Circulating tumor DNA dynamics in patients with liver-limited metastatic colorectal cancer resected after first-line systemic treatment

PURPOSE: Resection of liver metastases improves survival in metastatic colorectal cancer (mCRC) patients with liver-limited disease (LLD), but relapse remains common. Post-resection circulating tumor DNA (ctDNA) is a valid prognostic biomarker, however no data are available regarding the prognostic effect of pre-surgery ctDNA after upfront chemotherapy. EXPERIMENTAL DESIGN: Patients (N=116) with initially unresectable mCRC with LLD who underwent resection after upfront chemotherapy were included. Patients were sequenced with a tumor-naive ctDNA-based minimal residual disease (MRD) assay at baseline (pre-chemotherapy), pre-surgery, and post-surgery. RESULTS: Pre-surgery ctDNA status was not significantly associated with RFS. However, exploratory analyses demonstrated that a ≥50% reduction in VAF from baseline to pre-surgery was independently associated with improved RFS (median RFS: 21.0 vs. 9.8 months; HR: 2.19; p=0.014), aligned with multivariate analysis (p = 0.022). Post-surgery ctDNA positivity was strongly predictive of recurrence (HR: 6.66; p<0.001), with 100% specificity and 56.4% sensitivity. ctDNA dynamics from pre- to post-surgery further stratified recurrence risk, regardless of adjuvant chemotherapy status. ctDNA dynamics from pre- to post-surgery further stratified recurrence risk. Quantitative ctDNA measures (predicted tumor fragments per million) also correlated with RFS. CONCLUSIONS: Use of a tumor-naive MRD assay increased the accuracy of detecting ctDNA after chemotherapy treatment, when a significant drop of ctDNA amount is expected compared to tests routinely used in metastatic disease. Pre-surgery ctDNA dynamics were prognostic with respect to relapse-free survival and may help stratify patients' prognosis prior resection, potentially informing personalized treatment decisions.