Clearly,
p-values are going to need to be impressive--but how significant is "significant" in the world in which we are living and under the cloud that the FDA is operating, and will "poor" results for mortality differences really be fatal to a finding of overwhelming efficacy?
Let's remember:
- what we currently have as "FDA approved" emergency-treatments for COVID: Only 2 - Remdesivir and now convalescent plasma. The benefits of both are still open to question, the former only apparently provides at best a few days difference in early recovery at very early stages of COVID, NOT in relation to ARDS.
and
- what we currently have as FDA-approved TREATMENTS for COVID: NADA. ZIP. ZILCH. A BIG GOOSE EGG.
https://www.fda.gov/drugs/coronavirus-covid-19-drugs/coronavirus-treatment-acceleration-program-ctapHere we are, nearly 9 months into this scourge, and NOTHING has been deemed effective as a treatment. Trials ongoing, sure, but many have already failed and it's likely many more will.
So will the FDA really hold MSB to a requirement of
p <.0001 or even <.001 on mortality differences if people are clearly getting off ventilators more quickly (where time on ventilator may not be a death sentence, but every day is potentially incredibly harmful), are recovering more quickly, and are getting out of the hospital more quickly? If their levels on all other tested criteria are all significantly lower?
In other words, will MSB's treatment be penalized for across-the-board increases of SOC reducing mortality OVERALL?
I think significant differences on all of the below, non-mortality-related, yet still key tested values, will end up being as or more important under real-world SOC improvements, where each of these values are tested at baseline, 7, 14, 21 and 30 days - 5 separate points for Rem-L to show its working--and working in a way that is undeniable versus the placebo:
- Number of participants with resolution and/or improvement of ARDS;
- Change from baseline of the severity of ARDS;
- Length of stay (tested to 12 months, but discharge dates will clearly still be of interest and almost undoubtedly reported for all recovered patients within the initial 30 days for the 1st interim readout);
- Clinical Improvement Scale;
- Change in serum hs-CRP concentration;
- Change in IL-6 inflammatory marker level;
- Change in IL-8 inflammatory marker level; and
- Change in TNF-alpha inflammatory marker level.
That's a heady list, and if MSB's experience in AGVHD, and in the EAP, are any guide, we should see REMARKABLE differences on these points between those treated with Rem-L versus those only given the placebo.
I don't see the FDA being able to ignore stellar, non-mortality-related criteria under these circumstances, and can see a mortality difference significant at a
p-value of only, say, < .05, being entirely overshadowed by significant results in some or even all of the other secondary outcome measures, with Rem-L still being approved based on those results, at least for emergency use.
Such results would simply be too good to ignore. It would be cruel for the FDA to force patients in the trial to continue suffering on ventilators for days or even weeks after Rem-L patients have been discharged, even if those patients are likely to eventually limp home out of the hospital...with incredibly scarred lungs. Alive, yes. But at what cost? That's to say nothing of refusing to offer the same treatment to all Americans, beyond those at least fortunate enough to have a 50-50 shot in the trial.
I think MSB has a great shot here at this 1st interim readout, mortality rates notwithstanding.
(20min delay)