Understanding hepatocellular carcinoma development using longitudinal laboratory measurements

Hepatocellular carcinoma (HCC) remains a leading cause of cancer mortality. Current surveillance strategies rely primarily on single-timepoint risk assessments and may not fully capture the evolving nature of chronic liver disease. This study aims to characterize longitudinal trajectories of routinely collected laboratory biomarkers and evaluate whether longitudinal biomarker updating provides prognostic information beyond baseline assessment for incident HCC. We conducted a retrospective cohort study of 2,790 patients with chronic liver disease from the OneFlorida+ Clinical Research Network (CRN) who had at least 6 months of follow-up after the first qualifying liver disease diagnosis. Follow-up continued until incident HCC or, for patients without incident HCC, the last recorded laboratory measurement. Longitudinal laboratory measurements were extracted from electronic health records and evaluated using time-varying Cox proportional hazards models. Prognostic performance was assessed using Harrell’s C-index and 36-month time-dependent area under the receiver operating characteristic curve (tAUC), with optimism correction based on bootstrap resamples. Baseline and longitudinal models were compared in an aligned cohort with complete baseline and follow-up biomarker data. During follow-up, 232 patients (8.3%) developed incident HCC. Patients who later developed HCC exhibited laboratory profiles consistent with more advanced liver disease at baseline. Among individual longitudinal biomarkers, platelet count and albumin demonstrated the strongest prognostic performance. In the aligned comparison cohort ( n = 537), 185 patients were evaluable at 36 months, including 32 who developed HCC. The joint longitudinal biomarker model showed modestly higher optimism-corrected prognostic performance than the corresponding baseline-only model (C-index: 0.746 vs. 0.723; difference, 0.023 [95% CI, -0.036 to 0.075]; 36-month tAUC: 0.711 vs. 0.685; difference, 0.026 [95% CI, -0.072 to 0.117]). The largest gains from longitudinal updating were observed for albumin and lymphocyte count. Our results suggest that longitudinal laboratory measurements provide incremental prognostic information beyond baseline assessment and may improve characterization of HCC risk evolution in patients with chronic liver disease. These findings support further evaluation of longitudinal biomarker-based approaches for HCC risk stratification and surveillance, but the magnitude and reproducibility of the observed gains require further evaluation.