Abstract PS4-10-09: A cell-free DNA prognostic indicator for pathological response and recurrence in early triple negative breast cancer

Abstract Background: The addition of a checkpoint inhibitor to platinum containing neoadjuvant therapy (NACT) has significantly improved pathological complete response (pCR), progression-free (PFS) and overall survival (OS) in early triple negative breast cancer (eTNBC). However, clinical management of this breast cancer (BC) subtype remains challenging, as only between 40-50% of patients achieve a pCR and the rest witness rapid disease recurrence. Biomarkers to accurately predict pCR and recurrence, therefore, constitute unmet clinical needs. We herein deploy DNA-methylation patterns in tumor tissue as well as cell-free DNA (cfDNA) methylation before and after NACT for the identification of response and relapse biomarker in eTNBC. Patients and Methods: Blood and tumor tissue of 49 patients presenting with first diagnosis of eTNBC between Jan 2013 and Apr 2018 were evaluated for tumor DNA (n=27) and cfDNA methylation before (n=32) and after NACT (n=19). Paired plasma samples were available in 14 cases, paired tumor-plasma samples in 15 cases, respectively. Bead-based concentration and column isolation of cfDNA was achieved with the QiaAmp Mini Elute cfDNA Mini kit (Qiagen). Following fluorometric-based quantification, cfDNA methylation sequencing libraries were prepared with the Enzymatic Methylseq kit (New England Biolabs) followed by hybrid capture with an in-house panel. Libraries were sequenced on a Nextseq 2000 P4 flow cell and data analyzed with the Bismark suite. Missing values were imputed by K-Nearest Neighbors imputation (KNN). Differential methylation analyses was performed with the RnBeads suit and a penalized least absolute shrinkage and selection operator (LASSO) regression was used for biomarker identification. Maximal log-rank statistic was used for the determination of biomarker cut-off for all continuous variables. Results: Of all differentially methylated sites between pre- and post-therapy plasma samples, about 60% of sites showed a significant decrease in methylation levels. In paired analyses, tumor and plasma DNA methylation profiles mirrored each other for a substantial number of the selected markers. We identified sixteen differentially methylated sites associated with a pCR, most of which were located on chromosome 5. Single response marker analyses revealed markers with sensitivity and specificity higher than 0.8. Combination of at least two markers resulted in sensitivity and specificity over 0.9. The negative and positive predictive values were above 0.8 and 0.9 for single and combined markers, respectively. A prognostic index score combining all 16 response markers was significantly prognostic for OS (HR: 0.16, p value 1.25e-05). Furthermore, a 17 disease recurrence signature could be established with single markers attaining sensitivity and specificity of above 0.7 and 0.9, respectively, while combination of at least two markers reached 100% sensitivity and specificity. A recurrence risk signature combining all 17 recurrence markers was strongly prognostic for OS (HR: 0.16, p value 2.26e-05). Conclusion: Collectively, we have established cfDNA prognostic indicator signatures for pCR and recurrence in eTNBC. The validation of these signatures is ongoing in one of our independent validation cohorts and large multicentric cohorts could further foster clinical translation. Citation Format: S. Kasimir-Bauer, C. H. Zuendorf, A. Bittner, O. Hoffmann, R. Kimmig, J. Siveke, S. Lueong. A cell-free DNA prognostic indicator for pathological response and recurrence in early triple negative breast cancer abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS4-10-09.