Median overall survival across all trial patients is not cherry picking. If that's the best you can come up with you're pretty pathetic.