Figure 4. PF4/P IgG antibody levels in plasma from patients with CM are associated with decreased circulating platelets and platelet activation.

(A) Spearman correlation analysis of circulating platelet levels in patients with Ret⁺ CM versus PF4/P IgG levels (n = 98) in plasma. Values were log transformed for linear regression analysis. (B) Heterologous platelet activation assay showing relative activation levels (percentage of CD62p/CD41) of donor platelets when incubated with plasma from patients with UM (n = 13) or CM (n = 26) or from HIT⁺ patient plasma (n = 5) in the presence of recombinant hPF4 (15 μg/mL, PF4) or recombinant hPF4 plus HDH (200 U/mL). Shown for each data point is the mean relative activation from 3 independent experiments. Treatment with ADP (10 μM) served as an internal positive control for maximal platelet activation. Statistical significance between hPF4-treated clinical groups was determined by Kruskal-Wallis test with Dunn’s multiple comparisons. Analysis within a clinical group for hPF4 treatment versus hPF4 plus HDH treatment was determined by Welch’s t test. Bar graph represents the mean ± SD. (C) Spearman correlation analysis of the relative platelet activation (x axis) in the subset of samples from B, plotted against the corresponding PF4/P IgG levels (n = 39) in patient plasma. (D and F) Spearman correlation analysis of the relative platelet activation (x axis) against (D) circulating platelet counts (n = 32) and (F) against anti–PS IgG levels (n = 36) in patient plasma. (E) Spearman correlation of PF4/P IgG plasma levels (x axis) plotted against circulating platelet counts (n = 32) for the subset of samples from B. Spearman’s R_s coefficient, the associated P value, and calculated linear regression (dashed line) are shown within the graphs (A and C–F).