A Randomized Study of Immune Plasma for the Treatment of Severe Influenza

Beigel, John H.; Tebas, Pablo; Elie-Turenne, Marie-Carmelle; Bajwa, Ednan; Bell, Todd E.; Cairns, Charles B.; Shoham, Shmuel; Deville, Jaime G.; Feucht, Eric; Feinberg, Judith; Luke, Thomas; Raviprakash, Kanakatte; Danko, Janine; O’Neil, Dorothy; Metcalf, Julia A.; King, Karen; Burgess, Timothy H.; Aga, Evgenia; Lane, H. Clifford; Hughes, Michael D.; Davey, Richard T.

Twenty-eight (67%) participants randomized to receive plasma had documented resolution of tachypnea and hypoxia by Day 28, compared to 24 (53%) of control participants (p = 0·069) ( Figure 2A ). The hazard ratio of plasma + standard of care vs. standard of care alone was 1.71 [95% CI: 0·96, 3·06]. From the Kaplan-Meier analysis, with the caveat that the study evaluated tachypnea and hypoxia at Days 1, 2, 4, 7, 14 and 28, the estimated median time to resolution of tachypnea and hypoxia was 7 days (lower quartile 2 days) among participants randomized to receive plasma versus 28 days (lower quartile 7 days) among participants randomized to the control arm. Due to the higher than expected loss to followup, sensitivity analysis was performed with follow-up censored at last available assessment with sufficient respiratory status data (p = 0·086). Six participants randomized to receive plasma resolved the tachypnea and hypoxia present at screening by the time of baseline (prior to receiving plasma), as compared to one among participants that received standard care. With these participants excluded, 22 (61%) participants in the PEP randomized to receive plasma had resolution of tachypnea and hypoxia by Day 28, compared to 23 (52%) of controls (p = 0·26). The benefit was primarily seen in participants who were enrolled within 4 days of symptoms (treatment by subgroup interaction p-value = 0·038). ( Figure 2B ). As the participants randomized to standard care had slightly more severe disease at baseline, in order to evaluate the influence of these baseline differences, analysis was performed stratified by use of oxygen, mechanical ventilation, ICU requirement, and presence of ARDS. All demonstrated similar results as in the primary analysis (stratified logrank p=0·019 to 0·12).