Myeloablation and allogeneic hematopoietic stem cell transplantation causes a marked GPIIb/IIIa activation defect and procoagulant platelet phenotype

Introduction In response to allogeneic hematopoietic stem cell transplantation (HSCT), all blood cell lineages become replaced by donor-derived cells. During this process delayed engraftment results in a period of low platelet counts. Paradoxically, despite thrombocytopenia and associated bleedings, there is also an increase of thrombotic events including veno-occlusive disease/sinusoidal obstruction syndrome (VOD/SOS), of which the underlying mechanisms are yet unknown. While platelet counts are routinely monitored to evaluate their engraftment, platelet function after HSCT has not been studied so far. Here, we aimed to comprehensively assess platelet phenotype and function during the course of transplantation and engraftment ([Fig. 1]).

Fig. 1 GPIIb/IIIa activation defect and procoagulant platelets in HSCT patients; (A) Stimulation of platelets from HSCT patients with high doses of CRP-XL or (B) thrombin receptor activator peptide 6 (TRAP-6) revealed markedly impaired GPIIb/IIIa activation. (C) Platelets in HSCT patients exhibit a procoagulant phenotype as indicated by increased annexin V (AnnV) binding.

Method 13 adult allogeneic HSCT patients treated for acute leukemia were recruited at the University Hospital Würzburg (EV 301/21). Patients were conditioned by total body irradiation (TBI) with chemotherapy or chemotherapy alone before receiving 3.8 – 9.6 x 10⁶/kg mobilized CD34⁺cells. Platelet function was studied by flow cytometry before conditioning (t₁), after platelet engraftment (3 consecutive days over 20/nL, t₂) and after 30 (t₃), 60 (t₄), and 100 days (t).

Results Surface expression of GPIb/V/IX and GPIIb/IIIa was overall unaltered after HSCT compared to healthy controls (HCs). CD9 exhibited a transient decrease at t₂ while GPVI was downregulated at all time-points. We detected no differences in the number of PAC-1⁺or CD62P⁺resting platelets, indicating that platelets were not pre-activated in response to myeloablation and HSCT. Upon stimulation with high doses of ADP, TRAP-6, CRP-X_L or convulxin, we detected massively impaired integrin activation, still present at t₅, whereas α-granule release was only mildly compromised compared to HCs. Subpopulation analysis of activated platelets revealed a decreased percentage of CD62P^-PAC-1⁺and CD62P⁺PAC-1⁺but not CD62P⁺PAC-1^- events in HSCT patients, suggesting that integrin activation and P-selectin exposure become decoupled. Platelet δ-granule loading and release was monitored indirectly by flow cytometry using mepacrine. Mepacrine uptake was markedly impaired in platelets of HSCT patients at t₂, t₃ and t₄. In line with this, the relative mepacrine release in response to activation was significantly impaired until t₅, implying that there is a sustained paucity of δ-granules in these platelets. HSCT promotes activation of plasmatic coagulation. Intriguingly, we found massively upregulated annexin V binding and reduced mitochondrial membrane potential, indicative of procoagulant platelets.

Conclusion We demonstrate for the first time that platelets in HSCT patients exhibit a severe integrin activation defect and a paucity of δ-granules, which is uncoupled from α-granule release. Further, we provide evidence that this is associated with a massive increase of procoagulant platelets, which are prone to promote thrombosis. Our data suggest a mechanism why both bleeding and thrombotic events can occur after HSCT.

Publication History

Article published online:
26 February 2024