Preclinical to Human Translation: Correctable Errors?

• Activity Number: 343
• Type: Topic Contributed
• Date/Time: Tuesday, August 6, 2013 : 10:30 AM to 12:20 PM
• Sponsor: International Chinese Statistical Association
• Author(s): Mandy Bergquist*+
• Companies: GlaxoSmithKline
• Keywords: translation ; drug discovery ; preclinical ; communication

Abstract

In Phase 2 clinical trials, over 50% of new projects fail for lack of efficacy. The standard response is that preclinical models do not represent human disease adequately. While there are several challenges in translating preclinical efficacy to humans, statistical errors in the preclinical studies may make translation virtually impossible. A recent publication in Nature (Begley, CG and Ellis, LM (2012)) concluded that the lack of rigor around the generation and analysis of preclinical data today is similar to the situation in clinical research about 50 years ago. Reviewing the current state of design, analysis, and interpretation of preclinical pharmacology studies highlights the challenges as well as intriguing possibilities. The routine application of statistical concepts could dramatically improve translation, resulting in tremendous gains for drug research. The difficulty lies in getting from here to there - a task that likely requires more communication and influence than computational skill.