Clinical and CT imaging characteristics of early recurrence in patients with hepatocellular carcinoma after curative resection

Purpose To investigate the clinical and CT imaging factors of early recurrence in patients with hepatocellular carcinoma (HCC) after surgical resection and their predictive value. Methods We retrospectively analyzed 87 HCC patients who underwent curative resection. The associations of clinical and CT imaging factors with early recurrence were investigated by univariate and multivariate analysis. A binary logistic regression model was constructed based on the results of a multivariate regression analysis. The area under the ROC curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy rate, together with their 95% confidence interval (CI), were calculated. Results Postoperative early HCC recurrence was associated with alpha-fetoprotein, the platelet-to-lymphocyte ratio (PLR), aspartate aminotransferase, alanine aminotransferase, tumor diameter, unsmooth margins and satellite nodules ( P  < 0.05). The binary logistic regression model included 3 predictors from the regression analysis: PLR, tumor margin, and satellite nodule. The AUC, sensitivity, specificity, PPV, NPV, and accuracy rate were 0.854 (95% CI 0.776 ~ 0.933), 85.29%, 92.11%, 78.6%, 80% and 81.8%, respectively. Conclusion PLR, tumor margins, and satellite nodules were independent risk factors for postoperative HCC early recurrence. The binary logistic regression model based on these indicators showed good performance in predicting postoperative HCC early recurrence.