Prognostic Impact of Initial Laboratory-based Scores for Hepatocellular Carcinoma Patients: A Retrospective Study

Hepatocellular carcinoma (HCC) is a heterogeneous tumor with variable prognosis. We aimed to evaluate different laboratory-based scores and indices to predict 1-year overall survival (OS).

A total of 188 patients with HCC were included. Albumin-to-alkaline phosphatase ratio (AAPR), aspartate aminotransferase-to-lymphocyte ratio (ALR), albumin-to-bilirubin (ALBI) score, neutrophil-to-lymphocyte∗platelet ratio (N/LPR), neutrophil-to-lymphocyte ratio (NLR), lymphocyte-to-monocyte ratio (LMR), and platelet-to-lymphocyte ratio (PLR) were evaluated in relation to 1-year OS using receiver operating characteristic (ROC) curves, Cox regression analysis, and a prediction nomogram.

Using ROC curve analysis, area under the curve (AUC) values of AAPR, ALR, NLR, LMR, and PLR were (0.704, 0.697, 0.743, 0.677, and 0.693, respectively) for the prediction of 1-year OS. ALBI score and N/LPR failed to significantly predict 1-year OS, and the NLR has the highest AUC among all the scores. In the subgroup receiving trans-arterial chemoembolization as the initial treatment, NLR and PLR were the scores that accurately predicted 1-year OS with AUCs of 0.701 and 0.685, respectively. Baseline elevated alanine aminotransferase, a performance status ≥1, larger lesions ≥5, multiple lesions ≥2, the presence of portal vein invasion, the presence of lymph node invasion, an NLR >2.27, and an LMR ≤3.40 are independent predictors of 1-year mortality in multivariate regression analysis. The C-index of the nomogram, including independent predictors in the multivariate regression analysis, predicting 1-year mortality was 0.808 (95% confidence interval: 0.771-0.845).

NLR and LMR are independent predictors of 1-year mortality, and a novel nomogram prediction model based on these scores show a higher performance. NLR is superior to other baseline non-invasive serum-based scores for the prediction of 1-year OS.