I don't know, "reasonably likely to ..." is subjective. "substantial clinical evidence" is subjective in regards to "substantial". They both are subject to interpretation, which is what leave things up in the air regarding the adcom. The key issue to me is the error bars: error bars mean an average may be a fluke --or it could be a real effect. If its a real effect, the 2-5 years to do another study will grievously harm those not treated relative to those (effectively) treated. If its a fluke, no harm in waiting. In a safe drug, the risk of waiting to prove the error bar question is very high. In a safe drug the risk of using a drug later found ineffective is low.
Guillain-Barre syndrome can be extremely labor-intensive to treat. (Think a month in the ICU at $15K per day on a respirator , IV lines etc.) 25% of patients with GBS have respiratory failure. Like microcephaly, GBS is a medical nightmare, luckily quite rare.
A little thought experiment: Imagine it is two years from now. All etiplirsen boys are still walking, despite lower NSAA scores. Their course defies explanation. But ambulation was not a primary end point. All historical controls are now, already non-ambulatory. What does FDA do--require further study with ambulation as a predetermined endpoint? FDSIA seems to me to at least imply that FDA is now, unlike previous time, supposed to make accelerated approval possible for drugs that are 1. safe, 2. may well work, despite the existence of chances that they don't. 3. when the illness is grave 4. when no alternative treatment exists 5. when the risk of no treatment is significant morbidity and mortality. FDSIA and laboratory statistical inference are actually at odds with one another.
Could one make the point that the FDA is making a CONCLUSION about the 5 NSAA points per year based on limited data? Limited by the group's heterogeneity, by the epoch of time studied, by ASSUMING the decline slope is a linear, not a variable, function. No one has studied the decline-slope function for differing dystrophin parameters since this is a new area of study. Should FDA be able to raise questions based on hand-waving assumptions on DMD, a disease with many unknowns?
1. Using the historical control group comparison, the Etiplirsen group error bars overlap the historical controls' course. So FDA argues that the error bars show that even though the etiplirsen boys are still performing better (on average) than historical controls, the error bars extend far enough to include the chance that this average is a fluke. Since the FDA uses science and statistics to determine fact, the FDA must use these error bars to say etiplirsen is not proven to work, yet. SRPT can argue that approval, FDSIA-style-with-confirmatory-trial, should happen because the average MAY be true. The answer to this question will be a turning point for FDA precedent whichever way it goes. Will FDA go for glass half full or half empty on this issue? imho
In Japan (2014 I think) they have a law saying if a stem cell treatment is safe it can be used for ANY application.
Why is this so? To allow for experimentation, to speed innovation, to find the next "positive 'Black Swan"".
Nature, unlike Euclidean geometry, and the assumptions underlying classical statistics, does not work on a 'Normal distribution'; it works on a fractal distribution, wherein black swans are far more common than normal curves would predict. (See the book "The Black Swan").
FDA is doing a job it was purposed to do: used Normal-Population statistics to weed out ANOMALIES. But science advances by newly discovered anomalies. Probably some kids benefit from dispersion, probably many benefit from etiplirsen--which is safer. All new effects are anomalies at first; proving them not-anomalies delays innovation, especially in safe compounds.
Your timeline for NSI189: Depression trials could be done quickly if enrollment etc is efficient since results, like all antidepressants are done in the order of 24 weeks or even much less. I was actually puzzled why the study would take 18 months--academic medicine speed perhaps. NVAX just began a study though where their enrollment of 11 THOUSAND subjects took about two weeks(!)--amazingly efficient management. This timeline is a key question to me. Anyone with more detailed information about it--whether the basic data can be expected in 6 months or not (that timeline would be super-efficient validate the notion that this is a serious product candidate).
FWIW, "To revolutionize clinical trials and bring new therapies to cancer patients more rapidly, the F.D.A. should approve cancer drugs solely based on their safety and their ability to hit important molecular targets responsible for cancer rather than on their ability to independently improve survival or improve on an existing treatment,"
I have too many thoughts and feelings to mention, but trying to add to the discussion one point at a time:
The FDA has actually hindered the study they are criticizing as not being a full blown phase III already; they did this in 2013 when they said they weren't sure what outcome criteria should be studied. They said 6MW was potentially subjective, that dystrophin measurements were controversial. Unger himself said maybe other things like opening a milk bottle or ADL-realted things would be workable-- nobody mentioned NSAA points(!). No wonder no drug company rushed to do a study when FDA vetoed then-current outcome measures.
Children are still deteriorating. Scientific method is important, but in this case, FDA helped prevent it from being implemented (no endpoint= NO experiment).
FDA briefing docs appear to say the florescent dystrophin ("green circles:) never has been remotely close to normal--in contrast to SRPT's slides from pre-2012 showing normal-like circles in etiplirsen treated subjects, making it look like SRPT did indeed cherry-pick its "green circles" to give a misleading impression. This implication needs to be refuted by SRPT. Or maybe FDA cherry picked its examples to show lack of effects--which would be shamelessly misleading too.
logical tough questions about safety and efficacy are expected . absence of something unexpected could help the price.
Sentiment: Strong Buy
And now BMRN competition is.......wait for it................. GONE
Sentiment: Strong Buy
He helped patient advocates argue a year ago against the FDA’s stated desire for placebo-controlled trials, saying “at that time, they didn’t even want to hear about dystrophin levels. They didn’t know what was going on. They wanted clinical outcomes."
Now the FDA asked SRPT for an exon51-matched comparison group (DMD-knowledgeable request), and dystrophin-validation. FDA's Biomarin adcom questions were knowledgeable regarding DMD. If they didn't know what was going on a year ago, they seem to now.
FDA requesting this comparative group demonstrates their intelligence about DMD and specialized trials, imho. The group is of other exon 51 kids, whose natural history is actually more downhill than other exons. bodes well for next week.
Sentiment: Strong Buy
Dr. Von Deutekom, drisaperson's inventor, said Prosensa/BMRN could only find dystrophin in less than 30% of subjects AFTER drisaperson treatment, (if I remember correctly from the BMRN adcom) which explains why they avoided trying to use it as a surrogate. No statistically significant clinical benefit with drisa either. And mortality risk.
SRPT has dystrophin in multiple biopsies after etiplirsen treatment, and statistically significant, long term clinical benefit vs natural history despite small Number of subjects. and safety. Thus FDA has hard evidence in favor of approving conditioned on a confirmatory trial (also known as accelerated approval).
However, FDA demonstrated today they can say no if they need to, despite political advocacy. The negative case is this last point, the positive case is the above. Quite the epic odyssey eh?
bump: related to today's news release on published alzheimer's results.
Sentiment: Strong Buy
So BMRN has no statistically valid clinical results with a large study, and Etiplirsen has statistical validity at p =.01 level, along with safety, ongoing confirmatory study underway, manufacturing capability. Annals of Neurology is published, peer reviewed positive finding, despite small N. If this isn't a case for accelerated approval, even without dystrophin as surrogate, what is? Just asking.
A possible benign motivation for drisa support by advocates: when etiplirsen was not as well tested as it is now, drisa was a hedge against etiplirsen being a placebo. But now, mounting evidence, hard evidence, time tested evidence for safety and efficacy exists in favor of etiplirsen. The prior hedge-bets on drisperson, despite toxicity, have become void, even though they appeared reasonable in a potential competitive vacuum in the past. This reasoning overall makes the drisaperson purchase make sense in hindsight, but even better for etiplirsen, it makes the case for etiplirsen and against drisa now. imho
I think BMRN's patent claims all synthetic nucleotides even though their molecules are a specific subset of same--like saying because penicillin kills bacteria then the patent covers all antibiotics, even those that kill bugs penicillin doesn't
FDA's Dr. Temple used the example of other exons as an example of using surrogate marker for additional clinical use. In a conference about two years ago, when he was explaining what a surrogate marker was used for, he cited this sort of example. If the etiplirsen molecule causes dystrophin production--a surrogate marker--then dystrophin for the current exons might be enough to give Accelerated Approval NOW for the other exons, with follow-on confirmatory studies to be done (as already beginning for other exons.)
It is conceivable Accelerated Approval for the other exons could be granted very soon (now, along with etiplirsen for exon51) given these considerations.
Sentiment: Strong Buy