By Anne McNickle
This is an exciting time for those of us interested in RWE development. Digital technologies and advanced computing capabilities are enabling the generation of RWE continuously, and from sources that have traditionally been considered difficult or impossible to mine.
At the same time, regulators are strongly focused on RWE to inform drug development and regulatory decision making; in addition to statutory requirements (Cures law & PDUFA VI) to clarify and provide a clear path forward for the use of RWE, it is a top priority for FDA Commissioner Scott Gottlieb. And in fact, we have already seen FDA approve some drugs and devices based, at least in part, on certain types of RWE.
As Catalyst has been preparing to discuss RWE during the 10th DIA China annual meeting this week as part of a panel discussion alongside regulators, industry and academia, we wanted to highlight a few of the themes that have risen to the top for us.
First, use of RWE to help inform drug and device approvals is already a reality.
- On the drug side, CDER is focused on use of RWE in rare and ultra-rare diseases where enrolling enough patients in traditional clinical trials is difficult or impossible, and where there is unmet need
- For example, in April 2017, FDA approved BioMarin’s Brineura (cerliponase alfa) to treat a form of Batten’s disease, a rare inherited disorder that primarily affects the nervous system. Patient advocates credited the inclusion of registries as critical to this approval; FDA highlighted the fact that efficacy was established via a single-arm study that used a natural history cohort – a historical control group – rather than a traditional control arm.
- On the device side, Commissioner Gottlieb noted recently that FDA has approved or cleared “more than 8 new medical devices and expanded the use of more than 6 technologies” based on evidence derived from RWD.
- For example, Medtronic’s InPact Admiral, a drug-coated balloon for peripheral artery disease, received an expanded indication in 2016 using control data from a large national registry.
Second, FDA is engaged in a range of efforts to advance the field, for example:
- The agency is in the process of developing guidance on use of RWE in drug development, following issuance of a similar guidance for device development in August 2017.
- Under Cures and PDUFA VI, FDA will engage in a number of activities to advance the use of RWE, including: holding workshops; developing guidance; and developing a framework for evaluating use of RWE to support supplemental applications and post market requirements.
- Interestingly, FDA has also just initiated a new project with Harvard to try and validate RWE techniques, by comparing them to the outcomes of traditional randomized clinical trials.
- The agency has plans to establish new capabilities and tools to conduct “near-real-time” RWE evidence evaluation , and the oncology group at FDA is engaged in some interesting initiatives (see final section).
Third, sensors and wearables will vastly improve our ability to collect data and generate RWE in a continuous manner:
- The next big leap may come from exactly this area – making sense of the 24/7 data flow from things like sensors and wearables.
- With innovations such as the Proteus Digital Health/Otsuka collaboration, sensor technology is now a part of drug therapy for serious illness such as schizophrenia; thus, physicians and caregivers can now see how a drug is used in clinical practice (as we’ve blogged about).
- Sensors/wearables are also providing new insights into disease; the Michael J. Fox Foundation, for example, is doing some interesting work in this area to generate data on Parkinson’s disease that will be made available to the research community, in an effort to advance new product development.
- Companies are considering using sensors in their clinical trials; however, to include a sensor/wearable as a tool for data generation in a clinical trial, companies must ensure it is “fit for purpose” and that the data are objective.
Finally, artificial intelligence (AI)/machine learning will greatly improve our ability to collect and make sense of data:
- While there are many ways in which AI can be used to advance innovation in medical product development (as we noted in a recent blog), for purposes of RWE generation, AI is poised to open up hard-to-extract, unstructured data from EHRs and turn it into useable RWE.
- In fact, FDA has plans to do just this; the agency is seeking to develop the ability to conduct “near-real-time” RWE evidence evaluation “down to the level of individual electronic health records for at least 10 million individuals” in a broad range of healthcare settings, Gottlieb has said.
- Separately, as part of the oncology group’s INFORMED initiative, FDA is launching a fellowship program focused on developing and implementing machine learning/AI algorithms to help evaluate diverse data sets, including data from clinical trials, EHRs, and sensors/wearables.
The bottom line is that FDA is embracing use of RWE to help inform decision making across the medical product lifecycle, but the details of how and when RWE may be generated and used in drug development programs are still being worked out. If you are interested in exploring use of RWE in your drug development program, now is a good time to engage with the agency to ensure the appropriate tools and approaches are used, and that FDA will have a comfort level with the data for regulatory decision making.