A randomized phase 2 clinical trial has found convalescent plasma has no effect on progression to severe disease or all-cause mortality in COVID-19 patients. The data raise significant doubts about the efficacy of an intervention that has been given to more than 100,000 people in the U.S.
Plasma taken from patients who survived COVID-19 was an obvious early therapeutic response to the coronavirus outbreak. There is a long history of using convalescent plasma to treat infectious diseases, and, with many thousands of patients surviving infection with SARS-CoV-2, a large pool of potential donors was available. In the absence of other treatment options, 100,000 U.S. patients received plasma between April and August under the FDA’s expanded access protocol (EAP).
Now, a team from the Indian Council of Medical Research has shed light on whether the therapy is effective in a paper published in The BMJ. The researchers enrolled 464 adults hospitalized with moderate COVID-19. Half received convalescent plasma on top of standard of care. The other half just received standard of care.
In the convalescent plasma arm, 19% of patients progressed to severe disease or died in the 28 days after enrollment, compared to 18% of people in the control cohort. The lack of difference between the two arms caused the trial to miss its primary endpoint. Limiting the analysis to patients who got plasma with detectable antibody titers had no effect on the outcome.
Secondary endpoints linked convalescent plasma to small improvements in the resolution of fatigue and shortness of breath. However, as Beth Pathak of the Women’s Institute for Independent Social Enquiry said in an accompanying editorial, the unblinded design of the trial means “knowledge of treatment status could have influenced the reporting of subjective symptoms.” Plasma was no better than control against objective measures such as days on respiratory support and in hospital.
Yet the trial generated some more robust evidence that convalescent plasma acts on SARS-CoV-2. Notably, 68% of patients in the plasma arm tested negative for SARS-CoV-2 RNA on Day 7, versus 55% in the control group. The difference, while slight, suggests the neutralizing antibodies found in convalescent plasma may help patients eliminate the virus.
The failure of the apparent effect of plasma on viral clearance to translate into reductions in disease progression and death raises questions. Pathak’s analysis of those questions covers the thrombotic aspects of COVID-19 and the fact that therapeutic plasma has prothrombotic properties, at least when given to patients with acute bleeding and complex coagulopathies.
“It is notable that progression to severe disease or death occurred in 20% (13/64) of patients who received convalescent plasma with no detected neutralizing antibodies compared with 18% (41/229) of controls,” Pathak wrote.
Pathak called for further investigation of the prothrombotic risks of convalescent plasma. The Indian trial did not include thrombotic events as a pre-specified outcome, and analyses of the U.S. EAP data excluded most cardiac and thrombotic events from the reported adverse rates on the grounds they were deemed unrelated to plasma.
It is also possible that plasma would perform better in a different population of COVID-19 patients. The Indian trial enrolled hospitalized patients. Based on the typical progression of COVID-19, the subjects were likely infected days before entering the trial. Treating patients sooner could achieve better results, although early intervention may only be justified in high-risk individuals.