200PImplementation of artificial intelligence (AI) for lung cancer clinical trial matching in a tertiary cancer center

Background: Cognitive computing has promising potential to assist trial matching efficiency and accuracy by utilizing natural language processing and performing background analytics. The Watson for Clinical Trial Matching (CTM) cognitive system derives patient and tumor attributes from unstructured text in the electronic health record that can be matched to complex eligibility criteria in trial protocols. The Watson for CTM system was trained by Mayo Clinic subject matter experts in collaboration with IBM computer scientists/engineers and implemented in the Breast Oncology practice in July 2016. Metrics have shown an average monthly enrollment increase of 84% for breast systemic therapy trials. Methods: Training of Watson for CTM has continued with inclusion of additional cancers and expansion of trial types including Phase 1, supportive care, biomarker and observational trials. Watson for CTM was piloted in Lung Oncology in July 2018 and fully implemented in October 2018. Clinical research coordinators (CRCs) validated Watson-derived clinical trial matches on the day prior to patient clinic visits. A list of matched trials for each patient was given to providers to facilitate treatment decision making at point of care. Screening and timing metrics were tracked and compared with manual screening methods. Results: Watson for CTM facilitated screening of all lung cancer patients against 42 trials. Based on preset criteria, matches were validated by CRCs and provided to lung oncology providers in 69% (1818/2637) of patients’ visits from July through December 2018. Watson CTM-assisted patient matches resulted in a more complete list of potentially eligible trials and were completed in less than 50% of the time as compared to the traditional manual method. Enrollment data to define the impact of Watson for CTM and a screening team in the lung oncology practice is immature and will be subsequently reported. Conclusions: Implementation of the Watson for CTM system with a screening team enabled high volume patient screening for a large number of clinical trials in an efficient manner and promoted awareness of clinical trial opportunities within the lung oncology practice. Legal entity responsible for the study: The authors. Funding: Mayo Clinic. Disclosure: A.S. Mansfield: Funding to institution for participation on advisory boards: AbbVie, Genentech, BMS; Research funding to institution: Verily, Novartis. T. Halfdanarson: Research support: Ipsen, Thermo Fisher Scientific, Agios, ArQule; Consultancy (advisory boards): Lexicon, Advanced Accelerator Applications, Novartis, Curium. S. Coverdill: Employment: IBM Watson Health; Stock Ownership: IBM. M. Rammage: Employment: IBM Watson Health; Patent, royalties or other intellectual property: IBM. T. Haddad: Past consultant: TerSera Therapeutics; Research funding: Takeda Oncology. All other authors have declared no conflicts of interest.