IV but not IM injection of ChAdOx1 nCov-19 triggers ChAdOx1 nCov-19-platelet binding and thrombocytopenia. (A) Exemplary flowcytometric contour plots of human washed platelets with ChAdOx1 nCov-19, BNT162b2, or PBS. Two percent contour with outliers shown gate shows adenovirus positive platelets. (B) Quantification of adenovirus platelet binding according to panel A. One-way analysis of variance with post hoc Tukeýs test. Comparison of ChAdOx1 nCov-19 to both controls. n ≥ 3 human donors per group. (C) Platelet expression of activated GPIIbIIIa mean fluorescent intensity incubated with ChAdOx1 nCov-19, BNT162b2, or PBS control. One-way analysis of variance with post hoc Tukey’s multiple comparison test, n = 4 per group. Exemplary histogram of activated GPIIbIIIa is also shown. (D) Platelet counts of mice over time. Multiple t tests with Holm-Sidak correction of ChAdOx1 nCov-19 IV and IM is shown. n ≥ 3 per time point for of ChAdOx1 nCov-19 groups; n ≥ 2 per time point for other groups. (E) Platelet counts of mice over time with 0.5 µL ChAdOx1 nCov-19 injection. Multiple t tests with Holm-Sidak correction of ChAdOx1 nCov-19 IV and IM is shown. n ≥ 6 per time point and group. (F) Linear regression of platelet count at 24 hours after inoculation and adenovirus binding to platelets 1 hour after inoculation. 95% confidence interval shown; P value denotes line deviation from zero. (G) Adenovirus-platelet binding after IM or IV administration of ChAdOx1 nCov-19 1 and 24 hours after inoculation. Unpaired t tests. n ≥ 4 per group. (H) Ratio of transfused BNT162b2 and ChAdOx1 nCov-19 platelets (total transfused platelets are normalized to 100%) over time. Unpaired t tests. n = 4 per time point. (I) Platelet surface marker expression of mouse platelets 24 hours after administration of ChAdOx1 nCov-19 IV or IM. Normalized mean fluorescent intensity. Multiple t tests with Holm-Sidak correction. n = 7 mice per group. Error bars are mean ± SEM. *P < .05; **P < .01; ***P < .001. n.s., not significant.