IntroductionPatients with cancer commonly present with laboratory evidence of hypercoagulability. The HYPERCAN study (ClinicalTrials.gov, ID# NCT02622815) is an ongoing Italian prospective, multicenter, observational study, evaluating the predictive value of thrombotic markers for early cancer diagnosis in healthy subjects and for cancer prognosis and venous thromboembolism in patients with newly diagnosed cancer disease (Thromb Res, 2016).In this analysis of the HYPERCAN study, in a subcohort of consecutive patients with newly diagnosed metastatic non-small cell lung (NSCL) or gastrointestinal (GI) cancers, we aimed to assess whether the pre-chemotherapy levels of thrombotic biomarkers may be predictive of early disease progression (EDP, i.e. within 6 months) and early deaths (i.e. within 1 year).MethodsThe study cohort included 401 patients [median age: 65 (34-88) years] with metastatic NSCL (n=241) or GI (n=160) cancer. Clinical data were recorded at enrollment and during follow-up visits, disease progression being monitored by imaging. Fibrinogen (Clauss method, IL Werfen, Italy), d-dimer (immuneturbidimetric assay, IL Werfen, Italy), and thrombin generation (TG) potential (CAT assay, Stago, France) were measured in plasma samples collected at patient enrollment, before starting any curative chemotherapy and from a control group of 108 healthy subjects. The study protocol is approved by the local Ethics Committee. Informed written consent is obtained from all study subjects. Data are expressed as median (5th-95th percentile).ResultsSignificantly higher plasma levels of fibrinogen, D-dimer and TG potential were observed in the cohort of cancer patients compared to healthy controls (p<0.01). Among cancer types, fibrinogen levels were significantly greater in NSCL as compared to GI cancer patients [479 (258-934) vs 402 (220-668) mg/dl), p<0.001]. After two year follow up, the cumulative incidence of EDP was 31% (CI 95% 27-36). Of interest, patients with EDP (n=125) vs those without EDP were characterized by significantly (p<0.05) greater pre-chemotherapy levels of leucocytes [9.1 (5.6-21.4) vs 8.4 (4.6-18.2) x109/L], platelets [305 (131-586) vs 283 (148-513) x109/L], fibrinogen [475 (290-913) vs 421 (226-814) mg/dl], and TG endogenous thrombin potential (ETP) [1,919 (1,137-3,086) vs 1,752 (1,264-2,559) nM*min]. In GI cancer, but not in lung cancer patients, multivariate Cox regression analysis according to site of primary tumor identified d-dimer and TG ETP plasma levels, together with GI cancer diagnosis, as independent risk factors for EDP.Concerning mortality, after two year follow-up, the cumulative incidence of overall survival at one year was 63% (CI 95% 58-67), early deaths = 37%. The group of patients who deceased during the first year (n=145) showed significantly (p<0.01) higher leucocyte count [9.7 (5.4-21.6) vs 8 (4.6-16.1) x109/L], plasma fibrinogen [486 (250-846) vs 402 (223-816) mg/dl] and d-dimer [515 (97-3,174) vs 286 (55-1,456) ng/ml] levels. In GI cancer, multivariate Cox regression analysis identified d-dimer and TG ETP plasma levels, leucocytes and platelet counts, and GI cancer diagnosis, as independent risk factors for mortality within 1 year. Differently, none of the hemostatic biomarkers were predictive for mortality in NSCL cancer patients.ConclusionsThe current analysis of prospectively collected data shows the utility of detecting circulating thrombotic biomarkers in specific cancer types. In particular, in GI metastatic cancer, the levels of D-dimer and TG before starting chemotherapy, together with having a GI cancer, could predict EDP (within 6 months) and early deaths (within 12 months). This was not the case for NSCL cancer. Currently, new tests involving other coagulation biomarkers are ongoing and will be evaluated.Project funded by AIRC “5xMILLE” n. 12237 grant from the “Italian Association for Cancer Research (AIRC)”.* On behalf of the HYPERCAN investigators. DisclosuresNo relevant conflicts of interest to declare.