I put together a spreadsheet based upon a moderately extensive literature search of past neuroprotective trials (all stroke so far, because it is easier to find them) and all of their failures (more than 10 different stroke failures looked at and another 4 or 5 glanced through). I was hoping to find some general themes and lessons learned. Some questions:
1) Is there a long history of really promising Ph II results followed by failure in Ph III?
Answer: No. No Ph II that I have found so far had results anywhere near as promising as the Dex Ph II trial - for instance, none reached stat sig on any of their primary endpoints, if they even looked at all. They had to data mine to find positive results if they could find anything at all. And none had results that just had as many positive aspects as the Dex trial. OTOH some didn't publish their Ph II or the interesting results - e.g. Cerovive shortened their therapeutic window from <24hrs in Ph II to <6hrs in the on-going phase III (I'd love to see the data that supported that))
2) Was there a strong indication that shorter time to administration is likely to be a benefit?
Answer: There are hints. For instance, the Cerovive decision mentioned above. Or the tirilazad RANTTAS II trial with a dosing within 4 hours, which unlike its earlier failed Phase III (up to 6 hours til dosing), showed some trends. The trends are not as strong as the Dex trends even though it is a larger population, but ... . Surprisingly there are a lot(!) of trials with first dosing up to 24 hours after the event.
Some general commentary - about 25% of PIII trials had safety problems in addition to lack of efficacy. Excess deaths were on the order of 25% to 50% more deaths than placebo. This is not something that Dex could have seen in their trial - underpowered for that - and therefore a risk in the Ph III. What percent excess deaths would the FDA be willing to accept for, say a doubling in Good Outcome? Also note that many of these neuroprotectants seem to have bad effects in addition to their hopeful good effects. So especially if you administer it too late you don't get the saved lives and just get the extra deaths from, for instance, changed heart rhythm or lower BP.
More notes later, but in the meantime ... comments welcome.
Clark |
