April 2020 Edition Vol.11, Issue 4

An Interview with Amy Abernethy, MD, PhD, Principal Deputy Commissioner, U.S. Food and Drug Administration (FDA)

Amy P. Abernethy, MD, PhD, Principal Deputy Commissioner of the FDA, helps oversee the agency’s day-to-day functioning and directs special and high-priority initiatives that cut across offices overseeing the regulation of drugs, medical devices, tobacco and food.  Her areas of expertise include cancer data, real world evidence, clinical trials, health services research, patient reported outcomes (PROs), clinical informatics, and patient-centered care.  She is also the FDA’s acting Chief Information Officer with responsibility for overseeing the organization’s data and technical vision.

A hematologist/oncologist and palliative medicine physician by training, she is an internationally recognized clinical data expert and clinical researcher.  Prior to her FDA post, Dr. Abernethy was Chief Medical Officer at Flatiron Health, a company created to accelerate cancer research and improve cancer care through technology. At Duke University School of Medicine, where she was a professor of medicine before joining Flatiron Health, she pioneered the development of technology platforms to spur modern advancements in the care of people with cancer and other serious life-limiting illnesses.

In this interview, OBR asked Dr. Abernethy about initiatives at the FDA to accelerate how technology is being used to individualize strategies for treatment in cancer and other diseases.

OBR: How is the FDA participating in the increasingly high tech approach to precision medicine in oncology and other therapeutic areas?

AA: The growing emphasis on precision medicine in oncology is a living example of where healthcare is going overall. Targeted therapy, based on a better understanding of the underlying biology of specific cancers, permits us to streamline the evidence required for regulatory decisions. The trend toward more tailored treatment for well-defined targets increases confidence in decisions made at earlier stages in clinical trials and in smaller patient populations.

As we follow the science and data that supports targeted treatment, we can also take advantage of new FDA regulatory pathways, such as the Breakthrough Therapy designation, which is designed to expedite drug review. In addition, once a drug is approved, we are increasingly using novel data sources of information, such as electronic medical records, to understand how drugs perform post approval. These are changes that represent an evolution in the FDA regulatory process that has largely played out in oncology first, but I think we will be seeing this moving into all therapeutic areas.

OBR: What is it about cancer that has put it in the forefront of precision medicine?

AA: In the realm of precision medicine, particularly in oncology, it is no longer just one drug, one disease. A disease like lung cancer, for example, is sub-segmented into many different diseases based on differences in biology and molecular targets. At the level of the individual patient, stage and prior treatments are also relevant to a patient-centered precision approach. All of this has accelerated individualized decision-making in cancer care. The patient-centered approach is highly relevant to the regulatory process, which the FDA is adapting in order to keep pace with the evolving science.

The FDA’s Oncology Center of Excellence is one example of a unique initiative to ensure that we are enabling twenty-first century treatment. By fostering a collaborative environment, the Center was created to expedite development of diagnostic tools, drugs, biologics, and devices for oncologic and hematologic malignancies. I think this will be a proving ground that will help us learn how to apply the same principles in other areas, such as diabetes or Alzheimer’s disease.

OBR: What role do diagnostic tools play in advancing precision medicine?

AA:  Diagnostic tools and clinical decision support are critical to precision medicine, and the FDA has an important regulatory role in ensuring these are safe and effective. With all the current expectations around efficient decision-making in the delivery of precision medicine, we have a particular interest in helping physicians make decisions on the fly at the point of care. In the area of decision support, tools are becoming increasingly sophisticated but increasingly complex. We need to ensure that the software leveraging clinical data produces accurate and reliable guidance. We are interested in encouraging tools that can align with practical issues, such as reimbursement, that are important to treatment choices.

OBR: How is the FDA tracking health status through the use of digital tools?

AA: In an increasingly digital world, we are very concerned about encouraging tools that can be updated quickly to evolve with new information. We are pursuing several strategies. One example is the Software Precertification Pilot Program, called Pre-Cert. The concept is to identify digital health technology companies that have demonstrated a culture of quality and rigorous organization and work with them closely. Rather than focusing on the product, which is the approach with typical devices, we will look at the process. Essentially, the idea will be to crosscheck the data, the computations, and the algorithms during the development process so that the software can be adapted quickly in the event of glitches or new information. This is one example of the types of novel approaches we are developing, but there are others.

OBR: Where does the FDA stand with regard to telemedicine?

AA: As you know, the Centers for Medicare and Medicaid Services (CMS) expanded reimbursement for telemedicine services several weeks ago. This was a very forward-leaning action in my opinion. It certainly acknowledges that need to adapt healthcare strategies to reduce exposure to covid-19, but it also will allow us to protect many FDA initiatives. Not least of these, we can continue to monitor patients participating in clinical trials while maintaining social distancing. This is particularly important in oncology, where clinical trials are often the treatment of choice. The expansion of telemedicine might be one of the good stories that emerge from the covid-19 pandemic.

OBR: From the FDA’s point of view, how does real world data inform clinical decision making?

AA: Much of my career has been devoted to leveraging real world data to inform clinical decision making, so I am an enthusiast, but I am also a pragmatist. I think that real world evidence has promise as a complement to traditional clinical trials. Neither one will take the place of the other. All data, including data from clinical trials, has warts. From the FDA lens, I think we are still working to determine how data from different sources fit.

Certainly, real world evidence has been attractive for looking longitudinally at safety signals of approved drugs. One current example in oncology is the effort to follow the long-term safety of checkpoint inhibitors. There is also a great deal of interest in using real world data to inform the design of clinical trials, including the identification of eligibility.

As a substitute for a clinical trial, real-world data has been used for regulatory purposes when randomization is unethical or impossible. Real world data have also been used in selected cases for label expansion. However, the application of real world data requires a thoughtful, deliberate, and analytical plan. There are large amounts of real world data available but no fancy button to push that eliminates the problems.

OBR: What can you tell us about FDA initiatives in artificial intelligence?

AA: The potential for AI is so great that the FDA could not maintain our commitment to innovations in public health without helping to bring forward the potential applications. These are moving forward quickly in pathology and radiology, but I expect AI to play a role in drug development, in designing clinical trials, and in data analysis, including new strategies to learn from real world evidence. There are challenges for the FDA in determining how best to monitor and regulate AI that are analogous to the work we are doing to work with software companies innovating in clinical decision support, but we will meet these challenges. It brings me back to the initial discussion about precision oncology. It is not difficult to imagine how AI is likely to enhance almost every step toward greater optimization of individualized medicine.

OBR: We understand that you have an active role in the Technology Modernization Action Plan (TMAP) outlined last September. Can you tell us more about this?

AA: The Technology Modernization Action Plan or TMAP is an ambitious strategy that includes modernizing our technical infrastructure in ways that allow us to receive, analyze, and use data in new ways to advance the FDA’s regulatory mission. As part of the TMAP, the FDA is expanding its capability to develop technology products through a series of projects that demonstrate new ways of applying technology and data to the FDA’s regulatory mission.  These projects will span the FDA’s broad regulatory responsibilities, from supporting innovation in medical product development to bringing new ways to ensure food safety and increase the effectiveness of the FDA’s field operations.

Post a Comment

OBR Archives

To view previous issues of OBR green you can visit our archives. The entire library of OBR green articles is searchable.