Abstract PS1-09-23: Physicians’ Perception & Impact of Quality Improvement Intervention: AI-enabled HER2-low Metastatic Breast Cancer Patient Notification Program

Abstract Background: Trastuzumab deruxtecan (T-DXd) was initially approved (Aug 2022) for treatment (trt) of HER2-low (IHC 1+ or IHC 2+/ISH-) unresectable or metastatic breast cancer (mBC) who have received a prior chemotherapy (CT) in the mBC setting. At approval, physicians in the US faced barriers in identifying patients (pts) with HER2-low for appropriate management. Florida Cancer Specialists & Research Institute (FCS) implemented a quality initiative (QI) wherein a random sample of 365 pts with HER2-low mBC were identified from the EMR using artificial intelligence (AI), and on 10/24/2022, physicians received an email alert identifying pts with HER2-low and mentioning the newly approved T-DXd. A retrospective study was conducted to evaluate the impact of the AI-enabled HER2-low notification program. Methods: The FCS EMR data were utilized to identify - Cohort 1 (pts with HER2-low mBC receiving CT on 10/24/2021 who initiated next line from 10/25/2021‒10/23/2022) and Cohort 2 (receiving CT on 10/24/2022 and initiated next line from 10/25/2022‒10/23/2023). Trt patterns were assessed descriptively, and real-world progression free survival (rwPFS) was assessed using Kaplan-Meier curves for both cohorts. A sensitivity analysis was performed to assess immediate changes in trt utilization following intervention by calculating monthly T-DXd use per eligible pts. A survey was emailed to FCS physicians to evaluate their familiarity with HER2-low trt guidelines and their perceived value of such QI. Results: 77 and 69 pts in Cohorts 1 (median age: 63 yrs) and 2 (61 yrs), respectively, were included. Within 1 year, a significant increase in T-DXd utilization (p