Venous thromboembolism among patients with advanced lung cancer randomized to prinomastat or placebo, plus chemotherapy.

Two clinical trials have suggested that the combination of vascular endothelial growth factor inhibitor with chemotherapy is associated with venous thromboembolism (VTE). This retrospective cohort study investigates whether a similar association exists when matrix metalloproteinase inhibitor (prinomastat) is combined with chemotherapy. Patients (n=1,023) with stage IIIB, IV, or recurrent non-small cell lung cancer (NSCLC) were followed during 2 randomized, double-blind trials of prinomastat versus placebo orally bid, plus gemcitabine/cisplatin (GC) or paclitaxel/carboplatin (PC). VTE included deep venous thrombosis (DVT) or pulmonary embolism (PE) confirmed by imaging or autopsy. Risks identified in univariate analysis (incidence densities compared by t test) were confirmed in multivariate analysis (proportional hazards model). During 7,500.3 patient-months, 58 VTE (31 PE, 27 isolated DVT) were confirmed in 54 patients. On univariate analysis, VTE was associated with central venous catheter placed within 3 months, 15 mg prinomastat plus GC, and to a lesser extent, 15 mg prinomastat plus PC, baseline performance status, and histologic type. VTE incidence was not increased by 15 mg prinomastat alone (post-discontinuation of chemotherapy), by chemotherapy plus placebo, or by 5 or 10 mg prinomastat plus chemotherapy. On multivariate analysis,VTE hazards (95% confidence interval) were 5.69 (2.61, 12.40) with recently placed central catheter, 2.78 (1.42, 5.43) with 15 mg prinomastat plus GC, and 2.06 (0.98, 4.31) with 15 mg prinomastat plus PC; performance status and histology were nonsignificant. We can conclude that combined treatment with 15 mg prinomastat plus chemotherapy approximately doubles the hazard of VTE among patients with advanced NSCLC.