The FDA’s approval of Sarepta Therapeutics’ eteplirsen for Duchenne Muscular Dystrophy (DMD) turned the spotlight on a decision process that has tapped fundamental fissures within the FDA and the scientific community, fractures that could reverberate in far larger scale in the not-too-distant future.
To quickly recap–Sarepta filed an NDA for this drug, which addresses DMD, a currently untreated, crippling and eventually fatal rare disorder, based on an uncontrolled trial that enrolled just twelve patients (only 13% of DMD patients have the genetic anomaly in the dystrophin gene (exon51) that make them potential treatment responders). The FDA’s Advisory Committee voted 7-3 against approval (albeit with three abstentions, abstention not exactly looking like a badge of honor in this context), but a highly vocal and effective advocacy campaign revolving around the desperate parents of these children finally persuaded CDER chief Janet Woodcock to approve it. One of her subordinate physicians went over her head to appeal the approval, but FDA Commissioner Robert Califf finally announced that he would defer to Woodcock and leave her decision to approve untouched.
NIR has not covered DMD, and thus we have not followed this process in great detail. We are not in position to make informed comments about the quality of the data presented in the NDA, though there is certainly heated debate on the topic. The primary endpoint was the change in dystrophin-positive fibers, reflective-in-theory of improved expression of dystrophin in treatment-responsive patients. But there was no consensus amongst FDA AC members on whether the change in dystrophin levels was enough to be clinically meaningful, whereas a group of DMD researchers sent a letter to the FDA stating that the observed improvement in a walking-capacity secondary endpoint was enough to be clinically meaningful. In other words, this was not just a question of emotion vs. rationality: The interpretation of the endpoints measured was, ultimately, a subjective one when extrapolating to real-world functionality.
With all due respect to the sanctity of the scientific method, there is no immutable and agreed-upon gold standard here, and a stubborn insistence on some ambiguous expression of scientific purity is no more rational and objective than the fervor displayed by parents insisting that they had seen improvement in their children from the use of eteplirsen.
While the intra-agency tumult has not yet subsided, and indeed the main FDA reviewer on the case has resigned, Sarepta‘s valuation skyrocketed as they announced the approval, and their intention to charge an average of $300,000 per patient per year. With perhaps 2600 patients in the US who might be candidates (20,000×13%), 75% market penetration would mean a total health care cost burden of $585 million, which in itself, would be sustainable. Sarepta is required to conduct a post-approval placebo-controlled trial, which had already been open for enrollment. However, now that eteplirsen has been approved, this begs the question–what parent would take a 50% chance that their child would receive saline rather than active drug, thereby guaranteeing that their child would continue to deteriorate? Post-approval efficacy testing is thus even more fraught with obstacles than usual. Eventually, eteplirsen could serve as the standard of care control group against which new DMD candidates can be compared, but that’s not a nearterm scenario. And speaking of which, to what standard will trials for the next generation of DMD treatments be held?
In this instance of a rare, devastating pediatric disorder, using a drug with no apparent safety hazards attendant, it is hard to find serious fault with Woodcock and her decision. But it is not an open-shut case, approval does come at a potential cost. And what about large-scale disorders, where there are safety risks and big-time costs–would the same humanitarian concerns apply?
This is not a purely hypothetical question, we could well face it next year if solanezumab’s next Phase III data mirrors its impact in its previous round of testing. Our view has been, and is, that the best-case scenario for Lilly would be that the sheer size of the trial might provide enough statistical power to beat the magical, and arbitrary, benchmark of p=.05. And that these results might mimic the magnitude of therapeutic effect seen in the previous Phase III, which was a difference of 1.91 points on the ADAS-cog at one year. Given that the general premise has been that a difference of four points begins to augur real-world functional import, it is thus possible that solanezumab will be shown to predictably produce a treatment effect that is not ‘visible to the naked eye.’
This is where the eteplirsen precedent becomes tricky. Without any disease-modifying drugs on the market, even the prospect of imperceptible slowing would create a intense clamor and demand from the families of individuals with Alzheimer’s, particularly those early in the disease process, where it might be hoped that any effect on the disease could extend the time that they can stay independent, and remain themselves. Solanezumab has a relatively more benign safety profile than some other amyloid antibodies, without the vasogenic edema seen with aducanumab, for example. But providing a monoclonal antibody to a large number of elderly individuals is almost surely going to end up with some dire outcomes that could, however unfairly, reflect badly on the regulators. And if, to arbitrarily select some parameters, one million Americans were to receive solanezumab at a cost of $30,000 each (almost certainly undershooting what Lilly or any other Big Pharma would seek for a biologic), that would constitute a healthcare cost burden of $30 billion. That’s a more serious dent in the health care budget, one which would trigger a major debate about the standard that should be set: Must therapeutic benefit be ‘visible’ in order to justify spending $30 billion to get it? Whatever one may think of Woodcock’s verdict, this is a harbinger of the mega-debates yet to come, at a time when thoughtful discourse is in very short supply in the US.