Multi-Omics Biomarker Signatures for Precision Diagnosis and Prognosis in Primary Liver Cancer: A Literature Review.
TL;DR
It is argued that clinically useful signatures should be question-specific, stage-aware, and specimen-aware rather than universal multi-analyte panels and the need for prospective validation across clinically diverse populations is argued.
Abstract
Primary liver cancer (PLC) is a biologically heterogeneous group of malignancies dominated by hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (iCCA), and a smaller subset of combined hepatocellular-cholangiocarcinoma (cHCC-CCA), and its clinical burden remains high because current diagnostic and prognostic tools do not adequately capture molecular diversity. Conventional imaging, serum markers, and histopathological assessment remain insufficient for precise early diagnosis, subtype-resolved classification, and outcome stratification, while tissue and liquid biopsy approaches have expanded the range of analytes available for clinical assessment. Recent studies have identified candidate biomarker signatures across genomic, epigenomic, transcriptomic, proteomic, metabolomic, and circulating layers, suggesting that integrated multi-omics profiling may better represent tumor lineage, clonal evolution, immune context, and therapeutic vulnerability than isolated molecular readouts. However, these layers are not equally mature for clinical use: genomic testing is closest to routine therapeutic application in iCCA, plasma methylation assays are advancing for HCC surveillance augmentation, and many proteomic or metabolomic panels remain validation-stage tools. Their clinical value remains constrained by sampling bias, biospecimen-dependent signal loss, assay standardization, cost, and the need for prospective validation across clinically diverse populations. This narrative review critically synthesizes current evidence on multi-omics biomarker signatures for precision diagnosis and prognosis in primary liver cancer and argues that clinically useful signatures should be question-specific, stage-aware, and specimen-aware rather than universal multi-analyte panels.