Convalescent plasma transfusion for COVID-19: Impact of neutralizing antibodies on respiratory outcomes

ABSTRACT

Background/Case Studies In the past months, convalescent plasma (CP) has been used as an alternative therapy to treat COVID-19 patients. Previous studies highlighted the role of neutralizing antibody titers (NAbs) on clinical improvements. We analysed a series of cases of COVID patients treated with CP transfusion and associations between transfused NAbs and patient NAbs on respiratory outcomes. Study Design/Methods: Thirty patients diagnosed with SARS-CoV-2 by RT PCR and severe pneumonia were prospectively analysed at a single center in Brazil. Doses of 300-600ml of CP were transfused. To assess respiratory outcomes, PaO2/FiO2 ratios were determined at Day 0 (day of CP transfusion), Day 5 and by discharge and duration of ventilation support were analysed. Improvements were determined by variations of PaO2/ FiO2 ratios from Day 0 through day 5 (V0-5) and from Day 0 through discharge (V0-D). Neutralizing antibodies from patients prior to transfusion (NAbsP) and neutralizing antibodies of CP units transfused (NAbsCP) were analysed. NAbsCP considers the total amount of antibodies transfused to account for volume differences. We performed a generalized estimating equations (GEE) approach with logit link for binary data to model the effect of Nabs CP and variations on PaO2/FiO2 ratios between Day 0-Day 5 and Day 0-discharge. Regression models were performed to determine the predictive variables associated to improvements on PaO2/FiO2 ratios and duration of ventilation support. Results/Findings: Variations of PaO2/FiO2 ratios from day 0 through Day 5 and day 0 through discharge are displayed on Table 1. Significant improvements on PaO2/ FiO2 ratios were observed from Day 0 through discharge (p<0.001). NAbsP were associated to higher improvements on PaO2/FiO2 on V0-D (mean difference (MD)) 2.8 CI 95% -1.0-4.6 p 0.003), NAbsCP, however, were associated to minor variations in PaO2/FiO2 (MD -4.8 CI95% -7.9;-1.7 p 0.002) on the same interval. At each 100 unit increase in NAbs CP, variations on PaO2/FiO2 on V0-D were expected to be 4.8 units lower. When analysing V0-5, NAbsCP were again associated to minor improvements of PaO2/FiO2 ratios from D0 through D5 (MD -3.6 95%CI -7.2;-0.003 p 0.05). Other variables did not show statistical significance. When considering duration of ventilation support neither NabsP (Mean Ratio MR 0.985 CI 95% 0.962-1.007 p 0.185) nor NAbsCP (MR 0.967 CI95% 0.918 1.018 p 0.198) showed significant statistical differences. Conclusions: In our analysis, NAbs produced from the patient prior to CP transfusion are associated to better improvements on respiratory outcomes when compared to NAbs from transfused units. Regarding duration of mechanical ventilation, neither NAbsP nor Nabs CP had impact on outcomes.