Decreased Platelet Reactivity and Function in a Mouse Model of Pancreatic Cancer

Introduction: Cancer patients have a dysregulated hemostatic system and activated platelets. Using a tumor-bearing mouse model, we have demonstrated that tissue factor (TF)-positive extracellular vesicles (EVs) released from pancreatic cancer cells activated the coagulation system as well as platelets and enhanced venous thrombosis. However, arterial thrombosis has not been studied in our mouse pancreatic tumor model.

Methods: We used a human pancreatic cancer cell line called BxPC-3 that releases a high level of TF+EVs and injected into the pancreas of male nude mice. Tumors were allowed to grow until they were ≥2.0 -≤ 3.2g. Control mice were injected phosphate-buffered saline. Tumor-bearing mice and control mice were subjected to a ferric chloride-induced carotid artery occlusion model of arterial thrombosis. Hemostasis was assessed using a tail transection model. Markers of activation of platelets and endothelial cells were measured in the plasma. The proportion of reticulated platelet, platelet activation after agonist stimulation, and platelet receptors were measured using flow cytometry. Platelet aggregation was evaluated using light transmission aggregometry.

Results: Surprisingly, we found that tumor-bearing mice had significantly prolonged carotid artery occlusion time as well as tail bleeding time compared to control mice. Tumor-bearing mice had a significant decrease in platelet count and a significantly higher percentage of reticulated platelets compared to control mice. In addition, platelets from tumor-bearing mice were significantly larger. These findings suggest that tumor-bearing mice had increased platelet turnover. Tumor-bearing mice had significantly higher levels of platelet factor 4, soluble P-selectin, soluble VCAM-1, and soluble ICAM-1 compared to control, indicating that tumor-bearing mice had activated platelets and endothelium. Platelets from tumor-bearing mice had significantly reduced PAR4 agonist peptide and convulxin-induced but not ADP-induced α_IIbβ₃ activation. However, there was no difference in surface expression of P-selectin on platelets activated with PAR 4 agonist peptide, convulxin or ADP compared with platelets from control mice. These results suggest that platelets from tumor-bearing mice are impaired in their ability to activate integrin receptors. However, we did not observe differences in the platelet aggregation in vitro between platelets from tumor-bearing mice and controls. Finally, we also analyzed if platelets in tumor-bearing mice showed signs of in vivo activation. Platelets isolated from tumor-bearing mice had a significant decrease in surface expression of GPIbα compared to controls.

Conclusion: Our findings suggest that in our mouse model of human pancreatic cancer there is chronic platelet activation, leading to thrombocytopenia, decreased GPIbα and impaired platelet adhesive function. Our study may help to explain the increased bleeding observed in some cancer patients.