| Background and ObjectionAccording to the latest published Global Cancer Statistics,primary hepatocellular carcinoma(HCC),as the fifth most deadly cancer in men,remains keeping high incidence in the world.The geographic and biologic heterogeneity of HCC is complicated.There are several treatment strategies for HCC,including partial hepatectomy,orthotopic liver transplantation,tumor ablation,and hepatic intra-arterial therapies.Surgical resection,the potentially curative option,is widely used,especially for small HCC.The survival benefits of surgical resection are better than those of transarterial chemoembolization(TACE)in selected patients with intermediate-to advanced-stage HCC.To stratify the expected survival outcomes of patients with HCC,numerous staging systems have been established.Barcelona Clinic Liver Cancer(BCLC)stage is the most widely used staging system,and is recognized as the best guide for clinical management The 2017 staging criteria of the American Joint Commission on Cancer place greater emphasis on major vascular invasion as a potential predictive factor for poor prognosis.However,most conventional staging systems are not specifically targeted at patients with HCC after surgery.Thus,they are limited in their ability to predict postoperative survival.Nomograms are graphic depictions of predictive statistical models,and have been applied in cancers such as nasopharyngeal carcinoma,breast cancer and lung cancer.Nomograms have attracted the attention of physicians because of their advantages over traditional staging systems used to better and more accurately predict the outcomes of patients with cancer.Several nomograms have been established to predict survival or postoperative recurrence in patients with HCC.The predictive factors in these nomograms include characteristics of patients and their tumors.However,liver function and clinicopathologic factors are excluded.The survival of patients with HCC is affected by many factors.Hence,in this study,we developed a practical clinical tool that considers patient and tumor characteristics,liver function,and clinicopathologic factors.MethodsOur nomogram was based on a retrospective study of 426 patients who underwent primary complete resection of HCC at Nanfang Hospital,Southern Medical University,Guangzhou,China,between January 2007 and July 2014.The inclusion criteria were as follows:complete removal of macroscopic carcinomas;definitive histopathologic diagnosis of HCC in surgical samples;no preoperative anticancer treatments for HCC;and age ?18 years.The exclusion criteria were as follows:recurrent HCC;tumors with mixed types on histopathologic analysis or of uncertain origin;perioperative mortality;and incomplete follow-up data.Of them,284 were allocated to model derivation and 142 were allocated to model validation.The nomogram used independent predictors of overall survival(OS).Its predictive accuracy and discriminative ability were determined by concordance Index(C-index)and calibration curves,and compared with those of BCLC stage.The results were validated using bootstrap resampling.All statistical tests were two-sided.ResultsThe independent predictors of survival derived from univariate and multivariate analyses of the primary cohort were age,tumor size,tumor differentiation,serum Albumin level,and tumor embolus,and were assembled into the nomogram.Calibration curves for survival probability showed that nomogram-based predictions demonstrated good agreement with actual observations.The C-index of the nomogram for predicting survival was 0.738,which was statistically higher than that of BCLC stage(p = 0.001).This result was confirmed in the validation cohort(p =0.001).The areas under the curve of the nomogram for the prediction of 1-,2-,and 3-year OS were significantly better than those of BCLC stage.ConclusionOur predictive nomogram enables accurate prediction of the postoperative survival of patients with HCC and warrants validation at other hepatobiliary centers. |
PDF Full Text Request