Abstract 451: Platelet Activation and Thrombosis by Antibodies against VEGF and CD40 Ligand: Extending the mechanism of HIT
Introduction: The vascular endothelial growth factor (VEGF) monoclonal antibody, bevaci-zumab (Avastin), has been associated with arterial thromboembolic events in some cancer patients. Another therapeutic antibody, hu5c8, which targets CD40L, also produced unexpected thrombosis in lupus clinical trials. Because platelets play a crucial role in arterial thrombosis, we hypothesized that antibodies against VEGF and CD40L may activate platelets via a mechanism similar to that responsible for thrombosis in Heparin-Induced Thrombocytopenia (HIT).
Methods: Immune complexes (ICs) were prepared by combining these monoclonal IgG antibodies with their antigens: M90 (or hu5c8) with CD40L (“M90+CD40L”), or bevacizumab with VEGF and heparin (“BVH”). VEGF binds heparin (as does platelet factor 4, the HIT antigen). We measured platelet activation by serotonin release assay, platelet aggregometry, and flow cytometry. We also evaluated IC-induced thrombosis in hFc mice, transgenic for the human IgG receptor, CD32.
Results: Similar to HIT antibodies, these ICs potently induced platelet activation dependent on CD32, IC concentration (>10nM) and optimal stoichiometry. Intravenous injection of M90+CD40L or BVH into hFc, but not wild-type mice rapidly produced signs of thrombotic shock, thrombocytopenia and pulmonary thrombosis. However, wild-type control mice (lacking platelet CD32) were unaffected by IC injection. Similarly as with HIT antibodies, bevacizumab IC activity was reduced in the absence or excess of heparin both in vitro and in vivo, whilst M90+CD40L-induced platelet activation was abolished in vitro by blockade of the platelet CD40L receptor, CD40, demonstrating a requirement for Fab-dependent anchoring. Furthermore, VEGF121 (which lacks the heparin-binding domain of VEGF165) and bevacizumab with or without heparin failed to activate platelets or cause thrombosis in hFc mice.
Conclusions: Together, these findings demonstrate that Fab-dependent anchoring of anti-CD40L and anti-VEGF ICs is required for potent platelet activation and thrombosis, as is the case in HIT, suggesting common mechanistic elements. Clinical implications may apply in patients with cardiovascular comorbidity receiving immunotherapy.