Abstract P5-01-13: Longitudinal changes in whole-genome (WG) cell-free DNA (cfDNA) and methylation as a blood-based biomarker to identify early disease progression in advanced breast cancer

Abstract

Abstract Background: Liquid biopsies have potential clinical utility as dynamic biomarkers for treatment response in advanced breast cancer. We evaluated a plasma-only assay to track serial changes in WG cfDNA to identify disease progression prior to routine imaging. Methods: We prospectively enrolled and serially collected blood from 25 patients with advanced breast cancer. Blood was drawn and collected in Streck tubes prior to start of a new treatment, after the first cycle (median 22 days), and/or second cycle (median 52 days). 4 mL of plasma was separated from peripheral blood, after which cfDNA was isolated from plasma and used to prepare libraries (11 with bisulfite conversion) for WG sequencing (median 20x depth). Based on a patient-specific profile of WG features, including copy-number alterations and cfDNA fragment length, the fraction of tumor-derived cfDNA (ctDNA) was quantified over the initial course of treatment. For a subset of the cohort (11 of 25 patients), we also quantified changes in genome-wide methylation levels from baseline to subsequent timepoints to classify patients as progressors or non-progressors. Imaging was performed per standard practice with treatment response determined by an independent radiologist according to RECIST 1.1 guidelines. Results: Median age of patients in the cohort was 65 (range 30-83) and included all subtypes—HR+ (n=14), HR+ & HER2+ (n=6), and triple negative (n=5). 52% of patients were on their third line of therapy (range 1-5). On-study therapies were chemotherapy alone (8), targeted + hormone therapy (5), hormone therapy alone (4), targeted + chemotherapy (5), targeted therapy alone (1), or immune checkpoint + HDAC inhibitors (2). Patients with predicted progression by cfDNA (n=5), indicated by an increase in tumor fraction at either post-treatment blood collection, had worse progression-free survival (median 67 days) compared to patients who did not show an increase (n=20; median 207 days) (hazard ratio 7.9, [95% CI 2.2-28.5], log-rank p=3 × 10-4). For the patients who were predicted to progress, the ctDNA assay preceded clinical evaluation by a median of 53 days. All patients with predicted progression were later confirmed to progress at the first follow-up evaluation (5/5, 100% positive predictive value). For the remaining patients, 15 of 20 did not progress (75% negative predictive value). Therefore, sensitivity for the assay was 50% and specificity was 100%. Comparing molecular predictions for 11 of 25 patients based on genomic versus methylation features, 8 non-progressors were classified correctly by both types of features. For the 3 progressors, 2 were predicted correctly based solely on methylation, increasing the sensitivity to detect progression early in the treatment course. Conclusions: Analyzing ctDNA early in the course of a new therapy holds promise to identify patients with early disease progression across multiple types of treatment. The assay accurately identified patients with no durable clinical benefit earlier, indicating potential clinical utility to change therapy and to limit unnecessary side effects and costs associated with ineffective treatments, if validated in a prospective clinical trial.​ Finally, integrating methylation-based changes with information about genomic alterations may increase performance of ctDNA-based response monitoring. Citation Format: Andrew A Davis, David Chan, Michael S Oh, Robert W Lentz, Neil Peterman, Alex Robertson, Rohith Srivas, Abhik Shah, Nicole Lambert, Ayse Tezcan, Haluk Tezcan, Young Kwang Chae. Longitudinal changes in whole-genome (WG) cell-free DNA (cfDNA) and methylation as a blood-based biomarker to identify early disease progression in advanced breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P5-01-13.