Personalized MRD Assessment in Perisurgical ctDNA for Prognostic Prediction in Hepatocellular Carcinoma

Abstract Purpose: Detecting residual disease is a critical clinical requirement in the perisurgical management of patients with resectable hepatocellular carcinoma (HCC). Previous studies focused on specific genomic regions exhibiting limited sensitivity and failed to meet the minimal residual disease (MRD) testing threshold. We introduce a next-generation sequencing–based assay, informed by baseline samples, facilitating MRD detection in hepatectomized patients with HCC and offering prognostic predictions. Experimental Design: This study involved 88 patients with HCC who underwent surgical resections from January 2016 to May 2016 in Zhongshan Hospital, Fudan University. Tumor and normal tissue samples were collected during surgery, whereas plasma samples were obtained both before surgery and up to 7 days after surgery. Using a next-generation sequencing–based personalized ctDNA assay, we analyzed the MRD in both presurgical and postsurgical blood samples and its correlation with prognosis. Results: With a median follow-up period of 80.7 months, our findings demonstrated significant correlations between presurgical ctDNA tumor fractions, postsurgical plasma MRD status, and both recurrence-free survival and overall survival. Postsurgical MRD status emerged as the most significant risk factor for cancer recurrence (HR = 2.162; 95% confidence interval, 1.09–4.30; P = 0.027) compared with other clinical characteristics in multivariate Cox regression analysis. Notably, MRD status showed potential as a prognostic indicator among clinically low-recurrent-risk patients, such as those with Barcelona Clinic Liver Cancer stages 0 to A or China Liver Cancer Staging stages I to II. Conclusions: Evaluating personalized MRD provided crucial prognostic insights into recurrence-free survival and overall survival. It efficiently identified patients at high risk of recurrence, even among those initially perceived as low-risk cases. See related commentary by Pinato et al., p. 955