Yes! Now you're getting to the juicy stuff about this analysis!...

Yes! Now you're getting to the juicy stuff about this analysis! You're right that if you are willing to make some assumptions based on available evidence then there are some things that a "continue" result can tell us about the results of the trial. So, if we look at this table:

Results:
30% Readout

			35%	40%	50%	60%	65%	70%	75%	80%	85%
		15% sceptical	0.186	0.24	0.352	0.479	0.55	0.628	0.712	0.799	0.882
		15% uninformed	0.854	0.939	0.999	0.999	0.999	0.999	0.999	0.999	0.999
		20% sceptical	0.115	0.158	0.253	363	0.428	0.501	0.584	0.676	0.776
		20% uninformed	0.651	0.807	0.957	0.993	0.997	0.999	0.999	0.999	0.999

Provided you accept the assumption that patients receiving the cells survive 15-20% of the time, and that the desired measure for improvement is 10% as per other examples cited in the original post, and if we believe Mesoblast when they say mortality in the control group is > 60% then we could infer form this table that a sceptical prior is in use (though it may be more or less sceptical than the one I got form the white paper), because all the "green square" results are impossible now that we know there was not overwhelming efficacy in the 30% readout.

So then if we look at the 45% readout table:
45% Readout

			35%	40%	50%	60%	65%	70%	75%	80%	85%
		15% sceptical	0.277	0.375	0.563	0.735	0.811	0.878	0.931	0.968	0.989
		15% uninformed	0.901	0.971	0.993	0.999	0.999	0.999	0.999	0.999	0.999
		20% sceptical	0.151	0.228	0.399	0.578	0.669	0.756	0.837	0.905	0.956
		20% uninformed	0.683	0.856	0.982	0.999	0.999	0.999	0.999	0.999	0.999

we would need 15% treated group mortality, and 80% patient mortality to get a result. However... when you look at the 60% table:

60% Readout

			35%	40%	50%	60%	65%	70%	75%	80%	85%
		15% sceptical	0.373	0.511	0.742	0.895	0.942	0.973	0.990	0.997	0.999
		15% uninformed	0.932	0.985	0.999	0.999	0.999	0.999	0.999	0.999	0.999
		20% sceptical	0.188	0.303	0.545	0.757	0.841	0.907	0.953	0.981	0.994
		20% uninformed	0.708	0.890	0.992	0.999	0.999	0.999	0.999	0.999	0.999

We would only need 70% control group mortality to get a result with 15% treated mortality, or 75% control group mortality with 20% treated mortality to get a result. This is what makes me more confident of seeing overwhelming efficacy at the 60% readout than the 45% readout. The risk for overwhelming efficacy at the 45% readout is more about the mortality rate in the control group (we would need to see > 80%) than the mortality in the treated group.

The problem I have for giving even better modelling on what to expect is that we don't know the details of the prior that they are using. Can someone ask them on twitter or email?