Abstract
Abstract 34
Thrombotic complications are common side effects in cancer patients and up to 20% of all symptomatic cases of deep vein thrombosis (DVT) are thought to be cancer-related. Since Armand Trousseau established a direct correlation between thrombophlebitis and cancer in 1865, cancer-associated thrombosis has been the focus of several studies. All tumor types are associated with thrombosis but the incidence is particularly high (up to 57%) in patients suffering from pancreatic cancer (Sack et al Medicine 1977; Blom et al Eur J Cancer 2006; Cronin-Fenton et al Br J Cancer 2010). Cancer cells shed procoagulant microparticles (MPs) that can reach the bloodstream of pancreatic tumor-bearing mice and of patients with pancreatic cancer (Thomas et al J Exp Med 2009, Zwicker et al Thromb Res 2009). At their surface, these vesicles express a higher density of active tissue factor (TF) when compared to their parental cell (> 100-fold), and were shown to contribute to vessel injury-induced thrombosis (Thomas et al J Exp Med 2009). Furthermore, the concentration of circulating TF-bearing tumor-derived MPs has been correlated with cancer-associated venous thromboembolic events in patients (Del Conde et al J Thromb Haem 2007; Hron et al Thromb Haemost 2007; Langer et al Ann Hematol 2008; Zwicker et al Thromb Res 2009). In this study, we addressed experimentally whether circulating tumor microparticles play a role in DVT associated with cancer. We used a mouse model of DVT where thrombosis is induced by flow restriction (90% reduction in the lumen size) of the inferior vena cava (IVC) (Brill et al Blood 2010). In contrast to most DVT models, this model maintains blood flow in the IVC and it does not induce injury or endothelial denudation and thus is suitable for comparisons to human patients. The DVT model was used in combination with a model of murine pancreatic cancer (Thomas et al J Exp Med 2009). We observed that all the mice implanted subcutaneously with pancreatic cancer cells (Panc02 cell line) and submitted 35 days later to partial flow restriction of the IVC (n=8) developed an occlusive thrombus after 3 hours of flow restriction, while only 2 out of 7 control mice did (P < 0.05). We observed in this DVT model that mice infused with cancer cell-derived MPs have the same phenotype as the tumor-bearing mice since 100% of them developed a thrombus (P < 0.05 when compared to control mice), indicating that circulating tumoral microparticles play an important part in the triggering of cancer-associated DVT. A pro-thrombotic effect could be observed as early as 1 hour after stenosis in cancer cell-derived MPs-injected mice (P < 0.05 when compared to control mice). Moreover, thrombus formation after 1 hour of stenosis could be abolished when mice were pretreated with hirudin, suggesting the importance of the MP-generated thrombin likely by MP-derived TF activity.
In conclusion, our results support the hypothesis that tumoral microparticles trigger cancer-associated venous thrombosis. Thus targeting their recruitment or activity could help to prevent cancer-associated DVT and its complication, pulmonary embolism.
Disclosures:
No relevant conflicts of interest to declare.
Upregulation of the expression of Wnt5a promotes the proliferation of pancreatic cancer cells in vitro and in a nude mouse model