2nd of two posts from a different mb.
Initial had highlighting which makes it much easier to read, doesn't appear to xfer here.
--
"May I ask, based on your investment here, and your research and any SIMS you may have run on your own, do you have a prediction to HEAT and if so a brief explanation of how you got there? Thanks Has2"
It seems as though that lately, the more DD I perform, the worst my investments turn out so in that regard you’ll be happy to hear I haven’t run any models on CLSN (though I may this weekend if I have time).
I am long here with my purchase in the very low $2 range. Still haven’t sold any and, when push comes to shove, I would have to attribute my “hold” to managements bullish actions and my belief that it is, at least, possible that they were un-blinded to the interim data (either immediately after the IA or at some later point, perhaps upon full enrollment). I know many dismiss this, and they may be correct. Kid posted a well written argument against them being unblinded (I think either you or Trond reposted it here). He laid out the rational for blinding and made his case rather nicely. There is a flip side to his argument though, The FDA has, on occasion, allowed the sponsor access to interim data. Dendreon is a good example (and one where Fleming came out vocally AGAINST the unblinding). There are other lesser known examples, such as the recent Zytiga 301 and 302 trials. My point is, it happens. Perhaps not routinely, but on occasion it is allowed with the blessing of the FDA and a mechanism is detailed in the regs. The unblinding plans are detailed in the DSMB Charter, a document that is much more secret than the SPA or the protocol.
One argument FOR Celsion having had access (w/ the FDA’s blessing) to interim data is the very comments MT offered when they were denied the additional interim. He said (March 12 CC) the FDA said that doing so would risk the integrity of the trial (not a direct quote but the general gist). The main way the trial integrity would be at risk would be if the request for an additional interim was made with the knowledge of the interim efficacy data. In that case it would introduce a bias that was not quantifiable. The FDA hates that. So the denial is evidence (not proof) that they may have been unblinded.
Now, in the rare occasions where I think a sponsor HAS had access to interim efficacy data, such access was for the purpose of drug development (expedited filings) and care was taken to NOT let those at the sponsor involved with the trial to have access to the data.
I explored this in part with my earlier question about insider trading: if someone at the Sponsor (legally) had access to interim efficacy data would they be legally precluded from purchasing stock? I have given this much thought and conclude the answer is: No, they would not be precluded from buying. Suppose for an extreme example that the IA showed an HR or 0.55 and the p value approached or even crossed a stringent hurdle (say P=0.000003). The DMC would approach the Steering Committee at CLSN who might then consult with the FDA. The FDA might, correctly, counsel against stopping. The reason being it is possible that the PFS results do not actually relate to a survival difference. Ethically, the question about true efficacy is still open, and ethically they can continue to randomize patients to placebo and continue the trial. But now at least SOME at the sponsor would know. There is certainly no guarantee that the PFS results would not change at the final analysis with continued follow-up. Basically, they would know the interim but not the final results. The uncertainty allows more patients to be ethically randomized and it allows insiders to make purchases.
Now, I don’t necessarily think the above situation unfolded although the timing (delay) in the announcement of the IA1 “continue” is curious. I think it more likely (though certainly not definite) that the sponsor was allowed access to the interim data so that they could prepare their NDA. (in fact, didn’t they say recently that they have started preparing some sections? Or am I confusing that w/ another stock I follow).
Anyway, below is the FDA’s view on the Dendreon unblinding and how that trial was affected. It addresses most of Kid’s argument.
Interim analysis: A planned interim analysis occurred in May of 2008 after 247 death events had occurred. The hazard ratio, 95% confidence intervals were provided to the sponsor by the independent Data Monitoring Committee as per the IDMC charter. Since the primary endpoint was overall survival, accrual was completed and the study remained blinded, FDA does not believe that the study integrity was compromised by this procedure.
The reviewer goes on to reference the DSMB Charter (something we never get to see).
Reviewer’s comments: The IDMC charter version 4, submitted to the FDA, specifies that the IDMC will provide overall survival results (hazard ratio, 95% confidence interval) to the applicant. The study was closed to accrual before the date of the interim analysis. The annual report submitted in March 2008, confirms that the study was closed to accrual, and 466 subjects had received at least one infusion of sipuleucel-T or placebo. No protocol modifications were made subsequent to the submission of the interim analysis data; therefore, the integrity of the conduct of the study was maintained. In addition, the interim analysis did not result in unblinding of any of the study participants to treatment allocation for individual subjects; maintenance of the blind decreases the risk that management of the subject would be biased by the results of the interim analysis. Also, survival, the primary efficacy outcome measure, is relatively resistant to biased assessment. Furthermore, this review included confirmation of subject deaths by examination of death certificates and Social Security death indices. In summary, the applicant’s access to the results of the interim analysis is unlikely to have compromised the study integrity or the results of the primary efficacy analysis.
Would Fleming let it happen in a trial he was involved in? Probably not, but if it were allowed, he would insist that the trial investigators and enrollees not HEAR about it (his main objection to the Dendreon interim )
TL;DR: I guess that I want to believe that they either KNOW the interim efficacy data or they ARE OUT OF THEIR FRIGGIN MINDS for not diluting at these prices. So, I am holding (but I am well diversified)
