Wednesday, November 10, 2010

Creating Opportunities in Drug Design: Applying 21st Century Technologies to Drug Safety Testing

Adam M. Clark, Director, Scientific and Federal Affairs

Advancing medical progress can sometimes be akin to putting together 3,000-piece 3d puzzle. Each piece is critical and must be in its appropriate place. In medical research, we have many of the most critical pieces such as technological achievements in molecular and cellular biology that are revolutionizing research capabilities and expanding our understanding of disease.
  • Once hailed as the pinnacle aspiration of clinical genomics, the $1000 genome now seems to be right around the corner.
  • The application of protein sciences, aptly named proteomics, has demonstrated clinical effectiveness in evaluating for heart damage, as in the measurement of troponin, and in identifying responders to targeted treatments like Herceptin in certain breast cancers.
  • Advances in bioinformatics are providing researchers the ability to integrate large and complex datasets and model molecular networks.
  • And just this month, researchers presented a novel breakthrough using RNA to reengineer human fibroblast cells to pluripotent stem cells in a manner that is almost 100 times more efficient than gene transfer and does not alter the cell’s genome, bringing about tremendous potential for using differentiated stem cells for cellular modeling of disease or treatment response.
And yet, we have yet to position these key pieces in their appropriate places, see how they connect so they can build toward the ultimate goal: to improve patient outcomes.
Consider that the number of new drugs being submitted to the FDA for approval has fallen from 53 drugs in 1996 to 19 just last year. Very few drugs (only about 8%) make it from pre-clinical testing all the way to market and among the reasons for a drug’s failure to move ahead early is toxicity concerns in animal models during early stage testing.
Recently, I participated in a panel that focused on this issue as part of the Brookings Institute and Friends of Cancer Research Conference on Clinical Cancer Research. The discussion focused on the value 21st century technologies could add to improving drug design, screening, and approval – values that benefit both the drug developers and the patients receiving promising new therapies. In an issue brief the panel put together to complement the discussion, we highlight two case studies (an integrated approach to organ injury and oncology drug-induced cardiovascular toxicity) that may help shed some light on real-world implications of safety assessments.
Among the key concepts from the panel and echoed throughout the day-long conference were:
  • Current in vivo models for drug toxicity testing have changed very little in decades and have not taken into account advances in molecular and systems biology.
  • Animal models were inadequate predictors of toxicity responses in humans, both in terms of generating evidence for biologically accurate mechanisms of drug toxicity, as well as identifying concerns for low-incidence, but dangerous, toxicity responses in our diverse human population.
Panelists proposed that the U.S. Food and Drug Administration should prioritize toxicity testing, noting that improvements in this area could increase the number of drugs making it into clinical trials, decrease the time in pre-clinical testing, and provide a more accurate and biologically-based profile of the benefit:risk ratio for new drugs.

We await anxiously where this conversation may lead and look to the FDA to start picking up these puzzle pieces - genomics, proteomics, bioinformatics, and other scientific breakthroughs – and begin piecing them together to improve patient outcomes. Drug toxicity profiling, as technical as it may be, is a critical piece to help complete the cure puzzle.

Adam Clark joined FasterCures in September to lead its scientific and federal affairs programs.

1 comment:

Anonymous said...

Straight to the point and well written! Why can’t everyone else be like this?