Early but not late convalescent plasma is associated with better survival in moderate-to-severe COVID-19

Abstract

Background: Limited therapeutic options exist for coronavirus disease 2019 (COVID-19). COVID-19 convalescent plasma (CCP) is a potential therapeutic, but there is limited data for patients with moderate-to-severe disease.

Research question: What are outcomes associated with administration of CCP in patients with moderate-to-severe COVID-19 infection?

Study design and methods: We conducted a propensity score-matched analysis of patients with moderate-to-severe COVID-19. The primary endpoints were in-hospital mortality. Secondary endpoints were number of days alive and ventilator-free at 30 days; length of hospital stay; and change in WHO scores from CCP administration (or index date) to discharge. Of 151 patients who received CCP, 132 had complete follow-up data. Patients were transfused after a median of 6 hospital days; thus, we investigated the effect of convalescent plasma before and after this timepoint with 77 early (within 6 days) and 55 late (after 6 days) recipients. Among 3,217 inpatients who did not receive CCP, 2,551 were available for matching.

Results: Early CCP recipients, of whom 31 (40%) were on mechanical ventilation, had lower 14-day (15% vs 23%) and 30-day (38% vs 49%) mortality compared to a matched unexposed cohort, with nearly 50% lower likelihood of in-hospital mortality (HR 0.52, [95% CI 0.28-0.96]; P = 0.036). Early plasma recipients had more days alive and ventilator-free at 30 days (+3.3 days, [95% CI 0.2 to 6.3 days]; P = 0.04) and improved WHO scores at 7 days (-0.8, [95% CI: -1.2 to -0.4]; P = 0.0003) and hospital discharge (-0.9, [95% CI: -1.5 to -0.3]; P = 0.004) compared to the matched unexposed cohort. No clinical differences were observed in late plasma recipients.

Interpretation: Early administration of CCP improves outcomes in patients with moderate-to-severe COVID-19, while improvement was not observed with late CCP administration. The importance of timing of administration should be addressed in specifically designed trials.

Fig 1. Consort selection tree.

Patient disposition in the observational cohort analysis.

Fig 2

Kaplan-Meier curves of survival in early (A) and late (B) convalescent plasma recipients vs respective matched unexposed patients. Shown are Kaplan-Meier estimates of survival from time of index in (A) early CCP recipients (solid line) vs matched unexposed patients (dashed line) (B) late CCP recipients (solid line) vs matched unexposed patients (dashed line). Survival improved for early CCP at 14-days (15% vs 23%) and 30-days (38% vs 49%) compared to matched unexposed patients (HR 0.52, [95% CI 0.28–0.96]; p = 0.0367). There was no difference in mortality at 14-days (28% vs 29%) or 30-days (42% vs 47%) among late CCP recipients compared to their matched unexposed patients (HR 0.98,[95% CI 0.53to 1.83]; p = 0.95). Censoring is indicated by the tick mark “+” with number by each ten-day interval marked below the number at risk.

Fig 3

Histogram plots of WHO ordinal scale scores before and after intervention in the early (A) and late CCP (B) compared to matched unexposed cohorts. Shown is a histogram of the change is WHO COVID-19 severity score from time of index to discharge in (A) early CCP recipients (gray solid bars) vs matched unexposed patients (black solid bars) (B) late CCP recipients (gray dashed bars) vs matched unexposed patients (black solid bars). WHO scores in (A) early CCP recipients were significantly improved at discharge (difference in mean -0.9, [95% CI -1.5 to -0.3]; p = 0.004) when compared to unexposed patients. Relative to the unexposed patients, early CCP recipients were nearly twice as likely to demonstrate an improvement in their WHO scores at discharge from baseline index day scores (OR 1.9,[95% CI 1.1 to 3.3]; p = 0.02). Among (B) late CCP recipients, there was no effect of CCP on WHO scores at discharge (difference in mean +0.6, [95% CI -0.4 to 1.5]; p = 0.23). The odds ratio for improvement at discharge was 0.6 (95% CI 0.3 to 1.3; p = 0.22).