Background
Despite the frequency of venous thromboembolism (VTE) in pancreatic cancer, it is inconsistently reported as an adverse event in clinical trials. We hypothesized that reported rates of VTE in pancreatic cancer clinical trials are influenced by the objectives of the trial, with higher rates reported in thromboprophylaxis compared with chemotherapeutic trials. We performed a systematic review and meta‐analysis of randomized, controlled trials (RCT) in pancreatic cancer to quantify differences in reported rates of VTE in thromboprophylaxis and chemotherapeutic trials.
Methods
We systematically searched MEDLINE, EMBASE, and Clinicaltrials.gov. Eligible thromboprophylaxis RCTs were required to report rates of thrombosis in non‐anticoagulant pancreatic cancer cohorts. Eligible chemotherapy studies were RCTs evaluating chemotherapy regimens in advanced pancreatic cancer and reported thrombosis as adverse events. Pooled event rates of VTE and arterial thrombosis were calculated using a random‐effects model.
Results
The pooled VTE rate in 13 chemotherapy studies (5694 patients) was 5.9% (95% confidence interval [CI], 3.9‐9.0%) compared with 16.5% (95% CI, 11.7%‐23.3%; P < .001) in 9 thromboprophylaxis studies (631 patients). The pooled symptomatic VTE rate from chemotherapy studies was 5.4% (95% CI, 3.5%‐8.3%), which was significantly lower than the pooled rate from thromboprophylaxis studies of 10.5% (95% CI, 7.3%‐14.9%; P = .02).
Conclusion
The VTE incidence reported in chemotherapy RCTs in pancreatic cancer is significantly lower than reported in thromboprophylaxis studies. This finding highlights the underrecognition of VTE in chemotherapeutic trials and emphasizes the need to standardize approaches towards monitoring and reporting of VTE in clinical trials.