Differential genomic profiling of the progesterone receptor (PR) negative and estrogen receptor (ER) low metastatic breast cancer subtypes through circulating tumor DNA: A TNBC-like disease subtype?

Abstract

1039 Background: ER positive/HER2 negative metastatic breast cancer (MBC) is typically regarded as a single entity, although significant clinical differences are often observed, especially in the ER low subgroup (ER<10% regardless of PR status). This study aims to investigate genomic differences among HER2 negative subtypes (i.e., ERlow, PRpos [ER >10%/PRpositve], PRneg [ER >10%/PRnegative], triple negative [TNBC]) through circulating tumor DNA (ctDNA) next-generation sequencing (NGS). Methods: The study analyzed a retrospective cohort of 879 patients (pts) with HER2 negative MBC with ctDNA testing using the Guardant360 NGS panel within a large multicenter academic consortium. ER, PR and HER2 status was defined based on the most updated biopsy. Associations across single nucleotide and copy number variations (SNVs and CNVs) and ER/PR status were tested by multinomial logistic regression in terms of Relative Risk Ratio (RRR), adjusting for significant clinical characteristics (i.e., lines of treatment, metastatic sites, histology). Oncogenic pathway was defined based on reference (1). Prognosis was analyzed through Cox regression for overall survival (OS) defined from time of ctDNA collection. Results: Among the 879 analyzed pts, PRpos was the most prevalent subtype (N:420, 48%), followed by PRneg (N:216, 24%), TNBC (N:209, 24%), and ERlow (N:34, 4%). Compared to PRpos, lobular MBC was less common in TNBC (RRR: 0.33, P=0.001), while lung involvement was more frequent in TNBC (RRR: 1.53, P=0.019), and liver involvement in PRneg (RRR: 1.44, P=0.035). In contrast, bone metastases were less common in TNBC and ERlow (RRR: 0.26, P<0.001 and RRR: 0.45, P=0.03, respectively). After multivariable analysis, ERlow and TNBC were associated with a higher incidence of TP53 SNVs (RRR: 5.42, P<0.001 and RRR: 3.08, P=0.007, respectively) and a lower incidence of ESR1 SNVs (RRR: 0.23, P=0.025 and not detected, respectively). TNBC exhibited significantly fewer PIK3CA SNVs (RRR: 0.41, P=0.005) and CCND1 CNVs (RRR: 0.20, P=0.009), along with a higher frequency of CCNE1 CNVs (RRR: 6.47, P<0.001). PRneg had a lower incidence of ESR1 SNVs (RRR: 0.52, P=0.013) and more frequent FGFR1 CNVs (RRR: 2.61, P=0.004). Similar results were observed across oncogenic pathways. Compared to PRpos, OS analysis indicated poorer outcomes across all subgroups (PRneg HR: 1.29 P = 0.035, TNBC HR: 1.96 P<0.001, ERlow HR: 2.32 P<0.001). Conclusions: This study highlights several clinical and molecular similarities between TNBC and ERlow MBC, while suggesting specific endocrine resistance mechanisms in PRneg that are less reliant on ESR1 and more exposed to alternative mechanisms such as FGFR1 amplification. These findings underscore the importance of more precise disease subtyping to inform future drug development trials. 1. Sanchez-Vega F et al, Cell. 2018.