COVID-19 clinical trial slated at UPMC
PITTSBURGH — Searching for an effective treatment for COVID-19 during the ongoing pandemic, University of Pittsburgh and UPMC researchers have announced a “learning-while-doing” clinical trial that assesses drugs' efficacy while treating patients.
The trial will use artificial intelligence to simultaneously test different combinations of drugs, including the anti-malarial drug hydroxychloroquine, corticosteroids and immunomodulators, which alter the responsiveness of the immune system.
Patients admitted to any of UPMC's 40 hospitals with COVID-19 will have the option to participate in the trial, along with patients in at least 13 countries.
Dr. Derek Angus, professor and chairman of the university's School of Medicine's critical care department, said this type of trial allows researchers to determine what drug or combination of drugs are most successful and lets clinicians most effectively treat patients.
“This approach, of using adaptive platforms, is not what people ever did normally, but it is totally what many groups now want to do,” Angus said Thursday. “And by many groups I mean, for example, the FDA.”
How it works
Patients will be randomly assigned treatments when they enroll in the program. As the trial progresses, the more-effective drugs will be assigned more often, Angus said, and less-effective treatments may be removed entirely.
“The trial is essentially taking bets, even before it knows definitively, so that therapies that are on their way to winning, ... that therapy that is most likely to be doing well, is the therapy that is steered most often to patients,” he said.
Data sampled worldwide
Data from throughout the world are updated through the UPMC electronic health record program, which allows the AI to move in real time.
“The way our engine, the statistical engine of this trial, works is that it then takes the data from the entire world and then updates the probabilities of assignment, so that when a patient turns up at Passavant and is being randomized, the recipe they are assigned is based on information that we've been receiving from patients that have been enrolled in London, in Amsterdam and in Sydney,” Angus added.
Around 12.5 percent of patients enrolled in the trial will be the control group, not receiving any treatment, when they enroll. But after a few weeks of being in the trial, the researchers expect 99 percent of patients will have received drug treatment.
While the adaptive trial is novel in acute care, Dr. Donald Yealy, chairman of the Department of Emergency Medicine at Pitt's medical school, said this type of trial has been used in other areas of medicine, but not usually in the acute space area.
“So we're not the first to actually consider the design, we're the first to consider it in this particular deployment,” Yealy said. “And we think this is where it has its highest utility, to get to the most appropriate answers as quickly and as efficiently as possible.”
The program has some logistical issues to work out. Researchers are working to bring in video capability so patients and families will have the ability to consent to enrollment in the program with the most information possible, for example.
The idea to apply this type of adaptive trial in the time of a pandemic came about after the 2009 H1N1 influenza — or “swine flu” — pandemic, Angus said. Researchers worked to find therapies, but by time they were successful the flu had left their region.
After that, Angus said, researchers discussed how to find treatment more quickly for fast-moving diseases.
For now, researchers are focusing on how to improve their treatment and understanding of COVID-19 day-over-day.
“We want to do the very best possible treatment for the patient in front of us, but we want to do even better treatment for the patient who comes tomorrow,” Angus said.
