Association of anticoagulant use with clinical outcomes from crizotinib in ALK- and ROS1-rearranged advanced non-small cell lung cancers: A retrospective analysis of PROFILE 1001.

Abstract

ROS1- and ALK-rearranged advanced NSCLCs are associated with increased thromboembolic risk. We hypothesized that a prothrombotic phenotype offers an evolutionary advantage to subsets of these cancers. The impact of this phenotype could alter outcomes from targeted therapy. In a retrospective analysis of ROS1- and ALK-rearranged NSCLCs treated with crizotinib in a phase 1 trial, we compared progression-free survival (PFS) and objective response rate (ORR) based on the history of anticoagulation use (a possible surrogate of thromboembolism) at baseline (within 90 days before study enrollment) or within 90 days of study treatment. Twelve out of 53 (22.6%) ROS1- and 39 out of 153 (25.5%) ALK-rearranged NSCLCs received anticoagulation before or during the trial. Most ROS1 and ALK patients on anticoagulation received low-molecular-weight heparin (75% and 64.1%, respectively). In the ROS1-rearranged group, the median PFS (95% CI) values were 5.1 (4.4-14.4) and 29.0 (16.5-48.8) months, and the ORR values were 41.7% (95% CI: 15.2 to 72.3) and 80.5% (95% CI: 65.1 to 91.2) among those with and without anticoagulation treatment, respectively. In the ALK-rearranged group, the median PFS (95% CI) was 7.1 (5.4-7.7) and 12.0 (9.4-18.3) months, and the ORR was 41% (95% CI: 25.6 to 57.9) and 74.3% (95% CI: 65.3 to 82.1) among those with and without anticoagulation, respectively. Anticoagulation (as a potential surrogate of a prothrombotic subset) in ROS1- and ALK-rearranged NSCLCs may be associated with a lower PFS and ORR to crizotinib. gov: NCT00585195.