Tackling the Bottlenecks in the Drug Development Pipeline

Can you believe the average length of time from target discovery to approval of a new drug currently averages about 14 years? That is WAY too long. Even more shocking is that the failure rate exceeds 95 percent, and the cost per successful drug surpasses $2 billion, after adjusting for all of the failures. The National Center for Advancing Translational Sciences was specifically established one year ago to apply innovative scientific approaches to the bottlenecks in the pipeline. An example of game-changing innovation is the NCATS collaboration with the Defense Advanced Research Projects Agency (DARPA) to develop a biochip for testing drug safety. Devices like this and other tissue chips may someday reduce the amount of animal and human clinical trials necessary to determine if a drug works. That could be a huge step toward making drug development faster and cheaper—which is better for all of us.

3 thoughts on “Tackling the Bottlenecks in the Drug Development Pipeline

  1. The problem of translational disconnect is particularly acute in CNS [Central Nervous System] disorders where less than 8% of drugs that enter clinical trials make it to approval; in the case of Alzheimer’s disease this is only about 3% (only one in every 34 projects is successful – source PhRMA). With the emergence of all CNS Diseases as a major economic burden to the society (expected costs 900 billion/year – PwC study), we definitely need new out-of-the-box approaches for treating brain disorders. One possible solution is Quantitative Systems Pharmacology that combines preclinical neurophysiology with the vast amount of human clinical and imaging data in a ‘humanized’ computer model that can be calibrated using retrospective clinical trials. As an example a recent publication describes the blinded prediction of a phase II trial in schizophrenia, where the model correctly identified a substantial motor-side effect that was completely missed by the animal models, which was never done before. (Blinded prospective evaluation of computer-based mechanistic schizophrenia disease model for predicting drug response. Geerts H, Spiros A, Roberts P, Twyman R, Alphs L, Grace AA. http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0049732). While this approach is certainly not perfect, it offers a fresh new avenue for addressing the current drought of new CNS Discovery and Development projects.

  2. Ridiculous. It’s not how long, but how sensitive and precise the testing is that gives the net cost/benefit of a new drug. It will be at least 14 sequential Francis Collins careers before a chip performs as well as the best mammalian experiments & human trials. Are you asking for another Vioxx (et al.)? SSRI efficacy?

  3. Your posting about drug development is very interesting and noteworthy, I have learned new things from your posting, which will help me in my work process.

Comments are closed.