Ultra Deep Sequencing of Circulating Cell-Free DNA as a Potential Tool for Hepatocellular Carcinoma Management

In this unicentric prospective study, we analyzed the most prevalent mutations in HCC (TERT promoter, TP53, CTNNB1, AXIN1 and ARID1A) in plasma cfDNA by next-generation sequencing, aiming to elucidate their value as prognostic noninvasive biomarkers. Total cfDNA (cut-off value 2 ng/µL), number of mutated genes and number of detectable mutations on cfDNA were significantly related to mortality. Number of mutated genes and number of detected mutations in cfDNA and the ratio between number of mutations and total amount of cfDNA were also significantly associated with recurrence. Detection of more than four mutations in cfDNA correlated with a higher risk of death. Dynamic changes in cfDNA mutations were detected prior to radiological detection of HCC recurrence. We believe that these results support the proof of principle and launching of validation studies to confirm that total cfDNA and detection of prevalent HCC mutations could have prognostic implications in early-stage HCC patients. Background: Cell-free DNA (cfDNA) concentrations have been described to be inversely correlated with prognosis in cancer. Mutations in HCC-associated driver genes in cfDNA have been reported, but their relation with patient's outcome has not been described. Our aim was to elucidate whether mutations found in cfDNA could be representative from those present in HCC tissue, providing the rationale to use the cfDNA to monitor HCC. Methods: Tumoral tissue, paired nontumor adjacent tissue and blood samples were collected from 30 HCC patients undergoing curative therapies. Deep sequencing targeting HCC driver genes was performed. Results: Patients with more than 2 ng/µL of cfDNA at diagnosis had higher mortality (mean OS 24.6 vs. 31.87 months, p = 0.01) (AUC = 0.782). Subjects who died during follow-up, had a significantly higher number of mutated genes (p = 0.015) and number of mutations (p = 0.015) on cfDNA. Number of mutated genes (p = 0.001), detected mutations (p = 0.001) in cfDNA and ratio (number of mutations/cfDNA) (p = 0.003) were significantly associated with recurrence. However, patients with a ratio (number of mutations/cfDNA) above 6 (long-rank p = 0.0003) presented a higher risk of recurrence than those with a ratio under 6. Detection of more than four mutations in cfDNA correlated with higher risk of death (long-rank p = 0.042). Conclusions: In summary, cfDNA and detection of prevalent HCC mutations could have prognostic implications in early-stage HCC patients