Breast cancer surgeon Dr. Laura Esserman sings to her patients as they go under anesthesia. She tackles any song request, whether it’s a top 40 hit or a Broadway ballad. This same patient-centric attitude drove Dr. Esserman to participate in adaptive clinical trials, a game-changing way to test new medications.
It’s past time that other researchers think beyond traditional clinical trials. Adaptive trials can make drug testing both more efficient and accurate. If regulators were to encourage greater use of adaptive trials and facilitate sharing of electronic health record data, researchers could deliver better drugs to patients, faster.
In a traditional clinical trial, researchers plan out every element of the trial — from the number of participants to the type of data to be collected — before they begin testing. They stick to this rigid master plan until the trial is complete.
But in an adaptive trial, researchers pre-plan certain modifications that they can make part-way through the trial, based on the results they’ve uncovered so far.
Think of it in terms of mapping a run. A runner might look at a map and realize that at one point, the road will fork. Instead of deciding whether to go left or right before he ever begins jogging, he might choose to postpone the choice until actually reaches the fork. Perhaps he’d like to observe which path is muddier or which path has less traffic — and he’ll only know when he gets there.
Similarly, researchers administering adaptive clinical trials can modify their tests as they make observations. Like the runner, they have to pre-plan what choices they’ll make and when. But the more flexible trial formula enables them to alter the trial in response to real-world results.
Dr. Esserman’s trial, for example, personalized testing by splitting breast cancer patients into different groups depending on various measurements of their health. She tested a combination of therapies on these patients — and only continued testing those that were found initially effective. By not having to run a dud treatment through the course of a standard clinical trial, the adaptive design reduced the cost, time, and number of patients needed for the trial.
The FDA currently permits adaptive trials in limited instances. Letting researchers use this model more often would improve the drug development process.
Another way to improve the process is to enable doctors, insurers, and drug companies to share data on patients’ health outcomes after they take new FDA-approved medicines. By analyzing this data, they could uncover patterns that can’t be detected even in the largest clinical trials.
For instance, a clinical trial for a diabetes drug might include 1,000 participants, 10 of whom are Native American. Perhaps 60 percent of all patients respond well to the drug, but all 10 Native Americans get their blood sugar under control thanks to the medicine. It’s impossible to tell from such a small sample size whether the medicine really is vastly more effective for Native Americans, or whether those 10 patients just got lucky.
Changing regulations to permit greater sharing of such data would help researchers unveil rare side effects, complications, or “miracle” results that crop up.
Adaptive trials and increased data sharing would deliver better medicines to patients faster and with greater safety. That’s something to sing about.
Peter J. Pitts, a former FDA Associate Commissioner, is the president and co-founder of the Center for Medicine in the Public Interest. CMPI receives some funding from biopharmaceutical firms, which could benefit from adaptive trials.
