As COVID-19 rips into communities worldwide, in our own community Pitt and UPMC are accelerating the normally slow, methodical process of clinical trials to unforeseen speeds, using a “learning while doing” approach.
A new clinical trial developed by the University of Pittsburgh School of Medicine launches today at UPMC. It will fast-track an investigation of many potential therapies at the same time — like the much-debated (and frequently touted by President Trump) anti-malarial drug hydroxychloroquine.
“The solution is to find an optimal trade-off between doing something now, such as prescribing a drug off-label, or waiting until traditional clinical trials are complete,” said Derek Angus, professor and chair of Pitt’s Department of Critical Care Medicine. “We’ve developed a way to do that with an adaptive clinical trial model that relies on a type of artificial intelligence known as reinforcement learning to identify the best, evidence-backed therapy for COVID-19 much faster than using the traditional scientific approach.”
Before COVID-19 was even a threat, Angus was working with colleagues around the world on a project called REMAP-Community Acquired Pneumonia (REMAP-CAP), which was engineered to find the best treatments for severe pneumonia in pandemic and non-pandemic situations.
Now, REMAP-CAP has been swiftly adapted to target the COVID-19 virus. The platform gives researchers the ability to test multiple treatments simultaneously, at a lower cost, and with fewer patients than the usual clinical trials.
It’s called “adaptive platform trials.” For a vivid comparison, Angus likens this approach to a chef offering a prix fixe menu with appetizer, main course and dessert. The chef may try various combinations, serving sizes and options, sometimes leaving out the appetizer or dessert, and adjusting on the fly as plates come back scraped clean or barely touched, until hitting on the combination that sells best.
“We will simultaneously be testing combinations of the anti-malarial drug hydroxychloroquine, along with different combinations of corticosteroids, and then rolling in other, more targeted immune system modulatory drugs,” said Angus.
“We must throw out old ways of thinking and fuse clinical care and clinical research into one extremely efficient system. This is an unprecedented pandemic and we need an unprecedented response.”
The UPMC-REMAP-COVID19 trial will be integrated with UPMC’s electronic health record system, across its entire 40-hospital system. Every patient admitted with COVID-19, can be part of the program if they choose. New drugs that emerge are rolled into the platform, rather than waiting for separate clinical trials.
All COVID-19 patients will receive the same standard of care and may receive up to three experimental treatment options.
REMAP-CAP is a worldwide platform that uses machine learning to keep learning which therapies are working best across the world.
“Patients are being enrolled already, and they’re being enrolled in 13 countries,” said Angus. “The way 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, when a patient turns up at Passavant, and is being randomized, the recipe they’re assigned is based on information that we’ve been receiving from patients that have been enrolled in London, in Amsterdam, and in Sydney.”
It has gotten the attention of health systems all over the world.
“Just to give you a flavor of how this trial is being perceived, the chief medical officer of the British National Health Service last week sent a letter encouraging every hospital in Britain to participate in this trial,” says Angus.
If one treatment shows promise, patients are enrolled automatically into that treatment option.
“UPMC-REMAP-COVID then learns on the fly using a built-in artificial intelligence engine, such that within this combination of therapies, those treatments that are doing the best–even while the trial is still running–the trial then changes the odds towards getting those treatments more frequently,” said Angus.
“So the poor performing therapies are discarded quickly, and, essentially, our physicians are always betting on the winning horse. With this approach the trial ends faster, fewer patients are are exposed, and we have a finding that is statistically significant and the best treatments are given to the most patients.”
A very credible University of Washington model predicts that the number of COVID-19 cases and ICU beds needed in Pennsylvania will peak this weekend, notes Donald Yealy, chair of the Department of Emergency Medicine at UPMC and the University of Pittsburgh.
“Today’s April 9, and we have not seen yet that dramatic overwhelming surge of COVID-19 patients that we prepared for, and that we feared,” said Yealy. “We are seeing evidence that the pandemic is growing. But we see the possibility for a manageable future with no very high surge here in western Pennsylvania.
“I’m pleased to report that based on all of the activities–the social distancing and your participation–and other public health measures, it looks like we ‘flattened the curve.'”
A rise in cases is likely on the way, however.
“Our hospitals have plenty of beds and plenty of capacity, should we need to take care of more patients, including COVID-19-infected patients,” said Yealy. “We have an ample supply of masks, gloves, gowns and respirators to do this safely. We’ve launched innovative efforts to conserve all of our equipment so that we can maintain the supply into the future.”
“These efforts include sterilization processes to make sure that we can safely continue to use the N95 masks–that you’ve heard about–to help protect our employees. We also have partnerships with local universities to use 3-D printing to make thousands of nasal swabs…”
Last week, Pitt and UPMC announced that they had developed a contender in the race for a potential coronavirus vaccine — a dissolvable microneedle array that is applied to the skin like a Band-aid.