Randomized Phase II Cancer Clinical Trials PDF

Preface

A clinical trial is an important research tool in the evaluation of the efficacy and safety of cancer therapies. Cancer therapies include anticancer drugs, gene therapy, surgical procedure, radiation therapy, and combinations of these. An experimental therapy is evaluated through three phases of cancer clinical trials. A phase I trial has the objective of selecting the appropriate dose level of an experimental therapy (mostly an anticancer drug) to be used for further investigation. Once a dose level is chosen, the experimental therapy is tested through a small phase II trial requiring 50 to 100 patients. An experimental therapy that is shown to have a promising activity in phase II is evaluated by a large-scale phase III trial compared to a current standard therapy. Since evaluation of an experimental therapy through these clinical trials is limited to a specific type of disease, a single experimental therapy may go through multiple phase II and III trials for different disease types, especially when it is shown to be very efficacious from earlier trials.

If a new cancer therapy is shown to be inefficacious from a phase II trial, it is very likely that no phase III trials will be conducted for further evaluation. In this sense, phase II trials are called screening trials. Among the three phases of cancer trials, phase II trials usually comprise the largest volume of cancer research activity for individual cancer centers and cooperative cancer trial groups. In order to expedite the conduct of phase II trials, a traditional phase II trial has been designed as a single-arm trial to treat all the patients using the experimental therapy to be compared with a historical control. This simple phase II trial design has resulted in many issues, including increased false positivity of phase II trial results and many negative phase III trials. Pointing out these issues, many oncologists and biostatisticians began to propose to use a randomized phase II trial to compare an experimental therapy with a prospective control therapy.

It is critical to use accurate statistical methods for designing and analyzing phase II trials. Because of the small sample sizes, exact statistical methods have been used for phase II clinical trials. This book is intended to provide diverse statistical design and analysis methods for randomized phase II trials in oncology. Since a large part of methodologies for randomized phase II trials stems from those of single-arm phase II trials and many phase II cancer clinical trials still use single-arm designs, we also review the statistical methods for single-arm phase II trials in Chapters 1 to 5. This book will be useful for cancer clinicians as well as biostatisticians.