I had to split time between yesterday’s Advisory Committee meeting and my day job, so I can’t profess to having seen every presentation and testimony yesterday. But based on what I did see and what I gleaned through other sources (thank you Twitter and The Street’s excellent live blog), here are a few scattered thoughts on what unfolded:
- It seems like FDA really wants to approve eteplirsen and just can’t find the scientific evidence to do so. Some people are claiming that FDA has it in for eteplirsen, but I’m just not seeing that. Between Janet Woodcock’s statements and the general focus of FDA’s rebuttals it seems to me that FDA is looking for ways to approve this drug without setting the precedent that companies can get orphan drugs approved just by sheer will and public pressure.
- Keep this perspective in mind: Duchenne patients are rightly asking for a therapy, which is not what the panel voted against giving them access to—the panel is saying that eteplirsen is not, in fact, a therapy at all. A drug in development is called an “investigational agent”, not a therapy or treatment, for a reason—it may not actually be efficacious. Believe it or not, FDA does not want to hold effective medicines back from patients, but it does want to make sure companies aren’t peddling false hope.
- Don’t make the easy things complicated. Any data a company presents to FDA should be bulletproof. There’s no reason FDA should waltz into the facility that generated the company’s data and rip their methodology to shreds—it’s the company’s responsibility to make sure their data flawless, especially when they’re conducting a tiny, non-placebo-controlled trial under intense public and FDA spotlight.
- Don’t try to hoodwink the FDA or the public, because the second you do that it taints the rest of your story. Sarepta’s dystrophin data analysis was suspect; one of their expert consultants presented a misleading Kaplan-Meier plot; factual and chronologic evidence points to FDA having asked for a placebo-controlled trial years ago; and so on. The accumulation of questionable facts, all of which were being stacked in Sarepta’s favor, planted a seed of doubt that both FDA and I are having trouble shaking.
- What’s missing? How about any evidence of a difference? There are many points of contention between Sarepta and FDA—arguments over the validity of endpoints, he-said-she-said about expectations, etc. But, in my opinion, this is all just noise. The central question I’m asking myself is: what data would I want to see to make this a slam-dunk approval? And the answer I keep coming up with is: an actual difference between anything—between eteplirsen and historical controls, between eteplirsen doses—anything at all. And I’m not talking statistically significant difference, even just a good trend would do (when the data are properly analyzed).
- FDA shares the blame in eteplirsen not looking to be approvable. From the FDA’s briefing document:
“FDA strongly encouraged the sponsor to conduct an adequately powered, randomized, placebo‐controlled trial(s) to assess the clinical effect of eteplirsen. But in the context of an ongoing series of reports from the applicant and its academic associates describing marked effects on dystrophin production and stabilization of disease progression, many in the DMD community had strong reservations regarding the ethics and practicality of conducting another placebo‐controlled trial of eteplirsen. Given the apparent difficulty of doing such a trial, FDA expressed willingness to consider an externally controlled trial, although stating clearly that interpretation of the data could be difficult, and that the acceptability of the study would be a matter for NDA review.”
Between this statement and FDA’s presentations yesterday, I can understand where Sarepta’s confusion came from. As drug developers, when FDA says they’re open to something we tend to run with that, because FDA typically errs on the side of just saying “no” to things. Here it seems that FDA succumbed to public pressure and gave Sarepta a wishy-washy statement that FDA had no intention of following through with.
FDA’s presentations about the limitations of historically controlled trials were not Sarepta-specific—they described limitations that are inherent in any historically controlled trial. Whether FDA got cold feet or never intended to allow a historically controlled study to be the basis for eteplirsen’s approval we may never know, but what’s clear is that FDA was not.
- Presentation skills are important. I’m not singling anyone out here, but anyone who’s going to take the stand to talk about a polarizing subject in front of a passionate crowd should have their presentation and delivery completely airtight and polished. This was the case for many of the presenters, with at least one notable exception. Don’t let your message get lost in your delivery.