Assessment of Circulating Tumor DNA for Early Detection of Hepatocellular Carcinoma in Alpha-Fetoprotein-Negative Patients With Cirrhotic Nodules.

Aims: Early detection is essential for improving hepatocellular carcinoma's (HCC) prognosis. However, the diagnostic performance of the current interventions remains unsatisfactory, particularly in patients with cirrhotic nodules. We aimed to develop a predictive model for the early detection of HCC using circulating tumor DNA (ctDNA) profiles based on 86 cancer-related genes. Methods: This model was initially evaluated on a discovery cohort of 25 patients with HCC and 9 patients with chronic hepatitis and then validated in a prospective control cohort of 15 alpha-fetoprotein-negative patients with cirrhotic nodules. Results: In the discovery cohort, our ctDNA-based model achieved an area under the receiver operating characteristic curve of 0.8675, with a sensitivity of 88.0% and a specificity of 88.9%. CtDNA mutations in FGFR3, NF1, ALK, ERBB2, and NRAS were significantly associated with HCC prognosis. In the validation cohort, the ctDNA assay was positive in 8 of the 15 patients. Over a five-year follow-up period, 50% of ctDNA-positive patients were diagnosed with early-stage HCC, with a mean lead time of 26.6 months compared to standard clinical assessments. Conclusion: By enabling diagnosis months earlier than conventional methods, ctDNA-based testing may enhance HCC surveillance and facilitate a more personalized approach to cancer monitoring.