Latest News

Integrating Clinical Research And Care In A Perpetual Trial

June 25, 2019 | Mika Newton, CEO of xCures, believes R&D is about to hit a wall. Molecular subtypes and oncology drugs are exploding, and there just aren't enough patients to explore all the potential options.

Newton advocates for a perpetual trial—a combination of real world data collection and artificial intelligence platform that can provide patients and their physicians with individualized treatment options. Real-world data that is generated longitudinally helps inform what works in which patients, and can enable us to coordinate treatment prioritizing the most promising therapies.

On behalf of Diagnostics World News, Marina Filshtinsky spoke with Newton about how AI and real world data can work together.

Editor's note: Filshtinsky, a Conference Producer at Cambridge Healthtech Institute, is planning a track dedicated to Companion Diagnostics for Immunotherapy at the upcoming Next Generation Dx Summit in Washington, DC, August 20-22. Her conversation with Newton has been edited for length and clarity.

Diagnostics World News: Many companies are currently working on harnessing the power of Real World Data (RWD) for clinical research and precision medicine. Please tell us a little bit about your company and what is unique about your approach.

Mika Newton: xCures is uniquely focused on using Real-World Data (RWD) collected in a new type of study called a Perpetual Trial to improve outcomes for patients, as well as slash the time and cost of drug development. Our AI-based platform tightly integrates clinical research and care to continuously learn from all patients on all treatments. Our methods align incentives to the patient in a way that creates unprecedented value for all parties involved in both oncology research and care.

xCures' goal is to develop optimal regimens of tests and drugs that give patients the best possible outcomes. There are far more plausible regimens than can be tested in traditional trials. That's why we are running a Perpetual Trial to generate RWD and use AI to guide the experiments that can run efficiently on this platform.

We know that companies that make their products easily accessible to this approach will gain enormous value from a system that optimizes the use of their products in the context of the patients who need them.

How can AI be applied to the generation of RWD in the field of oncology?

Many companies are using AI and Machine Learning to try to understand "big data." This may be useful in understanding what has been done in the past, but it doesn't give us data on the newest treatment options or on novel combination therapies. The "next level" for the use of AI isn't in analyzing big data, but rather in telling us what data we need to advance knowledge, and to plan optimal experiments for acquiring that data with the fewest possible patients.

The use of AI in this way will power an "air traffic control system"—including garnering results from myriad investigator-initiated trials and investigator-sponsored trials—and from this, we can optimize the usage of the most valuable and limited resource we have for research: the patients.

"Perpetual Trial" is a new term; can you explain what it means and how this model can be applied on larger scale?

A Perpetual Trial is an adaptive platform (or umbrella) trial taken to its logical extreme: all patients, all treatments, all the time. As its name implies, there are no endpoints—there's an ongoing effort improve treatment regimens and outcomes—and the scale has virtually no limits.

New prospective hypotheses can be injected into the system where they are vetted by experts and used accordingly. Once a new "option" has been tried it enters an optimization cycle where if it is successful it will be used by more and more of the appropriate patients or if it is unsuccessful it will be eliminated for the types of patients it does not serve. New methods have been developed to interrogate and visualize data for signals with enough evidence behind them to produce packages for use in regulatory filings and dissemination to improve patient outcomes.

As we all know, biopharma companies are working tirelessly on new cancer therapeutics and diagnostics companies are trying to come up with new companion and complementary assays. How can these two industries work together to take advantage of RWD and pragmatic and perpetual trials?

Optimally, they would work best if they aggregated their data. There's little advantage in individually trying to run studies in which the utility of their individual technologies—when combined together—aren't well understood.

Combining RWD with real world experiences of patients is the future of care optimization and the development of new tests and treatments.

Observing real world usage is the only way in which the value of combinations of tests and treatments can be realistically understood. The cost of a clinical trial to understand general utility in this way is just astronomical, so nobody will do it. The connection between diagnostics and the treatments they lead physicians and patients to pursue must be observed in the context of treatment. Integration of research and care then enables all parties to work quickly to make the most of the knowledge they are gaining.

When companies set up, or participate in, an appropriate mechanism for data collection and aggregation of RWD, they are able to understand what happens in the real world. Consequently, real world evidence offers the opportunity for a company to put their product into the marketplace and see emerging trends of both the perception/use/utility of technology and the context in which its value is maximized.

Real world evidence is rapidly emerging as the most efficient way to strategically understand the marketplace, communicate the value of products in the context of current care pathways, and to get regulatory approval for novel and incremental uses of products.