Abstract P6-01-08: Gene signatures provide independent prognostic information in elderly breast cancer patients

Abstract Background: Elderly breast cancer patients (≥70 years old) are under-represented in clinical trials and remain an undertreated population. Gene expression signatures have been shown to add additional prognostic information beyond that of routine clinicopathological factors, however their utility in elderly breast cancer patients remains unclear. As such, the main aim of this study is to determine if gene signatures can provide prognostic information that may help to aid treatment decisions for elderly breast cancer patients. Material and methods: Research versions of the genomic grade index (GGI), 70-gene, 21-gene recurrence score (RS), cell cycle score (CCS), PAM50 and PAM50 Risk of Relapse score - Proliferation (ROR-P) signatures were applied to 39 open access breast cancer datasets totalling 9583 patients. After filtering based on age ≥ 70 years old, the presence of Estrogen Receptor (ER) and survival information availability 871 patients remained. The prognostic capacity of signatures was tested in all (N=871), Estrogen Receptor positive/Lymph node positive (ER+/LN+, N=335) and Estrogen Receptor positive/Lymph node negative (ER+/LN-, N=374) patients using Kaplan-Meier and multi-variable Cox proportional hazard modeling. Models were adjusted for tumor size, grade, ER and lymph node status in all patients, and tumor size and grade in the ER+/LN+ and ER+/LN- subgroup analyses. Recurrence Free Survival (RFS) censored at 10 years was used as clinical endpoint and defined as the time from date of curative surgery to the time of recurrence or death. Both loco-regional recurrences and distant metastatic events were included in this endpoint. Results: Tumours from patients ≥ 70 years of age showed high levels of ER (87%), were large (69% ≥ 20 mm) and were of intermediate or high grade (82%). All gene signatures were statistically significant in Kaplan-Meier analysis of all and ER+/LN+ patients (Logrank P < 0.001). This significance remained in multi-variable analysis (Cox proportional hazards, P ≤ 0.05) with the exception of PAM50 which showed a trend (P ≤ 0.1) in ER+/LN+ patients. In ER+/LN- patients the GGI, 70-gene, PAM50, ROR-P, and CCS signatures were significant in Kaplan-Meier analysis (Logrank P ≤ 0.05) but only the 70-gene, PAM50, ROR-P, and CCS signatures remained so in multi-variable analysis (Cox proportional hazards, P ≤ 0.05). Conclusions: In general, we found that gene signatures provide statistically significant prognostic information in Kaplan-Meier and multi-variable analyses of all, ER+/LN+ and ER+/LN- breast patients over the age of 70. Table 1: Multivariable proportional hazard (Cox) analyses for all gene signatures for patients above age of 70. NOTE: Bold values indicate P < 0.05. aAdjusted for tumor size, tumor grade, estrogen receptor status, and lymph node status. bAdjusted for tumor size and tumor grade. Citation Format: Miguel Castresana Aguirre, Annelie Johansson, Linda S. Lindström, Nick Tobin. Gene signatures provide independent prognostic information in elderly breast cancer patients [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P6-01-08.