P2.01-50 Thromboembolism in ROS1 Rearranged Non-Small Cell Lung Cancer

The risk of thromboembolism (TE) is estimated to be 0.001% per year and hereditary thrombophilia present in 0.2-5% of the general adult population. It is estimated that 8-15% of patients with advanced non-small cell lung cancer (NSCLC) will develop thromboembolism throughout their disease course. In ALK rearranged NSCLC this incidence has been reported to be three to four fold higher at 36%. We sought to investigate the incidence in TE in a cohort of the rare ROS1 rearranged molecular subgroup. Electronic records were reviewed in six tertiary hospitals to identify all patients diagnosed with ROS1 NSCLC. Predefined data were analysed in STATA15 software for Kaplan-Meier (KM) plot and Cox-regression modelling. Forty-two patients were identified at data cut (31/1/2018). Median follow up was 10.9 months (mo); 38% had died. Median age was 53 years (31-80); 74% were female; 67% non-Asian and 88% non- smokers and 29% CNS disease during diagnosis. Median overall survival (OS) overall was estimated 28.8mo (range: 0.1-180.4mo). Thromboembolism incidence was 48% (n=20); 19 venous, 11 with pulmonary embolus (PE). One patient presented with fatal arterial TE (ROS1 diagnosed pre-mortem). Median time to TE was 2.3mo. In four patients TE was the harbinger for their diagnosis and two had encountered unprovoked TE prior to diagnosis, one then recurrent TE throughout diagnosis. Screening for a thrombophilia (TP) was not mandatory, and tested in n=9/20 cases with TE, one without. A co-occurring TP was identified in 15% (n=3/20); two factor V Leiden; one anti-thrombin III (ATIII) deficiency. Of interest, one patient with TE had factor XII deficiency; one thalassaemia minor and one acute-promyelocytic leukaemia without evidence of disseminated intravascular coagulation. Age; race; baseline ECOG; smoking history; treatment received; brain metastases at diagnosis; hereditary thrombophilia and neutrophil to lymphocyte ratio were not predictive of TE in this small cohort. Median OS in patients with TE was 21.3mo (0.1-180.4) versus 28.83mo (1.2-63.5) with no TE; hazard ratio 1.16 (95%CI 0.43-3.15, p=0.77). The ROS-1 fusion partner was known in two cases, both CD74-ROS1, one encountering TE (PE). Twenty-five have archival tissue available to identify the fusion partner and explore an association with TE. To the best of our knowledge, this is the first report on the incidence of thrombotic events specifically in ROS1-rearranged NSCLC. In this small cohort intriguingly there was a high incidence of TE and a higher incidence than the general population of hereditary thrombophilia.