A randomised, controlled phase 2 trial involving 464 patients failed to show an effect on progression from moderate to severe disease or mortality.
In the absence of other effective therapies, convalescent plasma — first tested against diphtheria and tetanus in the late 19th century — has been tested for COVID-19. Observational reports and small studies have had variable results so far. Now, researchers have published the first randomised open-label trial of convalescent plasma for treatment of moderate COVID-19 (PaO2/FiO2 between 200 and 300 mm Hg or respiratory rate ) 24/min and oxygen saturation ≤93% on room air) involving 464 adults in 39 tertiary care hospitals across India. All patients received antiviral agents, antibiotics, and immunomodulators or corticosteroids. In addition, 235 were assigned to two doses of 200-mL convalescent plasma, one at randomisation and one 24 hours thereafter. The plasma was donated by 262 individuals (94% with mild disease) a median of 41 days after reverse transcription–polymerase chain reaction confirmation of COVID-19. Of the donors, 64% had neutralising antibody titers ) 1:20 (median titer, 1:40).
Since most participants had neutralising antibodies at enrollment, plasma administration may have come too late in the course of COVID-19, but even in the small subgroup given plasma within 3 days of symptom onset, no effect was seen. Besides, the antibody concentration in the convalescent plasma used in this trial and volume given may have been too low. Still, these results discourage the use of unscreened convalescent plasma for COVID-19, despite its approval in several countries.