Abstract P3-14-06: Concordance assessment of IBM Watson for Oncology with MDT in patients with breast cancer

Abstract

Abstract Objective: IBM Watson for Oncology (WFO) is an artificial intelligence cognitive clinical decision support system which provides evidence-based treatment recommendations in cancer. This study aimed to examine the level of agreement (concordance) between treatment recommendations made by WFO and a multidisciplinary tumor (MDT) board for adjuvant therapy in early stage breast cancer and first-line therapy in metastatic breast cancer. Methods: Treatment recommendations were provided by WFO and MDT board for 92 early stage and 40 metastatic breast cancer cases between Jan 2017 and Jan 2018. WFO treatment recommendations were provided in 3 categories: “Recommended”, “For Consideration” and “Not Recommended.” It was considered concordant if the tumor board recommendations were designated 'recommended' or 'for consideration' by WFO Version 17.10. Results: Overall treatment concordance between WFO and MDT was 76.51%, 73 cases with early stage breast cancer (79.35%) and 28 cases with metastatic breast cancer (70.00%) we concordant. Among all the 132 cases, 49 cases were in 'Recommended' category, 52 cases were in 'for consideration' category, and 26 cases were disconcordant. Subgroup analysis found that patients with stageIIIdisease or triple-negative breast cancer were the most likely to be concordant. Luminal A type and HER2-positive cases were less likely to be concordant than other types. In early stage breast cancer, based on the low recurrence risk, 11 cases and 2 cases were treated with TC and EC regimen instead of AC-T or CMF regimen recommended by WFO. 4 patients with HER2-positive were disconcordance by using TCbH or AC-TH regimen instead of AC-TH or CMFH regimen recommended by WFO. The main reason of disconcordance in metastatic cancer was that CDK4/6 inhibitors had not been approved in China. Conclusion: Treatment recommendations made by WFO and the tumor board showed highly concordant in breast cancer. Cancer stage and subtype had significant influence on concordance, and Watson for Oncology need to be optimized in localization. Citation Format: Xu J, Sun T, Hua S. Concordance assessment of IBM Watson for Oncology with MDT in patients with breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-14-06.