported by pharmaceuticalcompanies. For this meeting, parents gave heart-wrenching testimony about the brutal burdens of DMD and of the ways in which Sarepta’s drug led to important clinical improvements. Many parents supported their remarks with before-and-after videos of their children. There were no DMD parents who had chosen not to participate in studies. The visual evidence was compelling. After seeing these videos, I surely would have voted to approve the drug. My reasoning would have been that the drug seems to work for at least some children, the alternative is bleak, the safety profile is acceptable, and many parents are eager to have their children treated. As one parent said, “Thanks to gene therapy, we now have hope.”
But then, in the afternoon, after Sarepta and the Sarepta-supported parents, the FDA presented the results of its review of the evidence submitted. The review was meticulous and devastating. It showed that the objective evidence did not match the results given in the personal testimonies. It suggested that approval, now, might preclude the possibility of completing the confirmatory trial. It raised serious questions about the use of micro-dystrophin levels as a surrogate marker of clinical efficacy. Listening to this presentation alone, I surely would have voted against approval.
The Advisory Committee was deeply divided. Ultimately, they voted 8-6 for approval. Their recommendations are not binding. The FDA decided to postpone its final decision for a month.
Gene therapy for DMD and other rare disease will likely transform pediatrics in the next decades. We need the research. We need it to be done well. The AA approval program has helped bring new drugs to market for these diseases. As of March 2023, 295 drugs received AA. About half went on to full approval. But only a handful have been withdrawn from the market.
The problems with the AA program are widely recognized. There are two bills before Congress right now, with bipartisan support, to reform the AA process. The key is ensure that, if AA is granted, the confirmatory studies will be done in a timely manner. In DMD, the confirmatory studies have not been done. That has to change.