Surgery is a potential cure for hepatocellular carcinoma (HCC), but its postoperative recurrence rate is high, its prognosis is poor, and reliable predictive indicators are lacking. This study was conducted to develop a simple, practical, and effective predictive model.
Preoperative clinical and postoperative pathological data on patients with HCC undergoing partial hepatectomies at the Third Affiliated Hospital of Soochow University from January 2010 to December 2015 were retrospectively analyzed, and a nomogram was constructed. The model performance was evaluated using C-indexes, receiver operating characteristic curves, and calibration curves. The results were verified from validation cohort data collected at the same center from January 2016 to January 2017 and compared with the traditional staging systems.
Three hundred three patients were enrolled in this study: 238 in the training cohort and 65 in the validation cohort. From the univariate and multivariate Cox regression analyses in the training cohort, six independent risk factors, i.e., age, alpha-fetoprotein (AFP), tumor size, satellite nodules, systemic immune inflammation index (SII), and prognostic nutritional index (PNI), were filtered and included in the nomogram. The C-index was 0.701 [95% confidence interval (CI): 0.654–0.748] in the training cohort and 0.705 (95% CI: 0.619–0.791) in the validation cohort. The areas under the curve for the 1- and 3-year recurrence-free survival were 0.706 and 0.716 in the training cohort and 0.686 and 0.743 in the validation cohort, respectively. The calibration curves showed good agreement. Compared with traditional American Joint Committee on Cancer 8th edition (AJCC8th) and Barcelona Clinic Liver Cancer (BCLC) staging systems, our nomogram showed better predictive ability.
Our nomogram is simple, practical, and reliable. According to our nomogram, predicting the risk of recurrence and stratifying HCC patient management will yield the greatest survival benefit for patients.