P45.13 Efficacy and Safety of Crizotinib in Real-World ROS1-rearranged NSCLC: A Retrospective Canadian Cohort

Abstract

ROS1-rearranged non-small cell lung cancer (NSCLC), present in 1-2% of diagnoses, is a genetically distinct type of NSCLC which can be successfully managed through the use of targeted therapy, specifically the first-generation ALK-inhibitor, crizotinib, which doubles as highly effective inhibitor of aberrant ROS1 activity. Clinical trials have shown crizotinib to be safe and effective; however, the use of crizotinib and its safety and efficacy in real-world ROS1-rearranged populations requires further exploration. Consequently, this study initiated a retrospective, observational, population-based investigation into the safety and efficacy of crizotinib in ROS1-rearranged NSCLC patients diagnosed and treated in Alberta, Canada. Alberta patients with advanced or metastatic ROS1-rearranged NSCLC, receiving crizotinib as their first targeted ROS1-inhibitor between January 2014 and June 2020 were identified. Demographic, clinical, treatment and outcome data were extracted from the institutional Glans-Look Lung Cancer Research Database. 21 ROS1-rearranged crizotinib-treated patients were identified: median age was 51.6 years, 67% female, 86% never-smokers. 96% had metastatic disease at crizotinib initiation (38% M1c stage), 67% ECOG < 2, and 57% of patients had previous exposure to systemic chemotherapy (including platin-pemetrexed) and/or immunotherapy. 38% of patients were alive, 29% with crizotinib therapy ongoing, at the time of analysis. One-year survival was 47% following crizotinib initiation; median OS and PFS were 33.3 and 10.6 months, respectively. Disease control rate was 62%, while 14% showed primary resistance to crizotinib. ECOG was not significantly associated with outcome or clinical response. Adverse events (AE) were reported in 52% of patients, most common were gastrointestinal (29%) or investigational disorders (14%), and primarily grade 1 or 2. 19% of patients experienced grade 3+ AE and subsequently terminated crizotinib. Treatment breaks +/- dose modifications occurred in 16%. Patients remained on crizotinib for a median 6.9 cycles before discontinuation. Progressive disease was the most common reason for crizotinib termination (40% of crizotinib discontinuations); upon termination 24% received additional systemic treatment (range 1-3), all receiving additional ROS1-inhibitors. 33% of crizotinib terminations were due to death. 29% of patients had brain metastases at crizotinib initiation and 14% developed brain metastases while on crizotinib, predominantly managed with radiotherapy (67%). Presence/absence of brain metastasis did not significantly impact outcome, but baseline brain metastases resulted in a shorter duration of use and lower rate of clinical response to crizotinib. 48% had high PD-L1 expression (>50%) and 29% received immunotherapy (all prior to crizotinib initiation), with mPFS of 10.1 weeks and DCR of 33%. PD-L1 expression nor immunotherapy use had a significant impact on survival. This real-world study suggests crizotinib therapy for ROS1-rearranged NSCLC in a Canadian clinical setting is an effective, safe and tolerable therapy; however, this real-world cohort was unable to completely meet the outcomes found in crizotinib-treated ROS1-rearranged clinical trial cohorts. As a real-world population, lack of ROS1 testing, delayed access to crizotinib and heavy pre-treatment resulted in high disease burden and early failures on crizotinib, likely impacting crizotinib efficacy. This reinforces the need for real-world data alongside clinical trial data and highlights the need for expeditious access to appropriate targeted therapies in this population.