So you have two groups (test and control) with 600 and 400 people respectively in them. You have an exposure variable which is equal in proportion for both groups and you want to see if it explains median survival (as a binary variable) and response rate (please better define this).

So it sound like you will be running a survival analysis, correct?