If science is so neatly fit, and perfect, you have to question! This is why this observation must be observed by other investigators and with more patients in order to remove reasonable doubts on the validity of such data. That's the foundation of the FDA action in Nov. 2013.
I just point out there are other robust dataset besides the Ataluren data to suggest stability of 6MWD for those age 7 and baseline 350. The 80 meters decline observed in the N=4 placebo arm was not credible and the FDA has made the right call to question those. We will soon know how the Drisapersen phase III placebo arm look like for those with 6MWD at the same cut. A 20-30 meters decline after 48 weeks, in the age 7, 6MD 350 is probable, a 80 meters decline is not. That could be the conclusion based on the compilations of multiple datasets and "could" serve as the comparison baseline in a "placebo-less" confirmatory trial for Eteplirsen.
As for the two year data, I suggest you listen to Dr. McDonald's call. If you are basing on Eteplirsen 96 or 120 week data to suggest the much slower decline versus the two-year Mercuri data or the Goemans data than what say you with the Drisapersen 177 week data? GSK/RNA will soon reveal its extension trial data, will we find similar slower rate of 6MWD decline there? If yes, then what? Should these DMD trials run for 2 years instead of 48 weeks?
Size of course matters. Imagine FDA let an open label, single investigator, N=12 trial (and with all the problems of patients selection bias between the arms) to go through as the registration trial.? Every tin pot bio scams out there could do the same, and the FDA would have no time to review any real drugs.
MDA is correct to call the "trial" an experiment, It is an experiment with interesting results that have yet to be confirmed by another investigator in another investigational site. It is a science experiment that has yet proven to be an actual effect observation vs a chance observation.
We would not be in the same discussion had SRPT expanded the Phase IIb with additional patients with additional investigators. But no, mgmt. was too busy to move from OR to MA, too busy to jam the FDA with a N = 1, N=12 trial for AA, too busy to sell its story on social medial, hedge funds and selective advocates. They raised 300 Million dollars and yet have no time to consider expanding the phase II and to get more valuable data in 2013 from more patients.
Blame CG and his cronies like Kaye and others for the shortsightedness, and their hubris to think they can run around the FDA rules because of social media and the advocates.
This is the link for the Dr. Mercuri NH data.. LINKhttps://twitter.com/AndyBiotech/status/402884608233963520/photo/1 (remove LINK, copy and paste and read the link, picture taken during the McDonald NH live webinar on PPMD).
Age =7, 6MWD = 350: 52 week change (-13, -29 with N=42, 45 respectively).
Both Drs. Hoffman and Mendell are well respected investigators in the field, I don't believe pick one or another would make a difference. The real problem is the data. The FDA just do not believe it. The 80 meters 6MWD declined in the placebo arm was not credible. Folks complained about the Ataluren dataset because of one or two Beckers patients in there, so, I threw that away. Dr. Eugenio Mercui showed two other NH datasets with age = 7 and 6MWD = 350. The 52 week decline in those two groups (N=42, N=45) were 13 and 29 meters, respectively. Those were robust datasets with significant populations, not the N=4 case in the Eteplirsen Phase 2 placebo arm. People could throw away the Ataluren data and still could not argue the doubt FDA got was misplaced. If FDA could not take the placebo arm 6MWD data with confidence, and with significant concerns with the way Dystrophin was measured (counting fibers, not subtracting revertant fibers post treatment, biopsies in different muscle groups from baseline verse 48 weeks), then what do they have to put any confidence in to argue for an accelerated review of the application? Some people have suggested the 80 meters decline in the placebo arm was due to gross selection bias of the four patients there in terms of mutation status, weight, height, steroid dosage etc., I see those just cast additional doubts to an existing case already in deep concerns. It is pretty amusing to read all the blame and theories here, but the truth lies within the clinical trial data and conduct itself.