Estrogen receptor (ER) expression on circulating tumor cells (CTCs) and cell free DNA (cfDNA) mutational landscape in the PACE randomized phase II study.

Abstract

1054 Background: The PACE (NCT03147287) randomized phase II trial investigates CDK4/6 inhibition beyond progression in combination with endocrine treatment, with or without PD-L1 inhibition, in hormone receptor-positive (HR+)/HER2- metastatic breast cancer (MBC) (Mayer et al 2024). We previously reported that cfDNA alterations and CTC number correlated with survival and treatment response (Jeselsohn et al 2024; Gerratana et al ASCO 2023). Here we investigated ER expression on CTCs in relation to the cfDNA mutational landscape. Methods: Samples were collected at baseline. CTC enumeration and ER protein expression on CTCs (by immunofluorescence) was evaluated with the CellSearch and the ACCEPT software. Samples were classified as CTC high or CTC low (cutoff ≥5 CTC/sample). CTC high samples were defined ER+ if >15% CTCs/sample expressed ER to ensure good inter-group stratification. Concurrently, cfDNA was analyzed with the Guardant360 assay. Only pathogenic single nucleotide (snv) and copy number variations (cnv) with ≥3% prevalence were included and categorized into oncogenic pathways (Sanchez-Vega et al 2018). Differences in distribution across CTC groups were tested through Chi-squared and Fisher's test. Results: From 220 enrolled patients, 167 were evaluable for ER on CTCs. Of these, 91 were CTC low , 30 were CTC high /ER-, and 46 CTC high /ER+. ESR1 mutations were more common in CTC high /ER+ samples, while CTC high /ER- samples had higher incidence of alterations in SMAD4 , PIK3CA , BRAF and CDK4 compared to the other 2 groups (Table 1). CTC high /ER- had also higher mutant allele frequency compared to CTC high /ER+ and CTC low (MAF > 3% in 73% vs 54% and 31%, respectively, p < 0.001). CTC low samples had overall lower cfDNA alteration incidence. Similarly, alterations in the ER pathway were more frequent in samples with CTC high /ER+, whereas alterations in PI3K, cell cycle and P53 pathways were more common in CTC high /ER- samples. Alterations in the RTK/RAS/RAF pathway were more common in CTC high samples (23% and 28% for ER- and ER+ vs 9.9% for CTC low , p = 0.017). Similar results were observed with a 10% threshold. Conclusions: Distinct cfDNA alterations were identified based on ER expression in CTCs in HR+/HER2- MBC. Integrating CTC enumeration and cfDNA profiling may help elucidate resistance mechanisms, identify actionable targets, and predict benefit from continued CDK4/6 inhibition beyond progression. Incidence of cfDNA alterations across the 3 CTC-based groups. cfDNA alterations CTC low 1 CTC high /ER- 1 CTC high /ER+ 1 P value ESR1 2 36 (40) 15 (50) 31 (67) 0.009 SMAD4 2 0 (0) 3 (10) 2 (4.3) 0.009 PIK3CA 3 2 (2) 4 (13) 0 (0) 0.012 BRAF 3 1 (1) 1 (3.3) 5 (11) 0.022 CDK4 3 1 (1) 3 (10) 2 (4.3) 0.035 ER pathway 2 40 (44) 15 (50) 32 (70) 0.017 PI3K pathway 3 2 (2.2) 4 (13) 0 (0) 0.012 Cell cycle pathway 3 8 (8.8) 9 (30) 8 (17) 0.019 P53 pathway 2 27 (30) 16 (53) 13 (28) 0.040 1 n (%); 2 snv; 3 cnv.