Mechanisms of Platelet Activation by Antithymocyte Globulin (ATG).

We have previously shown that polyclonal rabbit ATG from Fresenius (ATG-F) can induce non-overt DIC in patients undergoing hematopoietic stem cell transplantation (HSCT) [Weber et al. 2003]. The compensated state of the coagulopathy suggested that the severe drop in platelet counts (>60%) observed in patients with normal pretreatment levels was not solely due to thrombin generation and platelet consumption. We thus further investigated the mechanisms of ATG-F-induced thrombocytopenia, hypothesizing that ATG-F had direct platelet-stimulating activity. Using an indirect flow cytometric binding assay we found that, compared to control IgG, ATG-F dose-dependently (1–100 μg/ml) bound to the surface of resting and TRAP-activated platelets, showing an up to 50fold increase in MFI at 50–100 μg/ml. In a washed platelet system, in which baseline positivity for CD62P was generally <15%, ATG-F alone had no effect on platelet aggregation. However, when platelets were primed with low concentrations of ADP (1–2 μM) or epinephrine (0.5 μM), ATG-F induced strong and stable aggregation of up to 80–100% in a dose-dependent manner (25–100 μg/ml), whereas control IgG did not. The priming effect of ADP was dependent on both P2Y1 and P2Y12 as evidenced by respective ADP receptor antagonists. Preincubation of platelets with inhibitory CD32 mAb completely abolished ATG-F-induced platelet aggregation, suggesting that clustering of, and signalling through, FcγRIIA was crucial for this platelet response. Similarly, ATG-F-derived F(ab′)2 had no effect on ADP-triggered platelet aggregation. Furthermore, preliminary studies indicated that responsiveness of platelet donors to ATG-F was associated with the FcγRIIA-R/R131 polymorphism, which has previously been shown to confer increased in vitro platelet responsiveness to heparin/PF4 antibodies from patients with HIT. In a 14C-serotonin release assay, in which prolonged incubation at 370C under shaking conditions increases baseline platelet activation and ensures constant platelet-platelet contacts, ATG-F (100 μg/ml), but not control IgG, induced strong FcγRIIA-dependent dense-granule release of up to 80%. In contrast, using diluted platelet-rich plasma under static conditions, ATG-F had only minor effects on platelet activation, increasing CD62P+ platelets from 2±1 to 15±5% (n=5). In a cohort of 16 consecutive HSCT patients receiving ATG-F, two patients were identified with normal pretreatment platelet counts and soluble CD62P (sCD62P) plasma levels of >51 ng/ml (mean+2SD of 23 controls). In these patients, an ATG-F-induced drop in platelet count to 38 and 19% of baseline was associated with a 1.7 and 2.2fold increase in sCD62P, respectively, indicating further platelet activation. In summary, strong binding of ATG-F to resting platelets may accelerate their clearance by the reticulo-endothelial system, thus contributing to the pathophysiology of thrombocytopenia. While ATG-F alone had negligible effects on resting platelets, it significantly enhanced activation of prestimulated platelets. Patients with normal-to-high platelet counts and evidence of in vivo platelet activation may especially be prone to this potentially hazardous side effect.