A bit deeper on hypercoagulability and cancer: focus on longitudinal trend of procoagulant microvesicles in gastro-intestinal malignancy

Introduction. It is fully recognized that cancer patients are at significant risk of developing thrombotic events (VTE). The prevention of such complications is of the utmost importance from a clinical point of view, seeing as they play a considerable part in the morbidity and mortality of these patients. The main issue is that the pathogenesis of the cancer-associated coagulopathy is complex and multifactorial. To assess the level of risk and identification of patients at high risk for thrombosis, the guidelines recommend including the detection of plasma thrombotic markers in “score systems” combining clinical and biological markers. However, current scores have been shown to perform poorly in predicting VTE in cancer patients. Thus, the identification of novel biomarkers associated with the grade of hypercoagulability in individual malignancies is required to drive the development of cancer type–specific scoring systems with improved predictive value. Aim of the study. We conducted a longitudinal cohort study to evaluate the trend of several coagulation parameters in patients with gastro-intestinal cancer with particular focus on circulating microvesicles (MVs) and MV-tissue factor (TF) activity. Our primary outcome was the description of coagulation fluctuations over a 6-month period following cancer diagnosis and the secondary outcome was the association between coagulation parameters and the occurrence of VTE complications. Material and Methods. Patients with a new diagnosis of gastro-intestinal cancer who underwent surgery were consecutively enrolled at the Padua University Hospital. Exclusion criteria were: cancer recurrence, severe liver or renal failure, Karnofsky Performance Status <60%, recent venous/arterial thromboembolism, pregnancy/puerperium, overt/recent sepsis. Longitudinal blood samples were collected at baseline, 7 days after surgery, 1 and 6 months after surgery. Each patient was followed for at least 6 months and for a maximum of 12 months. The primary outcome was the evaluation of coagulative parameters trend over a 6-month period following the diagnosis. Coagulation test performed included: factor (F)VIII and fibrinogen levels; D-Dimer and plasminogen activator inhibitor (PAI-1) antigen; hereditary thrombophilia; thromboelastometry; contact activation system (FXIIa-C1-inhibitor (C1INH), FXIa-C1INH and KAL-C1INH complexes); MV-TF activity; circulating MVs. Clinical outcomes recorded during the follow-up were: i) surgical radicality (i.e. complete or incomplete); ii) cancer severity (i.e. localized or advanced cancer); iii) any thrombotic event including superficial vein thrombosis, symptomatic or asymptomatic VTE or thrombosis in unusual sites. The secondary outcome was the association between the coagulative profile at the different time-points and the clinical outcomes. Results. Ninety-three patients (25 with pancreatic, 33 with colon and 35 with gastroesophageal cancer) were enrolled. The median clinical follow-up was 6 months [6-8.5] for pancreatic, 8 months [7-11] for colon, and 6.5 months for gastric cancer [6-9.25]. VTE incidence rate was 9.21 [95%CI 3.37-20.4] per 100 person-years for pancreatic cancer, 6.69 [95%CI 2.13-16.2] per 100 person-years for colon and 10.4 [95%CI 4.24-21.7] per 100 person-years for gastric cancer. The subgroup of pancreatic cancer at baseline showed increased levels of FVIII, D-Dimer, PAI-1 antigen, MV-TF activity and circulating MVs compared with the other cancer subtypes. In the overall cancer population, baseline contact system complexes were increased compared to levels measured in a reference healthy population, and pancreatic and gastric cancers showed the highest activation. Patients receiving chemotherapy at the 6-month time point showed significantly higher levels of FXIIa-C1INH and kallikrein-C1INH complexes compared to patients without chemotherapy. In a multivariate model, levels of MV-TF activity were independent predictors of incomplete surgical resection (OR 2.25 [1.25-7.0]) and cancer severity (OR 1.87 [1.20-3.8]). We observed that the majority of MVs detected were small (diameter 0.2-0.4 µm). Moreover, we confirmed that PS-negative MVs are the majority and thus PS is not the most suitable marker to detect the total number of MVs. MV-TF activity correlated with PS+MVs big and small, with endothelial MVs and big tumour MVs. Endothelial MVs, as well as MV-TF activity, showed a positive association with surgical radicality and cancer severity (OR 1.19 [1.04-1.36] and 1.30 [1.05-1.6], respectively). Levels of MV-TF activity ≥0.19 pg/mL conveyed a 2.38 [1.81-4.11] HR for VTE occurrence. Furthermore, baseline levels of FXIa >0.61 nM conveyed a 1.66 [1.02-2.9] HR to develop VTE over a median follow-up period of 6 months after diagnosis. This prediction model was adjusted for age, sex, BMI, cancer type, severity, and surgical radicality. Conclusions. Hypercoagulability in gastro-intestinal cancer is mainly mediated by high levels of FVIII, increased levels of complexes derived from the activation of the contact system, high MV-TF activity and increased levels of PS+MVs, endothelial and tumour MVs. Pancreatic cancer showed the most hypercoagulable profile. The prothrombotic factors remained altered up to 6 months after surgical resection of the neoplasm even in patients with surgical radicality, indicating that cancer-associated hypercoagulability persists months after tumour removal. Increased MV-TF activity and endothelial MVs are independent predictors of advanced disease and incomplete surgical resection. Finally, baseline increased levels of MV-TF activity and FXIa were independent predictors of VTE occurrence over the 6 months following cancer diagnosis. As TF is upregulated in cancer, it seems reasonable to hypothesize that concomitant activation of both the intrinsic and extrinsic pathways may act synergistically to produce a highly prothrombotic state in cancer.

NA