Support Grows for Wider Use of Real-World Evidence

The development and approval of more breakthrough therapies based on limited clinical trial data is heightening the need for more efficient methods of documenting treatment benefits in medical practice. At the same time, difficulties in designing and carrying out randomized clinical trials for promising medicines illustrates a need for innovative research strategies.

One response is to tap health system and insurance claims data systems for information on patient demographics, laboratory results and medical treatment to obtain clinical evidence that can augment the traditional clinical research system. FDA is working with sponsors to use real world evidence (RWE) to confirm results from initial clinical trials, particularly when product approval is based on limited studies. Such approaches are particularly useful for pharma companies seeking additional indications and tracking safety issues and rare events in drugs on the market.

RWE can help researchers generate new hypotheses, validate previous findings, support regulatory decisions and meet post-marketing study commitments, particularly in the development of new cancer therapies, according to experts at a conference on the utility of real world evidence in Washington, D.C. last week sponsored by the Friends of Cancer Research (FOCR) and the Alexandria Summit. The discussion is part of efforts to further develop breakthrough therapies for treating cancer and other serious conditions.

Biopharma companies are “still very tied to the clinical trial,” observed Jeff Helterbrand, senior vice president at Genentech. But difficulties in carrying out randomized studies for oncology are driving greater interest in new technologies to tap RWD, he said. A main problem is that in treating cancer, standard of care changes rapidly, often making a clinical trial “irrelevant” by the time a sponsor has gone through trial design and patient enrollment. Even registries become obsolete quickly due to fast-changing treatment standards, Helterbrand commented, while noting that RWE requires “sufficient rigor” to play a useful role in product development.

A main problem is that much RWE currently is highly variable, incomplete and unreliable compared to RCT results, and thus not considered sufficient to document safety and efficacy of a new therapy. This points to the need for linking data across health systems and payers, something that is missing today. The Patient Centered Outcomes Research Institute (PCORI) is working to move away from multiple data models by adopting for its research arm, PCORnet, the same data model used by FDA’s Sentinel system.

Janet Woodcock, director of the Center for Drug Evaluation and Research (CDER), identified a broad range of opportunities for tapping RWE to inform the performance of medical products. With marketed drugs, such data can be helpful in re-checking product performance and safety in approved uses, and also later in off-label uses. More advanced strategies may involve tapping data bases to learn about potential new uses prospectively, perhaps based on a hypothesis that then could be tested. Woodcock envisions future use of RWE to support investigational drug approval, but the system would have to bring in informed consent and study monitoring to be able to provide useful, generalizable information.

As part of efforts to support development of oncology therapies and orphan drugs, a working group organized by FOCR aims to design prospective pilot studies to test and validate data collection efforts and identify novel endpoints that correlate with clinical benefit, as described in a “Blueprint for Breakthrough” working paper [see

]. Collecting RWE more broadly may supplement post-market research, identify novel outcomes and avoid clinical studies that could expose more patients to less efficacious treatment. The bottom line is to spur development of new therapies mor