Incidence and risk factors of thromboembolism (TE) in advanced gastric cancer (AGC) receiving chemotherapy: A prospective observational study (NCT01047618).
31 Background: The incidence of TE in gastric cancer patients (pts) is known to be high. But because the previous reports were retrospectively analyzed in heterogeneous population, they give us only limited information. We therefore conducted a prospective study to investigate the incidence of TE and prognostic factors related with TE in AGC pts receiving chemotherapy. Methods: We checked D-dimer and coagulation battery at the start of chemotherapy and every 3 months thereafter. If there developed symptoms or signs of TE, or if D-dimer elevated 5 μg/mL or more we checked imaging studies to detect TE. The chemotherapy regimen mainly included fluoropyrimidine plus platinum-based for 1st-line, taxane-based for 2nd-line, and irinotecan-based for 3rd-line chemotherapy. Results: Between Nov 2009 and Apr 2012, 241 pts were analyzed. They received median 9 (range 1 - 42) cycles of chemotherapy. During the median observational duration of 16.7 months, 32 events (13.3%, 95% CI; 8.9 - 17.7%) of TE were detected. The types of TE were as follows; deep vein thrombosis (DVT) only in 18 (56.3%), pulmonary embolism (PE) only in 4 (12.5%), DVT and PE in 5 (15.6%), cerebral infarction in 4 (12.5%), and intra-abdominal arterial thrombosis 1 (3.1%) pts. The 1-year and 2-year cumulative incidences of TE were 15.0% (95% CI, 9.6 - 20.0%) and 20.0% (95% CI, 12.1 - 26.9%), respectively. The incidence rate of TE was 14.1 (95% CI, 9.6 - 19.9) events/100 person-years. In univariate analysis, the previous gastrectomy history, baseline CA72-4 level and baseline D-dimer level were statistically significant risk factor related with TE development. But in multivariate analysis, baseline D-dimer level was the only independent risk factor associated with TE development (Hazard ratio 2.46 [95% CI, 1.08 - 5.63], P= 0.033). Among 32 pts with baseline D-dimer 5.0 μg/mL or higher, 8 pts (25.0%) developed TE, while for pts whose baseline D-dimer was lower than 5.0 μg/mL, 24 out of 209 pts (11.5%) developed TE. Conclusions: The incidence rate of TE in AGC pts receiving chemotherapy was 14.1 (95% CI, 9.6 - 19.9) events/100 person-years. D-dimer was an important prognostic factor related with TE development. Clinical trial information: NCT01047618.