Mobile data collection's role in precision medicine and the clinical trial of the future

By Jonah Comstock
Share
Craig Lipset Pfizer Head of Clinical Innovation Craig Lipset

In 2011, Pfizer announced a novel clinical trial that would be fully remote, with every aspect of the trial being handled either online, over the phone, or by mail. But that trial was cancelled due to a failure to recruit enough participants, and a promised mulligan never emerged.

Craig Lipset, Pfizer's Head of Clinical Innovation, said at the BIO conference in Philadelphia yesterday that since 2011, Pfizer has incorporated many successful facets of that trial into its ongoing efforts -- and has put the failed aspects out there for other companies to learn from.

"For us, we took away a series of modules that worked and one that didn’t," he said. "That’s not meant to imply that online medication recruitment is a failure. In this particular protocol going after that particular patient population it did fail for us, but the other components did work. So our priority has been to take those components that did work, whether it’s electronic consent or other components, and scale those up. Because we’ve got 200-some clinical trials ongoing every day of the week, so how can we leverage what did work?"

He said he's been impressed with the other pharma companies that have made their own attempts at remote trials. Lipset didn't name any specifically, but notably Sanofi announced a fully remote trial in February.

"What didn’t work, we’ve socialized, we’ve published, we’ve shared in another form," Lipset said. "We shared it not just out of altruism, we shared it because, if it’s just a Pfizer thing, we’re not going to have support from regulators and other providers. We needed it to be more widely adopted. And I’ve been encouraged that there now have been at least three other pharma companies that have picked up where we left the ball on the field and have taken that farther down. And I’m optimistic that they’ll continue to share so that, as you said, it becomes more than just clever little pilots and much more of a mainstream tool."

Remote trials are becoming more important to pharma not just because of the cost savings, but because of the emerging conflict between personalized medicine and the traditional clinical trial infrastructure. Dr. Sanjiv Agarwala, chief of medical oncology & hematology at St. Luke’s Cancer Center in Bethlehem, Pennsylvania, described the future of personalized medicine as one where drugs will be made to order for individual patients based on their genetic markers -- a future where each drug treats so few patients that the clinical trial as we know it won't make sense.

"Regulatory needs to agree that [they] will approve a drug when you show that it works in a certain group of people because there is a genetic target and you hit it that it works, and [they] don’t need you then to take 500 patients and randomize half of them to that drug that you think has a pretty good shot at working and put the others on something that you know doesn’t work as well," he said. "We’re going to need to find a better way to test our drugs, treat the patients, collect the data post-treatment to understand and fine-tune it."

Lipset pointed out that remote trials could help in this area because connected devices collect so much more data than trials used to collect, and that data could help companies to identify the very specific phenotypes needed for a given trial.

"It’s really an evolution of [our remote trial] tools to not just create an opportunity to capture comparable data, but enable us to generate data of unprecedented variety and depth," he said. "We could use that with molecular data to find subgroups we didn’t have the resources to go after before."

Another group that is working on collecting large sets of data is the Joslin Diabetes Center. Joslin President and CEO John Brooks spoke at the conference about how Joslin is looking at ways to use that data to investigate the possibilities of precision medicine.

"The interesting thing about diabetes is we have this opportunity to really harvest this rich genetic information, but at the same time how do we couple that with all the environmental factors?" he said. "We know in many parts of the world air pollution, water pollution play a major role in the prevention of diabetes. But more importantly, we also understand that the aging process plays a significant role [as does] food, diet, exercise, and lifestyle behavior. And then the other element of diabetes, the fact that many people who have diabetes are now taking advantage of all these smart glucometers, CGMs, and all that data that’s being collected, wearable data, physiological data. How do we take that data, turn it into something that’s actually intelligible, and use that to give us better insights as we enfold that genetic information?"

There aren't any obvious solutions to the problem of how to investigate precision medicine drugs, but there are some promising leads. Computer modeling could eventually substitute live trials to some extent, or pharma companies could tap a variety of data sources to investigate the effects of drugs in real time or to easily find and recruit the right subjects for trials. All the speakers agreed, however, that the status quo wasn't going to cut it.

"We have models of clinical development that were built for a blockbuster era," said Lipset. "...We need new approaches that are able to help us identify patients more appropriately, and other ways of capturing and data. Whether it’s partnering with health systems or partnering with patients to aggregate and share data on their own. Because in today’s state in healthcare, we as patients have unprecedented access to our own data, and our own ability to serve as data hubs and aggregators of our own data. And patients have that willingness to share their data to support research. When you couple that with different molecular approaches, that creates tremendous opportunities for us to actually scale clinical trials."