There were some questions about effect size on the conference...

There were some questions about effect size on the conference call, but this can be calculated from what information we have, especially as we now know that the control group mortality is likely to be in the region of 60%-80%.

Effect size required for whole-cohort trial success
A tool such as G*-Power (https://stats.idre.ucla.edu/other/gpower/) will calculate the required effect size in the Remestemcel-L group to detect a difference with 90% power as indicated today in the conference call.

Assuming they are considering the all cause mortality proportion in each group and using something such as Fisher's Exact Test https://en.wikipedia.org/wiki/Fisher%27s_exact_test, we can calculate the following:

Assuming α < 0.01 is considered a whole-cohort trial success, with 150 in the control group and 150 in the Remestemcel-L group:

• with 80% mortality in control group, 41% (62/150) must survive in the Remestemcel-L group
• with 70% mortality in control group, 52% (79/150) must survive in the Remestemcel-L group
• with 60% mortality in control group, 63% (95/150) must survive in the Remestemcel-L group
• with 50% mortality in control group, 72% (109/150) must survive in the Remestemcel-L group
• with 40% mortality in control group, 81% (122/150) must survive in the Remestemcel-L group

Effect size required for early trial success due to Overwhelming Efficacy

This is more tricky as there will not be 90% power in the n=90 person trial. So let's do it the other way around and calculate the power for different control group mortality rates for an alpha of >0.001.

Model 1: 60% all-cause mortality in the control group
What is the chance that the trial will be halted due to overwhelming efficacy after 90 patient readout?


This graph tells us the chance of trial success at different proportions of Remestemcel-L group all-cause mortality

• If 90% of Remestemcel-L group patients survive, then there is a 96.6% chance that the trial will be halted
• If 80% of Remestemcel-L group patients survive, then there is a 68.1% chance that the trial will be halted
• If 70% of Remestemcel-L group patients survive, then there is a 26.2% chance that the trial will be halted
• If 60% of Remestemcel-L group patients survive, then there is a 5.2% chance that the trial will be halted

Model 1: 80% all-cause mortality in the control group
What is the chance that the trial will be halted due to overwhelming efficacy after 90 patient readout?

This graph tells us the chance of trial success at different proportions of Remestemcel-L group all-cause mortality

• If 90% of Remestemcel-L group patients survive, then there is a high% chance that the trial will be halted
• If 80% of Remestemcel-L group patients survive, then there is a 99.6% chance that the trial will be halted
• If 70% of Remestemcel-L group patients survive, then there is a 93.1% chance that the trial will be halted
• If 60% of Remestemcel-L group patients survive, then there is a 68.1 chance that the trial will be halted

Model 3: 50% all-cause mortality in the control group
Let us now take a more conservative view and look at 50% mortality.

This graph tells us the chance of trial success at different proportions of Remestemcel-L group all-cause mortality

• If 90% of Remestemcel-L group patients survive, then there is a 79.1% chance that the trial will be halted
• If 80% of Remestemcel-L group patients survive, then there is a 31.3% chance that the trial will be halted
• If 70% of Remestemcel-L group patients survive, then there is a 5.8% chance that the trial will be halted
• If 60% of Remestemcel-L group patients survive, then there is a 0.5% chance that the trial will be halted