Venous Thromboembolism and Haemostatic Disturbances in Patients With Upper Gastrointestinal Cancer
Armand Troussau observed back in 1865 a particular condition of the blood that predisposes
to spontaneous coagulation in patients with gastric cancer. Today it is known that
adenocarcinomas in the gastrointestinal (GI) tract including tumours of colon and rectum is
accompanied by an hypercoagulable state.
Cancer is a well known risk factor for venous thromboembolism (VTE) such as deep venous
thrombosis (DVT) and pulmonary embolism (PE). In fact PE in cancer patients is the second
most common cause of death.
In an analysis of 13 cohort studies published with observational data Otten and Prins
concludes that the prevalence of VTE in cancer patients varied between 10 and 20%. Variation
could be explained with inconsistency. Patients suffering from thrombosis secondary to hemi
paresis and paraplegia were included. Patients where included at time of VTE and not at time
of cancer diagnosis.
Levitan et al looked at discharge diagnosis in 1211944 cancer patients and found among
gastric cancer patients a frequency of VTE at 85 per 10000 patients and in pancreatic cancer
110 per 10000. Gastric cancer patients had a relative risk of 1.49 compared to non-cancer
patients and pancreatic cancer patients a relative risk of 2.05 compared to non-cancer
patients. This put pancreatic cancer among the most prone malignancy to cause thromboembolic
complications. In a newly published study of 202 pancreatic cancer patients 19 developed
venous thrombosis resulting in a 58 fold increased risk compared to general population with
an incidence of 108/1000 patient-years and cumulative incidence after 6 months of 74/1000.
15 out of 19 cases of VTE occurred in the first 6 months since diagnosis of cancer. Tumours
located in the corpus or cauda had twice the risk of caput. Chemotherapy resulted in a 4.8
fold risk of venous thrombosis.
Postoperatively VTE is higher in cancer patients than in non-cancer patients after surgery.
A multicenter study including Danish patients - using phlebography - has shown VTE in 23% of
patients after surgery for colorectal cancer. Coagulation activity - as assessed by
sensitive biochemical markers - was related to tumour load.
Stender et al has in a newly published study shown a high preoperative prevalence of DVT in
193 colorectal cancer patients with a prevalence of 16% in women and 2.6% in men. In another
newly published study asymptomatic DVT is found ovarian cancer patients with elevated levels
of D-Dimer before treatment. The mechanism by which cancer induce VTE is at present subject
to great interest.
Plasma D-dimer (DD), a degradation product of cross-linked fibrin, is routinely used to
support or exclude the tentative diagnosis of DVT, combined with clinical assessment and
The accuracy of DD assays in cancer patients is unknown, despite their high negative
predictive values (NPV) in non cancer patients. A negative DD seems to exclude PE in cancer
patients, but DD in combination imaging techniques such as US or computed tomography (CT)
seems to improve diagnostic work-up but needs further investigation.
Observational Model: Cohort, Time Perspective: Prospective
Ole Thorlacius-Ussing, Professor, DMSC, MD
Aalborg Hospital department of Aarhus University Hospital
Denmark: National Board of Health