OA16.07 Patient-Driven Epidemiologic Assessment of ROS1-Fusion Driven Cancers

Abstract

There are limited data on factors that contribute to the development of ROS1-fusion positive cancers. A group of ROS1+ patients approached the Bonnie J. Addario Lung Cancer Foundation (ALCF) for support of ROS1 research. As a first step in this unique, patient-driven effort, we designed a survey to understand epidemiologic/clinical factors that may contribute to the development and progression of ROS1 cancers. We aim to collect data with biospecimens and make these available through an open-access data-sharing platform to accelerate ROS1 research. With guidance from ROS1 patients, we designed a 204-question survey, obtained Stanford University IRB approval. The survey on the ALCF website from May 18, 2016, was widely publicized through conferences, patient support networks, social media and community-based outreach. The survey questions address demographic-, clinical-, diagnostic and treatment- factors, family and reproductive history, dietary, exercise, environmental exposure and hormone and substance use. We used Z-proportions test for statistical significance defined as p<0.05. In 53 days, 95 global patients with ROS1-fusion+ cancer responded to the survey (median age at diagnosis 56 years). Respondents were 71% female (n=52/73), 79% never smokers (49/62). 71% respondents were otherwise healthy before diagnosis (n=46/65). The majority of respondents were diagnosed with lung cancer (n=68/72); and one each with gastric, ovarian, cervical and liver cancer. 76% reported metastatic disease (n=52/68) at diagnosis with the most common site of metastases as contralateral lung (65%) and bone (46%). 52% patients reported their ROS1 cancers were not detected at diagnosis (n=35/67); 80% did not know their specific translocation (51/64). 71% patients received molecular testing on physicians’ orders (n=45/63), with 21% self-initiating molecular testing. Despite the availability of targeted treatments and clinical trials, most patients were prescribed chemotherapy in their first (62%), second (49%), third (60%) and further lines of therapy. 76% patients reported that crizotinib was the therapy that worked best, with 96% reporting significant improvement in symptoms and QoL. We found no significant correlations between oral contraceptive/hormone/anabolic steroid use, occupational exposure, geographic area of employment/residence, family history of cancer, and incidence of ROS1+ cancer. This is a unique patient and non-profit advocacy group-driven investigation that seeks to understand factors that may influence development and treatment of ROS1 cancers. The results highlight patient-centricity, the importance of upfront molecular testing and targeted therapies. We report patient-reported experiences with ROS1 testing and durable responses to targeted treatments e.g. crizotinib. As the study is ongoing, we will update results in December 2016.