“…There is often confusion about whether appropriate crossover constitutes confounding [30][31][32] -that is, whether receipt of the interventional therapy by patients in the control arm after progression inappropriately affects trial outcomes, particularly OS. But when a drug is known to have efficacy, this is not confounding, but rather the desired comparator arm of the study, both for the trial participants' sake and to ensure that the trial yields outcomes most helpful in informing clinical practice.…”