Preoperative predictors of early recurrence after resection for hepatocellular carcinoma

Abstract

BackgroundTo assess preoperative image tumor characteristics and alpha-fetoprotein (AFP) levels to predict early recurrence after liver resection (LR) for hepatocellular carcinoma (HCC). MethodsThis retrospective study's enrolled patients underwent LR for newly diagnosed HCC between 2011 and 2018. Multivariate logistic regression analyses using the Akaike information criterion were adopted to construct a nomogram to predict early recurrence (i.e. recurrence within 1 year). The performance of this nomogram was evaluated using calibration plots with bootstrapping. ResultsEarly recurrence was identified in 99 patients (11.2%). Four predictive factors, namely an AFP level of >400 ng/mL; image-diagnosed tumor characteristics, including a tumor size of > 5 cm; vascular invasion; and multiple tumors were adopted in the final model of the early recurrence nomogram, with a concordance index of 0.67. The calibration plots showed good agreement between the nomogram predictions and the actual observations of early recurrence. ConclusionWe have developed a simple nomogram with preoperative image tumor characteristics and AFP levels to predict the early recurrence of HCC after LR.