Re: review of IMbark ClinTrials, conclusions and questions...
Median overall survival means the median duration of survival, not the survival of half the enrolled patients. So if for example (totally made up numbers) they had survival on treatment for 10 months for patient A, 15 months patient X, 21 months for patient Z, 23 months for patient B, and 24 months for patient C; median overall survival for this subset would be 21 months. I don't know how many patient deaths they need to project median OS for the whole group, because I don't know what statistical model they would use to project the number. It probably depends on the variability in length of survival and they would need to be able to measure 2 std deviations from a bell curve of responses. I really just don't know what minimum number of patients they need and how they would project he median OS. But it doesn't mean they need half the enrolled number.
The answers to your other questions are out there if you do a little searching. I'm afraid I would incorrectly answer if I tried, so I'm just going to continue to watch the ticker. Whenever I get a minute at work.