Hepatocellular carcinoma (HCC) is one of the most prevalent malignancies worldwide and represents a major cause of cancer-related mortality (Chen et al., 2020). Its global burden continues to rise, particularly in regions with high prevalence of chronic hepatitis B and C infection, as well as among populations affected by alcohol-related liver disease and non-alcoholic fatty liver disease (Court et al., 2018). Despite advances in surgical resection, liver transplantation, and systemic therapies, the prognosis for patients with HCC remains poor, characterized by high recurrence rates and limited overall survival, especially in advanced stages (El-Mezayen et al., 2022).

Accurate risk stratification and early recurrence detection are critical for improving clinical outcomes in HCC (Jiang et al., 2024). Conventional surveillance strategies, including imaging and serum biomarkers such as alpha-fetoprotein, often lack the sensitivity and specificity required for timely and individualized management (El-Mezayen et al., 2022). These limitations have prompted the search for novel, dynamic biomarkers capable of capturing the molecular and biological complexity of HCC throughout the disease course (Jiang et al., 2024).

Circulating tumor DNA (ctDNA), comprising fragments of tumor-derived genetic material released into the bloodstream, has emerged as a promising non-invasive biomarker (Alemayehu et al., 2024a). In HCC, growing evidence suggests that the presence and quantitative characteristics of ctDNA may be associated with clinical outcomes, including disease-free survival and overall survival. However, existing studies vary widely in methodology, detection techniques, and clinical endpoints, leading to inconsistent findings and uncertainty regarding the clinical utility of ctDNA in routine practice (Alemayehu et al., 2024b).

In this study, we conducted a comprehensive meta-analysis to evaluate the prognostic and predictive value of ctDNA in patients with HCC. By systematically synthesizing the current evidence, we aim to clarify the role of ctDNA in forecasting clinical outcomes and to support its potential integration into the prognostic framework of HCC management.

This meta-analysis underscores the significant prognostic relevance of ctDNA in patients with HCC (Huang et al., 2022). The consistent association between ctDNA positivity and poor OS across studies suggests that ctDNA is not merely a reflection of tumor burden but a biologically meaningful marker of aggressive disease phenotypes (Zhang et al., 2021). It likely integrates multiple dimensions of tumor biology, including proliferative potential, vascular invasion, and metastatic propensity (Aalami et al., 2024). Notably, the prognostic value of ctDNA is not unique to HCC. A recent meta-analysis in pancreatic malignancies also reported that ctDNA-positive status was significantly associated with worse overall survival (HR = 2.00), reinforcing the role of ctDNA as a robust, pan-cancer prognostic biomarker across different solid tumors (Arayici et al., 2024).

The biological rationale supporting the prognostic role of ctDNA lies in its origin and dynamics. ctDNA is released into circulation during cellular apoptosis, necrosis, or active secretion by tumor cells, with its abundance influenced by factors such as tumor size, vascularity, immune clearance, and cellular turnover (Tan et al., 2021). In HCC, where tumors are often heterogeneous and vascularized, ctDNA may serve as a proxy for intratumoral heterogeneity and the presence of aggressive subclones (Cheng et al., 2019). The observed association with OS may therefore reflect an underlying biological shift toward systemic dissemination and resistance to host defenses or therapy (Cui et al., 2020).

The moderate heterogeneity observed among included studies is likely attributable to differences in detection platforms, patient populations, and definitions of ctDNA positivity (Chandrapalan et al., 2022). Some studies utilized PCR-based methods, which target specific mutations with high sensitivity, while others applied broader sequencing techniques that allow for comprehensive mutation profiling but with variable analytical depth (Li et al., 2022). Despite these methodological differences, sensitivity analyses confirmed the robustness of the association with OS, suggesting that the prognostic value of ctDNA remains consistent regardless of assay modality (Oussalah et al., 2018).

In evaluating the predictive performance for RFS, the pooled AUC indicated only modest discriminative accuracy (Liu et al., 2022). This may reflect, in part, methodological heterogeneity among the included studies, particularly regarding the timing and frequency of ctDNA sampling. Studies that performed ctDNA detection during the postoperative or surveillance phase generally reported higher AUCs, suggesting a more effective role for ctDNA in identifying minimal residual disease. Mechanistically, this aligns with the concept that residual malignant clones shed fragmented DNA into the circulation before they are radiologically detectable. ctDNA may thus serve as a leading indicator of molecular relapse, offering a window for earlier intervention (Lumkul et al., 2024). In contrast, pre-treatment ctDNA levels may reflect bulk tumor burden rather than residual disease potential, limiting their utility in forecasting recurrence. These observations underscore the importance of temporal context in ctDNA interpretation and point to the need for standardized protocols regarding assay timing and clinical thresholds.

From a translational perspective, ctDNA holds several theoretical and practical advantages over conventional surveillance tools in hepatocellular carcinoma. Biologically, ctDNA reflects tumor-derived genetic and epigenetic alterations, such as mutations in TP53, TERT promoter, or CTNNB1, as well as methylation changes and fragmentomic signatures. These features allow for a more nuanced understanding of tumor behavior, clonal evolution, and response dynamics (Lumkul et al., 2024). Unlike static biomarkers such as AFP or imaging modalities that rely on structural changes, ctDNA enables real-time monitoring of tumor kinetics and treatment efficacy, potentially capturing molecular resistance before clinical progression (Jin et al., 2019). Furthermore, in patients with impaired hepatic function or elevated bleeding risk, where invasive biopsy is contraindicated, ctDNA offers a safe and repeatable alternative for tumor genotyping and longitudinal assessment. The ability to track tumor-specific molecular profiles through serial blood sampling may ultimately support a shift toward more personalized, adaptive management strategies in HCC (Lumkul et al., 2024). However, the implementation of ctDNA-based monitoring in routine practice requires rigorous validation of its sensitivity, specificity, and cost-effectiveness in prospective studies.

Despite these promising implications, this meta-analysis has several limitations. First, the number of included studies was limited, and their predominantly retrospective design may introduce bias and confounding. Additionally, substantial variability in ctDNA detection methods, thresholds, and sampling timepoints, coupled with the frequent lack of key methodological details such as the blood volume used for cfDNA isolation, may compromise the comparability and sensitivity of ctDNA assessments across studies. Moreover, most studies did not adjust for clinical variables such as tumor stage or treatment type, limiting interpretation of independent prognostic value. Furthermore, the assessment of ctDNA’s predictive accuracy for recurrence was based on only two studies with significant heterogeneity, which precludes strong conclusions in this regard. Finally, the generalizability of our findings may be constrained by the geographical concentration of the included patient cohorts, which were primarily from Asian and European populations. Therefore, future research is needed to validate the prognostic value of ctDNA in broader geographical and ethnic cohorts.

In conclusion, this study supports ctDNA as a promising biomarker for prognostication and recurrence prediction in HCC. Its detection correlates with worse survival outcomes and shows potential for early identification of molecular relapse. However, broader validation in prospective, well-annotated cohorts is necessary. Standardized workflows, defined reporting frameworks, and integration with imaging and clinical parameters will be key to realizing the clinical utility of ctDNA in the management of HCC.