This review covers blood-based biomarkers for hepatocellular carcinoma (HCC) and biliary tract cancers (BTC/CCA) across the full cancer care continuum — risk stratification, surveillance, early detection, prognostication, treatment selection, and minimal residual disease (MRD) monitoring.
Few biomarkers have reached routine clinical practice. AFP (alone or combined with ultrasound) and multi-marker panels (GALAD, GAAD, ASAP) show pooled early-stage HCC sensitivity of 70–74% and specificity of 83–87%. Methylated DNA panels (HelioLiver, Oncoguard® Liver) outperform ultrasound for early-stage HCC detection (sensitivity 48–73% vs. 22–29% for ultrasound). For CCA, ctDNA liquid biopsy detects actionable alterations in ~44% of advanced BTC cases; concordance with tissue for IDH1/BRAF is 87–100%, but complex alterations (fusions, CNVs) are less reliably detected. ctDNA positivity after resection predicts recurrence in HCC (HR 2.2–27) and eCCA (HR 1.8–7.7).
Most candidate biomarkers are validated only in small, retrospective, or case-control cohorts, with significant spectrum bias (exclusion of indeterminate lesions inflates performance). Technical variability across platforms, lack of standardised EV isolation and ctDNA quantification protocols, and biological confounding from underlying chronic liver disease all limit reproducibility and cross-study comparison.
For HCC surveillance, adding AFP to ultrasound improves sensitivity; methylated DNA panel tests (HelioLiver, Oncoguard® Liver) are promising adjuncts but await prospective RCT validation (TRACER trial ongoing). For advanced BTC, when tissue is unavailable or insufficient, plasma cfDNA with high-quality assays can identify actionable targets (FGFR2, IDH1, BRAF), and liquid biopsy is also useful for detecting acquired resistance mutations during FGFR inhibitor therapy.
Explore related topics
