LAUSR

INTRODUCTION Increased thromboembolism (TE) has been reported in ALK+ and ROS1+ non-small cell lung cancer (NSCLC). MATERIALS AND METHODS Odds ratios (OR) and hazard ratios (HR) of TE were calculated from meta-analysis and time-to-event analysis respectively for either ALK+ or ROS1+ NSCLC patients. RESULTS We identified eight studies (766 ALK+, 143 ROS1+, 2314 non-ALK+ and non-ROS1+ NSCLC patients) for the meta-analysis. For ALK+ NSCLC, the pooled OR was 2.00 (95% CI: 1.60-2.50) for total TE (TTE) by random-effects model, 2.10 (95% CI: 1.70-2.60) for venous thromboembolism (VTE), and 1.24 (95% CI: 0.80-1.91) for arterial thromboembolism (ATE). For ROS1+ NSCLC, the pooled OR was 3.08 (95% CI: 1.95-4.86) for TTE, and 3.15 (95% CI: 1.83-5.43) for VTE. Six studies (739 ALK+, 137 ROS1+, 561 EGFR+, 714 "wildtype" NSCLC patients) were included in the time-to-event analysis. The TTE incidence rate was 17.4 (95% CI: 15.3-19.5) per 100 pateint-years for ALK+ NSCLC, and 32.1 (95% CI: 24.6-39.6) per 100 patient-years for ROS1+ NSCLC with a 50 % cumulative incidence rate at year 3 of diagnosis. HR for TTE was 2.35 (95% CI: 1.90-2.92, p < 0.001) and 3.23 (95% CI: 2.40-4.34, p < 0.001) for ALK+ and ROS1+ NSCLC, respectively. Comparing ROS1+ NSCLC to ALK+ NSCLC, HR for TTE was 1.37 (95% CI: 1.05-1.79, p = 0.020). CONCLUSIONS ALK+ and ROS1+ NSCLC patients had an increased risk of TE. ROS1+ NSCLC had further increased risk of TE over ALK+ NSCLC.