A nomogram to predict prognosis after surgery for young patients with hepatocellular carcinoma

BACKGROUND: Only few studies have been evaluated the clinical characteristics and prognosis of hepatocellular carcinoma (HCC) in young patients. The purpose of this study is to identify prognostic factors and develop an efficient and practical nomogram to predict cancer-specific survival (CSS) in young patients with HCC. METHODS: Four hundred and forty-one young patients with HCC who had undergone surgery from 2004–2015 were selected from the Surveillance, Epidemiology, and End Results (SEER) database. The competing risk model, Lasso and Cox regression were used to screen prognostic factors for CSS, and a prognostic nomogram was established using these factors. Thirty-nine young patients with HCC from the National Cancer Center, Cancer Hospital, Chinese Academy of Medical Science were used to validate our model. To further evaluate the predictive performance of our model, the concordance index was calculated and the calibration curves were drawn. The clinical usefulness was evaluated by decision curve analysis (DCA). Finally, all patients were grouped by our nomogram. The survival of different risk groups was analyzed using the Kaplan-Meier method, and the differences among survival curves were compared by the log-rank test. RESULTS: The median survival times of the SEER training group and the external National Cancer Center validation group were 41 and 52 months, respectively. Histological grade, tumor size, Alpha-fetoprotein (AFP), T stage, and M stage were selected as independent factors for CSS, and a prognostic nomogram was established. The concordance indices of the training and external validation groups were 0.76 (95% CI, 0.72 to 0.80) and 0.92 (SE=0.085), respectively. The calibration plots showed good agreement. DCA revealed that our nomogram resulted in a better clinical net benefit than the AJCC 7th edition and Barcelona Clinic Liver Cancer staging systems. Patients were divided into two risk groups according to the cut-off value of 125 of the total points from our nomogram. Kaplan-Meier plots for CSS were performed using the log-rank test, the P-value of which was <0.001. CONCLUSIONS: The practical nomogram resulted in a more-accurate prognostic prediction for young HCC patients after curative liver resection.