Single arm designs, in which all patients are given a potential new cancer treatment (i.e. with no control arm), have commonly been used in oncology Phase II studies. These studies often specify tumor response (shrinkage) as the primary endpoint. Rationale for such single arm designs relies on the expectation that tumor response is directly attributable to treatment effect. Conclusions in such studies are based on comparison with historical data, or in some cases, based on there being no effective alternative treatment (thus any response is clinically meaningful). However, several issues challenge the justification for utilizing single arm designs. Due to changing diagnostic tools (MRI, CT-scans, etc.), patients in a given category of disease state now may include very different patients (more or less healthy, etc.) than were included in that category several years ago when the histo