Jiajing Zhao, Liping Liu
We read with considerable interest the study by Lin and colleagues in Alimentary Pharmacology & Therapeutics, ‘Liver Steatosis-Related Polygenic Risk Score Predicts Hepatocellular Carcinoma Risk After Hepatitis C Virus Eradication’, which examined whether a five-variant steatosis-related polygenic risk score (PRS-5) can stratify hepatocellular carcinoma (HCC) risk after sustained virologic response (SVR). The study addresses a clinically relevant question, particularly as current practice increasingly favours risk-adapted surveillance rather than uniform follow-up for all patients [1-3]. First, the proposed high-risk threshold (PRS-5 ≥ 0.66) raises questions regarding its practical applicability. Although selecting the top 10% may improve specificity, it may reduce sensitivity for identifying patients who will eventually develop HCC. In daily practice, this may create uncertainty: patients below the threshold could still have a clinically meaningful absolute risk over time, yet they would not meet criteria for intensified surveillance. Such a dichotomous approach may lead to variation in clinical decision-making and possible under-recognition of individuals at intermediate risk. It may therefore be more informative to define risk categories based on absolute risk estimates, such as low, intermediate, and high risk, aligned with cost-effectiveness considerations and guideline-relevant incidence thresholds, and to report net reclassification across predefined strata [4-6]. Second, although PRS-5 remained predictive in noncirrhotic subgroups, the clinical implications for surveillance are not fully delineated. Current major guidelines primarily recommend surveillance in patients with cirrhosis and selected high-risk groups, largely based on cost-effectiveness and the balance between benefit and harm [2, 3]. In the post-SVR setting, where risk decreases but does not disappear, clinicians require clearer guidance regarding which noncirrhotic patients should undergo surveillance, when surveillance should begin, which modality should be used, and at what interval. An explicit framework integrating PRS-5 with fibrosis severity, such as FIB-4 or elastography, together with age, and supported by sensitivity analyses against established incidence benchmarks, would enhance the clinical interpretability of these findings [2, 4, 5]. Third, the incremental value of PRS-5 beyond established post-SVR risk models requires further clarification. In routine practice, clinicians commonly rely on parsimonious models incorporating age, platelet count, albumin, and fibrosis markers, and several such tools have undergone external validation. Without direct comparisons using measures of discrimination, calibration, decision-curve analysis, or net benefit, it remains difficult to determine whether PRS-5 meaningfully improves risk stratification over existing approaches. A pragmatic next step would be to integrate PRS-5 into a validated clinical model and formally evaluate its incremental contribution, ideally using a simplified targeted genotyping strategy to address feasibility and equitable access [5, 7-10]. In summary, Lin et al. provide important evidence that steatosis-related genetic susceptibility may refine HCC risk assessment after HCV eradication. Greater clarity regarding clinically actionable thresholds, explicit surveillance strategies for noncirrhotic patients, and demonstration of incremental value over established models would further strengthen the translational relevance of these findings and facilitate consideration in future guideline development. J.Z. writing – review and editing, writing – original draft; L.L. writing – review and editing. The authors have nothing to report. Declaration of Generative AI and AI-Assisted Technologies in the Writing Process: During the preparation of this work, the authors used ChatGPT to check grammar and spelling. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. The authors have nothing to report. The authors declare no conflicts of interest. This article is linked to Lin et al. papers. To view this article, visit https://doi.org/10.1111/apt.70566. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.