Thrombotic Risk Assessment in a Prospective Cohort of Newly Diagnosed Ambulatory Cancer Patients Candidate to Chemotherapy

Introduction: The occurrence of venous thromboembolism (VTE) during chemotherapy may result in treatment delays with unfavorable effects on cancer outcome, therefore VTE prevention is a relevant issue. Current guidelines recommend primary thromboprophylaxis in select high-risk cancer outpatients undergoing chemotherapy. Hence, the use of risk assessment models involving both clinical and biological parameters is increasingly important.

In this analysis, in a large prospective cohort of patients with newly diagnosed metastatic non-small cell lung (NSCLC), colorectal (CRC), gastric (GC) or breast (BC) cancers, we assessed whether pre-chemotherapy levels of thrombotic biomarkers may help to stratify patients at different risk levels of VTE during the first 6 months of anti-tumor therapy.

Methods: The study cohort included patients with advanced cancer enrolled from January 2012 to December 2017 in the HYPERCAN study (ClinicalTrials.gov, ID# NCT02622815), an ongoing Italian prospective, multicenter, observational study (Falanga et al, Thromb Res 2016). Clinical data and blood samples were collected at enrollment and during scheduled follow-up visits. Baseline plasma levels of fibrinogen (Clauss method), d-dimer (immunoturbidimetric assay), prothrombin fragment F1+2 (ELISA), FVIIa-AT (ELISA) and thrombin generation (TG, CAT assay) were measured. The study protocol was approved by the local Ethics Committee. Informed written consent was obtained from all study subjects. The outcome measure of the current analysis is the time to the occurrence of first symptomatic and/or incidental VTE, objectively confirmed.

Results: A cohort of 922 patients [median age: 66 (30-92) years] with metastatic NSCLC (n=416), CRC (n=262), GC (n=105) and BC (n=139) was available for analysis. After a median follow-up of 327 days (range 44-1,548), VTE occurred in 122 (76 M/46 F) patients [median age 64 y (range 40-85)] providing a cumulative incidence of 19.7% (CI 95%: 16-23). According to tumor site, VTE frequency was: NSCLC (n= 64, 15.3%) > CRC (n= 38, 14.5%) > GC (n= 12, 11.4%) > BC (n= 8, 5.7%). Biomarkers' analysis of baseline plasma samples from 598 patients showed significantly higher levels of fibrinogen, d-dimer and TG in the cancer cohort compared to healthy control subjects (p<0.01). NSCLC patients had significantly (p<0.001) higher levels of d-dimer compared to CRC and BC patients, and of fibrinogen (p<0.001) compared to CRC, BC and GC patients. Cumulative incidence of VTE at 6 months was 10% (CI 95% 8-12). In particular, patients who had VTE in 6 months presented with significantly higher pre-chemotherapy levels of D-Dimer than VTE-free subjects (p<0.05). Cox-multivariate analysis identified as independent risk factors for VTE, high d-dimer (HR: 2.1, CI 95% 1.2 - 3.7; p=0.008), high peak of TG (HR: 1.8, CI 95% 0.99 - 3.2; p=0.052), and site of primary tumor (non-BC vs BC) (HR: 3.3, CI 95% 1.01 - 10.5; p=0.048). A score for VTE prediction was created with these variables and three risk categories (low, intermediate, and high) were generated. The cumulative incidence of VTE at 6 months in each risk category was 3.9%, 9.8%, and 19%, respectively (low vs high HR: 2.36 p=0.005; intermediate vs high HR 2.16 p=0.006). Differently, the application of the Khorana score failed to identify high risk patients.

Conclusions: Our prospective study in a large cohort of metastatic cancer patients shows the high burden of VTE during the first 6 months of chemotherapy. Furthermore, laboratory data from this analysis enable us to create a risk scoring system based on d-dimer and TG together with tumor site, that significantly identifies patients at the highest VTE risk. Further investigation are worth to externally validate this score for clinical use.

Project funded by AIRC "5xMILLE multiunit extension program" n. 12237 grant from the "Italian Association for Cancer Research (AIRC)".