Abstract PO3-10-07: Leveraging AI to identify factors influencing access to care and their association with overall survival- a multiracial Breast Cancer cohort

Abstract Introduction: Data from the Centers for Disease Control and Prevention (CDC) labels each US state based on its 5-Year Relative Survival of Breast Cancer (BC) patients (pts). Data is categorized into four “survival groups'', each representing a range of 5-year survival percentages. Using real-world data, this study characterized the populations in each survival group, examining pt knowledge and education about their disease as well as preferences and factors influencing their decision to explore clinical trial (CT) opportunities. Methods: Leal is an AI-based platform that employs self-reported medical profiles to match cancer pts with CTs. This study involved a diverse cohort of 14,509 BC pts who completed the questionnaire. The geographic distribution of pts across US states was analyzed, taking into account the state’s 5-year BC survival %. Four groups were defined as follows: Group A (86.4 - 89.4%), Group B (89.4 - 90.4%), Group C (90.4 - 91.5%), and Group D (92.0 - 95.0%). Parameters such as race, willingness to travel, insurance coverage, performance of genetic testing, and biomarker knowledge were considered in the analysis. Results: We found significant variations in registration rates on the Leal platform across groups. The Highest registration rate was recorded for Group D which has the highest 5-year survival rate (p 0.0001). Group A which has the lowest 5-year relative survival had a significantly higher proportion of African Americans compared to group D (14% vs 4%, p 0.0001). Moreover, while there was no difference in the percentage of pts undergoing Next-Generation Sequencing (NGS) in both groups, a larger proportion of pts in Group D demonstrated proficiency in interpreting the test reports (p = 0.05). Although no significant differences were found in the proportion of pts who were aware of their treatment (tx) history between the two groups, whites in both groups exhibited greater knowledge of their tx (p = 0.001). Additionally, pts in Group D showed a significantly higher willingness to travel longer distances to participate in CTs (p=0.001), and a larger percentage of them had insurance coverage. Conclusions: This study used data from CDC that categorized US states into four groups based on BC pts’ 5-Year Survival. CT participation is associated with better clinical outcomes, especially for pts with poorer prognosis. In agreement, we show a higher signup rate to the Leal platform in States with higher 5-year relative survival, possibly pointing to higher CT education and knowledge among pts in those states. This is further supported by higher proficiency in interpreting NGS reports, resulting in better-optimized CT options on the Leal platform. Group A was also characterized by a reduced willingness to travel for a CT and a larger proportion of pts without health insurance. Those are significant barriers for CT enrollment, and ultimately limit access to optimal care and may impact outcomes. Gaps in knowledge of key disease parameters such as tx history were observed for African Americans, regardless of geographic location, highlighting the need to focus on minority populations even in states where the survival rate is relatively high. In agreement, the proportion of African Americans in group A was markedly higher, correlating with current disparities in access to tx and poorer outcomes. AI-based platforms are optimized for the exact identification of barriers to access to care and for enabling relevant healthcare and support organizations to effectively address individual patient issues, consequently increasing access to CTs, particularly among different racial and ethnic populations. Citation Format: Michal Safran, Yelena Lapidot, Tzvia Bader, Avital Gaziel. Leveraging AI to identify factors influencing access to care and their association with overall survival- a multiracial Breast Cancer cohort abstract. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO3-10-07.