Background: The rate of thrombosis after cancer surgery is high and cancer-associated thrombosis (CAT) is independently associated with mortality. Pancreatic cancer is most commonly associated with CAT, but this association is not well described for histology other than adenocarcinoma (ACPC).
Aim: To define the association of non-adenocarcinoma pancreatic cancer (NAPC) as a risk factor for post-operative CAT.
Methods: We conducted analysis of a prospectively collected large database of pancreatic cancer surgery. We randomly collected NAPC and compared 1:3 to ACPC. Along with comorbidities, demographics and cancer extension; we calculated the pre-operative Khorana score. All thrombosis were objectively confirmed with imaging. CAT, including imaging positive DVT or PE was the primary outcome. Categorical variables were presented as percentages, continuous variables as median and range. For Chi square and logistic regression SPSS was used.
Results: A total of 441 patients were included. Pyloric preserving Whipple and standard Whipple were the most common surgeries (n=230,52%), age 65.9 + 11.5, male 57% (n=252), 8% (n=36) had metastasis. IPMN and neuroendocrine were the most common NAPC. The median follow up was 449 days in which 90 (20%) patients developed CAT. The odds (OR 1.1, 95% CI 0.6-1.9, p=0.7) and time to VTE were not different between NAPC and ACPC. Given the variation of prophylactic strategies through the life of the database we analyzed for trends by year of surgery; there was no trend for NAPC (p=0.4) or ACPC (p=0.06). The Khorana risk score was not associated with CAT. In the multivariate analysis, only history of PAD (ORadj:13.9; p<0.001) length of stay >12d (ORadj:12.8; p<0.001), and CAD (ORadj:5.7; p=0.02) were associated with CAT in this highly selected population.
Conclusion: The risk of CAT after pancreatic cancer surgery is not governed by histology, moreover, the rate of CAT after pancreatic cancer surgery was high despite the standard extended prophylaxis practice. It is plausible that the organ and surgery complexity are the main reasons for the observed rate of CAT in pancreatic cancer.