What data, indeed? I'll get to that in a bit.
After reading up on the difference between statistical and clinical significance after that call, I think that this distinction supports my stated hypothesis around the company asking for broader labeling. Essentially, clinical relevance or significance is a squishier topic that statistical significance. The latter purports to show the likelihood that the outcome was due to chance, and thus unlikely to be repeated in future studies. Statistics is, of course, a highly-developed area of mathematics, and the basic principles of statistical power for small, simple experiments are broadly agreed upon. I think it's safe to assume that both the company and the FDA had a uniform understanding of the statistical power of the trial design from the outset.
Clinical relevance, on the other hand, seeks to answer questions like "how big of a difference is there from the control?" Or "do the side effects or risks justify the difference"? Or "is the result enough to get physicians and patients to change practices"? In other words, highly subjective. The least subjective part would be "how big is the difference from control"? This is something that is essentially negotiated during the construction of trial protocol and the definition of primary and secondary endpoints. The FDA agreed to the protocol from the outset with regard to the required reduction in viral load at one week, and did not change this in the October meeting. Therefore, I don't believe that this portion of it was actually in question during that meeting, and therefore isn't the issue when they reference clinical relevance. This is supported by the fact (unless everyone is simply lying to us) that the data was still blinded, and the FDA could not have seen the data to determine that the observed difference was not clinically relevant.
So what could the FDA now need to see, in terms of trial design, to confirm "clinical relevance"? Here, I'll reference my previous post where I hypothesize about the patient population, labeling discussions, and the relaxing of criteria for the additional 10 patients. In short, however, I think that for the FDA to deem the results "clinically relevant" in a broader patient population, they had to see at least some patients treated who fit that patient profile. Clinical relevance is just the semantics used to cover for the (hypothesized) fact that the company handled the overall negotiations and ODD application poorly from the beginning.
As for the question of "what data"... again, pure speculation. It's clear to me - and this is in no way meant to belittle Dr. Pourhassan - that Nader's first language is not English and that he is not a terrific speaker. In fact, if I were Tony, I don't think I'd let him speak on investor calls. Since (I assume) that the primary endpoint data was still blinded, my take on the comment you highlighted was that Nader was referring to the patient profile of the 40 enrolled as "the data". When the FDA looked at that, they then said (in my hypothesis) that it wouldn't provide clinical relevance for the (assumed) broader ask re: labeling.
Tons of speculation and assumption here, obviously and admittedly. It's just the scenario that I find best fits with any and all facts that we, as investors, think we know.
