Covid Researchers Pool Data To Test How Well Recovered Patients' Plasma Works

While experience and anecdotal evidence suggests that plasma works in treating Covid patients, securing enough of it has been difficult.

Pooling data among traditionally competitive researchers may be one way to determine plasma's efficacy and thus help the healthcare system focus its efforts on solutions that really work. JL

Amy Marcus reports in the Wall Street Journal:

Researchers studying plasma to treat Covid-19 have struggled to find enough patients. So a new project is pooling data in real time from disparate clinical trials to form a larger data set.Pooling data from  randomized controlled trials allows investigators to glean insights from more patients than they were able themselves. The efforts reflect a sense of urgency among researchers who need answers faster than independent efforts produce.Transfusing antibody-rich plasma from recovered patients has been used for more than a century but a question has persisted: Does it really work?

Researchers studying the use of convalescent plasma to treat Covid-19 have struggled to find enough patients as the coronavirus outbreak wanes in some areas. So a new project through NYU Langone Health, Albert Einstein College of Medicine and Montefiore Medical Center is trying an unusual tactic: pooling data in real time from disparate clinical trials to form a larger data set.

The New York City trial is looking to pool data with ongoing convalescent plasma trials from at least 10 other major medical institutions. Before reaching out to other trials, It had enrolled 170 patients as of early July, according to an NYU spokesman.

Pooling data from different trials for analysis is done in medical research, but typically after the individual studies have been published. This time, investigators want to do it in real time, with data sent every two weeks to a secure server.

The efforts reflect a sense of urgency among researchers who might not have collaborated so extensively but recognize that they need answers faster than independent efforts typically produce.

A data-safety monitoring committee, independent of the investigators and a feature of many studies, will have access to the information and will be able to recommend that investigators stop a trial if patient-safety concerns arise or efficacy signals emerge early from the data.

Transfusing antibody-rich plasma from recovered patients has been used for more than a century during outbreaks of diseases such as influenza and Ebola. But a question has persisted: Does it really work?

Covid-19, the disease caused by the novel coronavirus, seemed to offer a chance to attain the gold standard of proof because of the large number of people infected and potentially able to enroll in rigorous clinical trials, according to scientists involved in studying the therapy.

But there is growing concern within the research community that many of the trials now under way are too small to prove conclusively whether the transfusions work. By the time regular clinical trials are organized, funded and approved, the outbreak often has waned in a particular place.

Some studies also lack a control group of patients. Control groups enable scientists to make definitive comparisons of outcomes between those who receive an experimental treatment and others who receive the standard of care or a placebo.

More than 40,000 Covid-19 patients have received convalescent plasma through a national expanded access program, also known as compassionate-use, sponsored by the Food and Drug Administration and led by the Mayo Clinic in Rochester, Minn. That study has amassed a lot of safety data, but because every patient gets the plasma it isn’t possible to meet the gold standard of proof of a randomized controlled trial.

“We are concerned that in the rush to get trials up and running, there were some corners cut in the design, and the studies will not produce data that will allow a robust assessment of the value of the treatment,” says Michael Murphy, professor of transfusion medicine at the University of Oxford and one of the authors of a recent survey of 64 convalescent plasma trials around the world published in Transfusion Medicine Reviews.

The data-pooling initiative, which began outside the sponsorship of a government agency or a corporation, seeks to address some of these challenges. Pooling the data from the various randomized controlled trials allows investigators to glean insights from more patients than they were able to enroll themselves.

Other organizations are taking steps on their own to find more patients for trials. The AABB, the transfusion community’s national group, created a grid looking at which of the convalescent-plasma trials registered at clinicaltrials.gov share similar outcome measures, such as mortality at seven days and 28 days, or need for supplemental oxygen. It has begun contacting principal investigators to suggest they pool data, said Claudia Cohn, director of the blood bank lab at the University of Minnesota and chief medical officer of AABB.

After the surge of hospitalizations ebbed in New York City, investigators at Columbia University started sending convalescent plasma to collaborators at a research institution in Brazil, where the pandemic is raging, to enroll more patients in a trial. “They need answers as do we,” says W. Ian Lipkin of Columbia University, the principal investigator of the study.

Investigators at Weill Cornell Medicine in New York considered going it alone but realized it would be difficult “for a single center to get enough numbers to really answer the questions we have,” said Melissa Cushing, director of the clinical laboratories at Weill Cornell Medicine and New York-Presbyterian. Investigators there joined forces with McMaster University in Canada, which is working with sites in Canada and the U.S. to enroll 1,200 patients in a randomized controlled trial of convalescent plasma. The trial has 58 people enrolled so far, investigators say.

There are hurdles to successfully pooling data among existing clinical trials, even when they are open to collaboration. Priscilla Hsue, a co-principal investigator on a convalescent plasma trial at the University of California, San Francisco, said her team plans to send data every two weeks to pool with others in the NYU-Einstein-Montefiore consortium.

Dr. Hsue said the UCSF investigators don’t plan to change their study design, and are still examining which data they are collecting in common. “Everyone’s trial is slightly different,” she said.

Katharine Bar of the University of Pennsylvania said enrollment stalled in that institution’s randomized controlled convalescent-plasma trial, as the number of hospitalized patients declined. They have joined the pooled-data project. She said her group wants to enroll 80 patients and is currently at 26.

“It takes time and energy and personnel to keep a study going,” Dr. Bar said. “It is harder to do at low capacity.” She said her group still believes they can reach their enrollment goal. “By pooling data, we hope we can get more quickly to an answer to the question of efficacy,” she said.