Introduction: Venous thromboembolism (VTE) is a leading cause of death in cancer patients, but routine thromboprophylaxis is untenable due to an unacceptably high rate of bleeding complications and judicious patient selection is critical. The Khorana score (KS) predicts the risk for chemotherapy-associated VTE, but the score has limitations because it is assessed only at one time point prior to initiation of treatment. Thrombin generation (TG) is a novel biomarker that assesses global coagulation activation and predicted increased VTE risk in cancer patients in the Vienna Cancer and Thrombosis Study (CATS). In CATS, patients with high peak TG had a 3- fold increased risk for VTE compared to those with low peak TG. A limitation of the study was that it enrolled a heterozygous cancer patient population, and peak TG was not reported by tumor type. Additionally, TG was only measured at enrollment into the study, not over time. The primary aim of our study was to determine if assessment of TG (both peak TG and endogenous thrombin potential (ETP)) improves VTE risk prediction compared to the KS alone during systemic therapy for lung and pancreas cancer. A secondary objective was to evaluate the impact of systemic therapy on peak TG and ETP over time.

Methods: This is an ongoing prospective study enrolling adults with newly diagnosed locally advanced or metastatic adenocarcinoma of the lung or pancreas who will receive their cancer therapy at Dartmouth Hitchcock Medical Center. Those with a history of active VTE or use of full dose anticoagulant medication within 30 days prior to enrollment are excluded. After informed consent, KS is calculated and blood is collected in sodium citrate tubes at 3 different time points (at initiation of therapy, and at the beginning of the 2nd and 3rd cycles of systemic therapy). Platelet-poor plasma is prepared by centrifugation and stored at -80°C until analysis by calibrated automated thrombogram (CAT; Thrombinoscope BV, Maastricht, Netherlands) using 1 pM tissue factor and 4 uM phospholipids to trigger coagulation reactions. Measurements are performed in triplicate for each specimen, and raw data are converted to peak TG and ETP. Information about symptomatic VTE events within 6 months of study entry is collected from the medical record.

Results: We present the results for 15 patients who have completed study-related procedures to date, including 6 lung and 9 pancreatic cancer patients with a median age 61 and 69 years, respectively. The majority of patients had metastatic disease and all patients received systemic therapy. The KS breakdown for the cohort is: KS1, 7%; KS2, 60%; KS3 27%; KS4 7%. Mean peak TG were 341, 324, 641 and 890 nmole, and mean ETP were 1459, 1685, 2155, 2051 nmole/min for KS 1,2,3 and 4, respectively, and both peak TG and ETP decreased over time after cancer therapy. VTE occurred in 4 patients (27%), and the median KS was 2 for patients with and without VTE. Mean peak TG prior to initiation of chemotherapy was similar in patients with and without VTE (359 and 389 nmole, respectively) but mean ETP was higher in patients with VTE (1959 vs 1760 nmole/min, respectively).

Conclusions: Though our sample size to date is small, preliminary results suggest that there is a direct correlation between KS and peak TG and ETP, and that ETP at diagnosis may be a more reliable predictor VTE than KS in lung and pancreatic cancer patients. Peak TG and ETP decreased progressively after initiation of systemic therapy, suggesting that the degree of coagulation activation is proportional to tumor burden. Though preliminary, these data suggest that TG is a promising biomarker of VTE risk. Enrollment to the study is continuing.

Funding: This study was funded by Northern New England Clinical Oncology Society.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.