In early phase clinical trials in cancer, efficacy of a therapy is often assessed by the response rate (RR) of the subjects exposed to therapy. If patients are not responding to therapy at a high enough rate should future subjects be exposed to therapy? If another therapy in the trial is showing greater response should more subjects be exposed to this therapy? In this phase II study design, Bayesian methods are utilized across multiple interim analyses to adaptively update the randomization allocation based on accruing data and form complex decision rules for early trial termination using predictive probabilities. The talk will be presented as a case study but will retain statistical rigor, presenting study design operating characteristics shown through simulation. Special attention will be paid to achieving desirable operating characteristics while monitoring futility and efficacy.