My understanding of the PR:
1. Trial did not meet primary endpoint.
2. But it is doing right for the critical subpopulation with a smaller sample size.
3. We have 24% mortality improvement. This is not said to be statistically significant. Why not, if it is 12 times better - according to Nader?
Note: We are not told the actual number of deaths, nor the distribution of this sub-group between placebo and LL. What if 50 got placebo and only 12 got LL? Or assuming 31-31 normalized split, what if 12 died from placebo and 9 from LL? It may or may not impress the FDA. But if in a larger trial, we get 40 in placebo and 30 in LL - would that be stat-sig or good enough?
We have stat sig in time to discharge. That is good. Gilead had sucked in the world's attention for this alone (although based on a fuller trial). Note we are not told that "time to discharge" is better for LL than placebo for the whole patient population; everything said is about the ventilation group.
We have improved probability (28% to 11%) for discharge-alive at day 28. Again, not mentioned as statistically significant.
This means if we were to do the trial again on this sub-population, we may get the same sort of results (though ideally results will improve if LL is actually impactful). Would that be good enough? Not clear how to interpret the improved numbers if they don't meet stat-significance; but for my untrained eye, they all look good to recommend drug for such near-death patients.
4. In effect, this phase 3 trial is turning out into a phase 2 trial that tells us what the key group our drug is most effective for.
5. BUT if we think back on all the "amazing" anecdotal data, it was all about critical patients "getting off echmo" or ventilation. So we could have focused on that group to begin with. We can say "Trial design" was not correct for our drug and we overreached.
6. FDA may still give EUA based on the 62 patients since we are talking about near-death people with no second option. Perhaps a phase 4 trial makes sense. The drug is safe; let it be used for ventilated patients
7. The news release is not giving us the hard numbers for either the full trial or for the relevant subgroup. We have to compute them based on the %. I wish they would just say the numbers straight; nothing to hide at this point. Or is there?
I am not upset or astonished by these results, because the trial is supporting the thesis that the drug is doing something significant in actually changing mortality. We got something to speak about rather than a dud. eInd and OLE are still there for a reason. The results are consistent with the anecdotal data for this subgroup.
We have a case to make that LL is effective on a particular route relevant to those who require ventilation whereas we may expect general mortality from Covid happens through multiple pathways and not all who die end up getting into ventilation. In a pandemic situation, people are desperate for exactly that subgroup for which our drug is doing something good.
Next questions are how will FDA, MHRA, HC respond and how will the stock price respond next week?
Market Data 
Markets 
AI
