Opinion: For early testing of convalescent plasma, we were ‘building the plane while we were flying it’

During the Covid-19 pandemic, clinicians have seen many remarkable things. Among them is how a 19th-century therapy called convalescent plasma came to be used in more than 500,000 hospitalized Covid-19 patients the United States and elsewhere.

Because we’ve had front-row seats on this journey, we’d like to offer answers to two key questions: How did this happen? Is it safe and does it work?

How did this happen?

Convalescent plasma is antibody-rich plasma (the liquid portion of blood) from patients who have recovered from an infection. It is a rough-and-ready form of antibody therapy. For more than a century, convalescent plasma and related products have been tried with some success as treatments during numerous unexpected outbreaks of infectious diseases.


Antibody therapy was also a mainstay of infectious disease treatment before World War II, and still has some niche uses especially in post-exposure prophylaxis for several viral diseases.

At the beginning of the Covid-19 pandemic, convalescent plasma therapy was used almost immediately in China. It was also percolating in the U.S., highlighted by a late February 2020 op-ed in the Wall Street Journal by one of us (A.C.) recounting how, in the 1930s, a measles outbreak in a boy’s school was stemmed by the use of convalescent blood products.


Because we are friends, this led to an email serve and volley between us. By mid-March, protocols for using convalescent plasma for prophylactic and early treatment of Covid-19 were in development at the health centers where we work: Mayo Clinic (M.J.J.) and Johns Hopkins (A.C.). Early in the email exchange, one of us (M.J.J.) noted that the blood banking infrastructure in the U.S. was big enough to potentially use convalescent plasma at scale.

Little did we know or anticipate what would happen in the months that followed.

Colleagues from all over the country connected with us and, throughout March, we were having nightly calls about getting clinical trials off the ground. A group of us created the Covid-19 Convalescent Plasma Project, which provided a nexus for physicians interested in this therapy and served to disseminate critical information on protocols, efficacy, and the history of this therapy. We were also participating in calls with various elements of the Food and Drug Administration and National Institutes of Health, along with other interested parties.

Within a few of weeks of the virus taking off in the U.S. in mid-March, our plan was to conduct small or modest-sized trials with two use cases in which antibody therapy had previously shown to be effective: preventing Covid-19 in health care workers at risk of occupational exposure and early treatment of those with Covid-19.

Our group wasn’t the only one pursuing this therapy. The initial media reports of compassionate use of convalescent plasma occurred in late March of 2020, when the first patient was treated at Houston Methodist Hospital.

Not long after that, the FDA was getting hundreds of requests for compassionate use of convalescent plasma from physicians all over the country. Patients and their families were using social media to find recovered Covid-19 patients in desperate efforts to secure their plasma.

The workload associated with the FDA’s reviewing and fulfilling compassionate use requests was immense, no systematic data were being collected relevant to safety and possible efficacy, and there was simply no sustainable way to manage the process.

As a result of this chaos, at the end of March 2020, Peter Marks, director of the FDA’s Center for Biologics Evaluation and Research, contacted one of us (M.J.J.) with a “can you take a look at an expanded access protocol for convalescent plasma” email. Expanded access allows physicians to use unapproved therapies to treat patients with severe or life-threatening disease when there are few alternatives and there is reason to believe the therapy might work.

The protocol was straightforward and, as we learned later, convalescent plasma was an ideal candidate for expanded access in the middle of a pandemic because it was likely to be safe and had a chance to work. But two things were needed: a coherent way to provide access to it across the country outside of trials, and creating a mechanism for collecting information on its use and outcomes.

After receiving a “looks good” response about the expanded access protocol, Marks indicated that an institutional review board (IRB) was needed to assess the protocol, someone needed to hold the investigational new drug (IND) application, and an organization was needed to collect and analyze the data. One of us (M.J.J) agreed to be the IND holder and Scott Wright, who leads the Mayo Clinic institutional review board, quickly agreed to review the protocol and, almost as quickly, indicated that if it was approved the Mayo Clinic could serve as the central IRB.

Scott is an experienced trialist and had many practical ideas about how to establish a streamlined and online process. He also helped line up the internal regulatory support and administrative personnel needed to get things rolling. Online survey research tools used by our colleagues DeLisa Fairweather and Katelyn Bruno could be adapted to run the program. One of us (M.J.J.) then contacted a long-term statistical collaborator, Rickey Carter, who figured that we could “send him the data” and his team would find a way to get them analyzed. The fellows, residents, and staff in M.J.J.’s lab got busy setting up an engine room to run things.

By early April, the expanded access protocol had been approved with a cap of 5,000 patients.

The use of convalescent plasma explodes

The next part of “how it happened” is about the factors that converged to ultimately lead to 105,000 patients enrolled in the expanded access protocol and nearly 95,000 patients treated under it during the spring and summer of 2020. (An estimated 400,000 other patients have received convalescent plasma since the end of August when the protocol ended and the FDA granted this therapy emergency use authorization.)

Delivering convalescent plasma to patients requires donors and collectors. The donor part was solved in part by community partners like Chaim Lebovits from the Orthodox Jewish community in and around New York and Diana Berrent from Survivor Corps. These advocates found and recruited donors and worked with the blood banking community to collect convalescent plasma at scale. Equally important, when our non-medical partners had ideas from their real-world experience about how to get things done, we listened, and the blood bankers listened, and then we acted.

Word of mouth, as well as media and social media coverage of hopeful stories about convalescent plasma also promoted its use. We believe that the user-friendly design of enrollment into the expanded access protocol and the reporting process made it easier for patients to get treatment. Other organizations also started to promote plasma collection and use.

All of this was underpinned by financial support from the federal Biomedical Advanced Research and Development Authority for implementing the expanded access protocol and for the blood bankers. Philanthropy also provided support.

As favorable outcomes of convalescent plasma therapy began to appear in the press and medical journals, they gave hope to patients and physicians at a time when there was little more available for treating Covid-19 than supportive therapy. To be sure, these early reports were case series and noncontrolled trials, which provided suggestive information at best.

Thus, the explosive growth in the use of convalescent plasma in the U.S. was driven by a plentiful supply, the hopes of patients and physicians who embraced it, and our streamlined approach to enrollment and reporting. As our colleague Rickey Carter pointed out to us, we went from having an idea to enrolling more than 100,000 patients and the closeout of the study in 150 days — less time than a full-term pregnancy.

No competition for randomized trials

Some have argued that the expanded access program for convalescent plasma slowed and even hindered doing randomized clinical trials (RCTs) to test the efficacy this therapy. That’s just not true.

Many randomized clinical trials of convalescent plasma were set up in the early days of the pandemic in surge areas that subsequently controlled the epidemic, and so had no more patients to enroll. Most of these early RCTs were stopped before enough patients could be enrolled. Also, in the early days of the pandemic, the necessary laboratory tests to measure the quantity and quality of antibodies in plasma were not widely available.

The best evidence that the expanded access protocol did not hinder RCTs comes from the fact that after the FDA issued its emergency use authorization for convalescent plasma on Aug. 23, 2020, several U.S.-based trials continued to enroll patients and the outcomes of those trials are eagerly awaited today. In fact, the expanded access protocol produced important information for both efficacy and safety that is essential for the design and interpretation of the results from randomized controlled trials.

There are important caveats to remember, as the protocol grew from a demonstration project to something much larger. The patients treated under the expanded access protocol were geographically and ethnically diverse: Most were treated at facilities that would never have been part of a clinical trial. In addition, many of the researchers who were early proponents of the protocol were in parallel designing more formal randomized trials or enrolling patients in them. In this way, the expanded access protocol created access to the therapy for many U.S. patients who would not otherwise have had access to clinical trials.

Is it safe and does it work?

One unequivocal contribution of the expanded access protocol to the fight against Covid-19 was to establish that the administration of convalescent was as safe as administering regular plasma.

In March 2020, there were fears that giving Covid-19 patients the antibodies found in convalescent plasma would make them worse through a mechanism known as antibody-dependent enhancement (ADE), which can occur with coronaviruses like SARS-CoV-2, the virus that causes Covid-19. In fact, the possibility of ADE had hindered earlier efforts to develop vaccines for other coronaviruses. When the expanded access protocol established that convalescent plasma was relatively safe and that there was no hint of ADE, this cleared the path forward for the rapid development of vaccines and monoclonal antibodies therapies.

Had the protocol turned up any toxicity associated with antibodies to the Covid-19 virus, that would have slowed the development of vaccines and monoclonal antibody therapies by necessitating additional safety testing for them. The fact that convalescent plasma was safe provided critical reassurance as the world began racing to develop vaccines for Covid-19 in the spring of 2020.

Before diving into the efficacy question, it’s important to point out that expectations for an answer to it grew as enrollment escalated beyond anything we initially anticipated. The idea that there must be some way to find a signal of efficacy someplace in the data seemed to accelerate as patient enrollment took off.

In this context, we found that mortality was lower in patients treated early in the course of Covid-19 with plasma containing higher levels of antibodies. This finding was subsequently confirmed by studies in Israel and Argentina and is essentially a rediscovery of what was known in the 1920s and 1930s about antibody therapy.

It is also part of a larger set of convergent information about the effectiveness of this therapy based on the antiviral effects of plasma, results from animal models, observations in rare patients who do not generate endogenous antibodies, and insights from retrospective and prospective trials. These data all indicate that antibody therapy can be effective if used early in the course of disease and patients receive sufficient doses of antibodies. This conclusion is also supported by data from monoclonal antibody studies.

A third question?

The obvious third question is about what we have learned from this saga and what we might have done differently. Even as the pandemic continues to rage, the sense of urgency, the desire to do something, and political confusion have abated to some extent. Yet it is perhaps too soon to apply the retrospectoscope to the expanded access protocol for convalescent plasma.

What we can say for certain is that better ways are needed to rapidly design and implement randomized controlled trials in an emergency. For convalescent plasma, we were building the plane while we were flying it. The hardest part was that there was no playbook, or even an outline of a playbook.

Equally important, the need to provide broad access to promising therapies to geographically and ethnically diverse groups of patients receiving care outside of academic centers must remain a priority as we prepare for the next pandemic.

Michael J. Joyner is a physician and professor in the Department of Anesthesiology and Perioperative Medicine at the Mayo Clinic and led the U.S. Expanded Access Program for Convalescent Plasma. Arturo Casadevall is chair of the Department of Molecular Microbiology and Immunology at the Johns Hopkins Bloomberg School of Public Health. The views expressed here are those of the authors alone.

Source: STAT