Abstract
8526 Background: Estrogen receptor (ER) can activate MAPK signaling but the contribution of the two classical receptors—ER-alpha ( ESR1) and ER-beta ( ESR2) is unclear. Past trials targeting ER and EGFR in NSCLC lacked efficacy. We evaluated the association of ESR1& 2 expression with the genomic landscape and overall survival (OS) in NSCLC. Methods: NSCLC tumors (N = 21603) were tested at Caris Life Sciences (Phoenix, AZ) with NextGen Sequencing of DNA (592-gene or whole exome) and RNA (whole transcriptome). Mutation prevalence (-Mt) was calculated for pathogenic SNVs/indels. ESR1& 2 expression was split into quartiles (transcripts per million, top (-H) and bottom (-L) quartiles were compared). A transcriptomic signature associated with MAPK activation (MPAS, arbitrary units: AU) was applied. The X 2 test was applied, p-value was adjusted for multiple comparisons ( p < .05). Real-world OS was obtained from insurance claims and Kaplan-Meier estimates were calculated. Results: There was a greater proportion of females in ESR1-H (53%) versus (v) -L (45%, p < .05) but not in ESR2-H v -L (50% v 50%). There was a greater proportion of adenocarcinoma (AD) in ESR1-H (65%) v -L (44%) and a greater proportion of squamous (SCC) tumors in ESR2-H (27%) v -L (15%) ( p < .001 all). The prevalence of EGFR-Mt and KRAS-Mt was greater in ESR1-H v -L and smaller in ESR2-H v -L. (Table) Of all KRAS-Mt, there was a greater percent of KRAS G12C-Mt in ESR1-H (44%) v -L (39%, p = .001). In EGFR-Mt, ESR1-H/ ESR2-H had a higher MPAS (1.4 AU) than ESR1-H/ ESR2-L (.52), ESR1-L/ ESR2-H (.59) or ESR1-L/ ESR2-L (-1.6, p < .05). In KRAS-Mt, ESR1-H/ ESR2-H (1.7 AU) and ESR1-L/ ESR2-H (1.8) had a higher MPAS than ESR1-H/ ESR2-L (.7) or ESR1-L/ ESR2-L (-.7, p < .001). ESR1-H had longer median OS v -L (22 months (m) v 16 m, p < .001) as did ESR2-H v -L (23 m v 15 m p< .001). ESR1-H/ ESR2-H had the longest OS (25 m) followed by ESR1-H/ ESR2-L (18 m), ESR1-L/ ESR2-H (16 m) and ESR1-L/ ESR2-L (14 m, p < .001). In EGFR-Mt, a significant difference in survival since treatment (SST) with osimertinib was seen for ESR1-L/ ESR2-H (40 m, N = 13), ESR1-H/ ESR2-L (36 m, N = 62), ESR1-H/ ESR2-H (34 m, N = 161) and ESR1-L/ ESR2-L (30 m, N = 138, p = .03). In KRAS G12C-Mt, a significant difference in SST was seen with sotorasib; ESR1-H/ ESR2-H had the longest SST (median not reached, N = 17), followed by ESR1-H/ ESR2-L (17 m, N = 17), ESR1-L/ ESR2-L (13 m, N = 34) and ESR1-L/ ESR2-H (1 m, N = 1, p = .002). Conclusions: ESR1-H had a higher percentage of females, AD, and EGFR/ KRAS-mt v ESR1-L while ESR2-H had no sex difference, more SCC, and fewer EGFR/ KRAS-mt v ESR2-L. ESR1-H/ ESR2-H tumors had the highest MPAS and longest OS and there were SST differences with EGFR and KRAS G12C inhibition. ESR1& 2 may play key roles in activating the MAPK pathway and future trials could consider targeted therapy combined with ER inhibition based on ESR1&2 expression. [Table: see text]