Fostering such solutions is the goal of a newly launched FDA initiative, the Critical Path, which refers to the route from laboratory prototype to medical treatments. The initiative aims to identify the industrys largest stumbling blocks that cause drug and device companies to abandon late-stage projects, and to develop tools so that dead-end projects can be scuttled before years and millions of dollars are invested in them.
The project is unusual because other FDA efforts to speed development have focused on subpar bureaucracy rather than underdeveloped science.
The report says that basic science has far outpaced the means to assess it, stalling or derailing potential new therapies.
"Not enough applied scientific work has been done to create new tools to get fundamentally better answers about how the safety and effectiveness of new products can be demonstrated, in faster time frames, with more certainty, and at lower costs. In many cases, developers have no choice but to use the tools and concepts of the last century to assess this centurys candidates," the report said.
The report says that the FDA has a unique vantage point to guide the process because it reviews confidential data from failed and successful products. Industry groups representing biotechnology and pharmaceutical companies welcomed the initiative.
However, under questioning from Science Board members, Woodcock acknowledged that the FDA does not have the resources to create and validate predictive toolkits on its own. The agency plans to create a Critical Path Opportunity List and to use it to foster an "aggressive, collaborative" effort to move predictive tools forward.
The effort will rely heavily on next-generation information technology. The report specifically calls for computer models that can predict the effects of device modifications and so allow small changes to be adopted without requiring human testing.
Biomarkers to predict a drugs efficacy are particularly important. For certain diseases, clinicians can use indirect measures to argue that a drug will work. For example, for cancer the FDA will often approve a drug that causes tumors to shrink or even to stop growing, even if there is no direct evidence that the treatment help patients live longer. Thats because a drugs effect on tumor growth can be known long before its effect on patient survival, and quicker approval gives more patients access to the drug.
Most biomarkers are much more subtle than tumor growth; typically they detect a protein, genetic variation or other molecule: Vaccine trials measure antibody levels; some specialized cancer drugs are prescribed after certain proteins or active genes are detected in tumors. Woodcock said without common quantitative biomarkers that predict a drugs efficacy for both animal models and human subjects researchers "are floundering around in the dark" until their device or drug reaches large clinical trials. But, she said "we have very few of them and there are not that many on the horizon."
Even for late-stage trials, a biomarker may become increasingly popular with researchers, but still be insufficiently validated for clinical trials. "Who is going to conceive and conduct research that would actually close these [knowledge] gaps, and make this into a biomarker that developers can rely upon?" Woodcock asked. She also cited a need to standardize clinical infrastructure "so that we have standardized case-report forms and data-collection and data-format standards for different clinical trials so that developers are not constantly inventing new case-report forms, new data standards." Such standardization could have a huge impact on companies providing services for e-clinical trials.
The first step, said Woodcock, is to draw up a list of the biggest barriers and determine which can be most easily removed. The FDA has established a public docket to collect feedback and suggestions through July.