Phase I Studies: Goals, Patient Selection, and Implications for Design

• Activity Number: 73
• Type: Contributed
• Date/Time: Sunday, August 1, 2010 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Paul H. Frankel*+ and Stephen Shibata and Susan Groshen and Jeffrey Longmate and Stella Khoo and Edward Newman
• Companies: City of Hope and City of Hope and University of Southern California, Norris and City of Hope and City of Hope and City of Hope
• Keywords: Phase I ; Phase 1 design ; Phase 1 ; Cancer ; Dose escalation ; survey

Abstract

Background: Phase I trials in cancer are often intended to select a dose of a toxic therapy, assuming an unknown but monotonically increasing drug-toxicity and drug-efficacy curve. Statistical designs are often advocated for this situation without a thorough investigation of the clinical processes and goals. Method: A brief anonymous survey was sent to 670 oncologists at major US research institutions. Results: 19 % responded. 66% stated that moderate toxicities would alter subsequent patient selection, conflicting with the statistical assumption of random patient selection. Additional responses suggested that usual Phase I targets and designs, while statistically and operationally convenient, are inconsistent with the clinical intent of the research. Conclusion: Understanding the goals and behaviors of today's Phase I investigators is critical for proper Phase I designs.