Ann: Mesoblast Update on COVID-19 ARDS Trial, page-435

Looking at the protocol on clinicaltrials.gov I see this trial started 30 April 2020.

So I imagine patients joined at intervals until recently. The data should allow for discovery of improvements in maximal care (without cells) over that timeframe and against the base set that made the trial seem worth doing in the first place.

I don't hold MSB but I want MSCs to work (so I'm biased probably to making excuses rather than concluding the whole MSC story is just wrong or has been oversimplified) however, maximal care as a comparator when maximal care is going to be quite naturally improving with clinician knowledge over the testing period seems like a bad confounded test to me.

Seems like the 30 day survival rates of patients in April (no cells at all) might naturally be better than that in the early period of Covid and that May might be better than April and June better than May etc - just because doctors are learning and passing their knowledge to their colleagues. Doctors trying to help patients would naturally be learning - they'd be motivated. I just don't think the primary endpoint of all-cause mortality (given a learning rate) is fixed enough.

MSB seems to be spending a huge amount of human time and effort but running into statistical or study design problems.

With Dream HF its like the wrong primary endpoint was chosen - MSB was too cautious. In this trial its like they MSB was wildly optimistic in hoping to see within 30 days better than a 43% improvement in all-cause mortality when all the patients were getting maximum (and improving over time) non cell treatment.

Seems like MSB is doing a heck of a lot of bleeding on the edge.