16 December 2013
I wrote about a month ago voicing my concerns related to FDA’s interpretation of ‘recent data’ related to eteplirsen. Such had been referred to by FDA-DNP as being a part of your ‘premature to file NDA’ and ‘need for placebo group for a Ph III study’ statements. As you may remember, I felt your interpretation(s) failed to properly apply the data—as presented in McDonald’s, Henricson’s, and Mazzone’s 2013 publications---for the grouped natural history of DMD boys—at least as compared to eteplirsen’s data. However, this letter is about other concerns in your process(es) for detecting possible clinical benefit of drug candidates for DMD.
Though the recent PPMD forum on December 12 was a high-profile event and included FDA’s presence and involvement, several statements attributed to FDA personnel raise significant concern that the basic FDA-DNP ‘old’ way of thinking is not in any way changed, or, even in process of changing. However, if new therapies such as for DMD are to receive a fair hearing, I’m convinced such thinking will have to change.
For example, a tweeted statement (attributed only to “FDA”): ‘We are looking at the natural history data very carefully and we are really trying to understand it.’ What’s wrong with that, you ask? It is this: natural history in DMD (and similar conditions) is not only that definable by group stats, but also that natural history unique to each individual affected. The former seems to me reasonably clear, but the latter is in fact crystal clear. There is a pattern of decline in each boy which is not known to spontaneously reverse—this pattern is itself a definable natural history—for that boy.
slightly off topic, but can you help me gain access to complete versions of both or your FDA letters recently? I am compiling a package that I will be submitting to CBS here in Boston, and perhaps at the national level, for an investigative journalism piece. I can find your writing but it's in pieces, and it would be easier if I could just have one, or perhaps two, complete documents. I had a most interesting conversation this morning with a local, popular TV figure who requested that I put together a package that gets them up to date on the SRPT/Etep/FDA story, and as best I can, I intend to do so. I believe what you have had to say in your two letters is meaningful. Thanks, Grey
WOW!!! and i thought your first letter was suberb. You out did yourself with this masterpiece. I wrote a smidgen of a letter to Dr Woodcock compared to yours and basically asked how their neurololoical division could arrive at a decision regarding eteplirsen without the complete data from GSK and how they could define a new natural history using McDonalds data when it included Becker boys. Have you received a response to your first letter? I have not.
Again masterful job Doctor and I applaud your tenacity!!! You make me proud!!!
To all, lets keep them coming: media (grey and all). letter and phone calls to congress people (everyone) and Dr Woodcock herself (everyone)
jrrt1 - Along with your timely applied knowledge this post reflects an outrageous stubbornness - bordering with stupidity and in this extreme case crossing the line of the criminality within the WALLS of the FDA. The 2 years old Sarepta's clinical data will stand in court against FDA's crime. Bless your heart.
Sentiment: Strong Buy
Bravo Doctor. As you well know you speak for me with this letter. The truth of the individual/single-case-design viewpoint you express is actually self-evident. (But woefully out of vogue.) Sooner (hopefully) or later it will become accepted and later still seen as obvious.
E.g., a boy who had declined to the point of requiring assistance to ambulate, then becomes ambulatory without assistance—or even begins to run again; or, another boy with severe toe-walking then resumes ability to walk almost ‘flat’ footed---then the clearly known natural history (for those individuals) has been altered! If a boy who couldn’t lift his arms above his shoulders can now ‘reach for the ceiling’ then his natural history has been altered. Something had to have occurred for such previously unknown improvements to have happened.
Based on data (including recent), such dramatic changes are not adequately explainable by ‘chance’, steroids, or, effort/placebo effect. Whether such clinical effect is from ataluren, eteplirsen, or other agent; or, whether it is in a majority or just a minority of affected boys, it should be clear that ‘natural history’ has been altered--in at least those boys! I.e., clinical benefit has been demonstrated compared to ‘natural history’!
Another statement of concern is attributed to Dr. McDonald to the effect that: ‘If 6MWD is greater than 350M at baseline, will have to show improvement or do longer trials’. This statement is another ‘group’ natural history statement—limited to trying to discern clinical benefit by alteration of natural history among GROUPS only. Though his field is DMD natural history (for ‘groups’ at least), this demonstrates the same worrisome narrow view of “natural history”. It is, in my opinion, the central weakness, if not failing, at the core of the FDA’s thinking. You/he are still only looking for clinical benefit in GROUPS, and only then, if finding it, are willing to apply it to INDIVIDUALS in the group (e.g., by approving its use). This shows an ‘old’ mindset which is inadequate for some of the diseases which 21st century medicine is beginning to attack. It is especially a setup for missing otherwise clearly demonstrable clinical benefit in DMD.
" ‘If 6MWD is greater than 350M at baseline, will have to show improvement or do longer trials’. "
THis statement suggests that unreal expectations must be met prior to a drug being granted approval. Since when, was it determined that for DMD treatment to be successful, actual "improvement" had to occur. This is completely ridiculous, in that, simply to halt the progression of this dreaded disease, and stop the decline would and should be considered a wonderful intervention in the progression of DMD. Now all of a sudden, McDonald suggests that for certain populations at the 350 line, they must show actual improvement?
I find that conclusion to be ridiculous. Maybe they should require that these kids dunk the basketball, or throw a flip on their BMX bikes, before recognizing that a drug has a positive impact on DMD? To simply stop further progression, would be considered by anyone being reasonable, to be a resounding success. To require more than that, is unreasonable.
As a physician, I understand the belief that double-blinded, placebo-controlled studies represent the ‘holy grail’ of proof—I actually agree with its applicability to many circumstances. However, in a rare disease whose course is very predictable based on an individual’s natural history, yet which remains confusing (to you, at least) from a group standpoint, then perhaps it is because this particular ‘holy grail’ can’t actually solve this particular problem!? Or, at least not ‘solve’ it without much unneeded loss/damage in the search for such a solution. This statement attributed to Dr. Unger is exactly on point to this whole concern: ‘We have seen a lot of variability in placebo groups from one trial to the next--it creates problems.’ Not so much a problem unless using an ‘old’ paradigm of natural history. Not so much either if you (FDA) stop looking from the group to the individual for clinical benefit, and start looking from the individual to the group!
So what am I suggesting? If an agent shows clear clinical benefit across a number of INDIVIDUALS, based on any of several areas never known to reverse spontaneously; and, if there is an adequate safety profile, then grant tentative approval for more general use—including boys not yet on the downward slope of previously irreversible decline (e.g., all 7yr old). Then require detailed clinical analysis on all users and before long, the FDA will have a sufficiently large database to compare with prior GROUP natural history data. This should allow you—starting from the clear-cut individual natural history alteration—to decide if that clinical benefit is also a GROUP characteristic. If so, then general use would continue; if not then this does not invalidate the benefit for that subset of individuals who DO benefit, and limitations on labeling may be felt appropriate.