Venous thromboembolism (VTE) is strongly associated with
cancer [1–3]. Among the types, in pancreatic cancer (PC), VTE
incidences are among the highest [4–6]. Various biomarkers
like soluble P-selectin, thrombin/antitrombin complex, prothrombin fragment 1 þ 2 levels, and D-dimer have specifically
attracted investigation regarding their capacity for VTE prediction or its early detection [7–9]. Among these, general
cancer-related VTE studies have focused the most attention
on D-dimer, but D-dimer is not PC-specific.
In our earlier preoperative panels, in cases with no overt
signs of thrombosis, D-dimer failed to reveal the presence of
PC, whereas elevated fibrinogen and FVIII activity did
together with carbohydrate antigen 19-9 (CA 19-9) [10]. PC is
associated with a hypercoagulable state, accompanied by
neutrophil extracellular taps (NETs) [11,12], and the activity
of circulating tissue factor (TF) [13], which mostly resides on
microparticles [14].
PC is either the fourth or fifth leading cause of cancer
mortality in developed countries [15]. In particular, the development of symptomatic [16] or even of asymptomatic VTE
the first months after cancer diagnosis are associated with
lower survival rates and worse prognosis [17,18]. Prevention
of VTE recurrence would improve quality of life and extend
overall survival. To the best of our knowledge, no clinically
applicable PC-specific biomarkers yet exist for VTE prediction.
PC cells produce CA 19-9, also called sialyl-Lewis A (a
modified Lewis (a) antigen), a monosialoganglioside/glycolipid, which can be assessed in circulation. In PC, CA 19-9 is a
commonly used plasma biomarker [19,20], mainly to monitor
disease burden during treatment, and relapse during followup [21]; it has sensitivity limitations due to lack of sialylLewis A expression in 6% of the most populations [19,20].
Earlier, we analyzed cancer-associated incidence of VTE
based on a hospital electronic database of 42 245 cancer
patients treated at Turku University Central Hospital during
2005–2013 [4]. We confirmed that VTE incidence is high in
cancers such as PC. In this case-control study, our aim was to
survey, among patients with PC, the incidence and timecourse of VTE, and to evaluate, whether laboratory variables
in routine clinical care, including cancer biomarkers such as
CA 19-9 can predict VTE in PC patients. We had the advantage of several electronic hospital-record (EHR) databases in
one single tertiary care center (Turku University
Central Hospital).