LAUSR

Introduction Sarcomatoid hepatocellular carcinoma (SHC) is a rare subtype of liver cancer with extremely poor prognosis. This study aimed to identify the prognostic factors and develop an exclusive and efficient nomogram to predict cancer-specific survival (CSS) for SHC. Methods The data on patients diagnosed with SHC from January 1973 to December 2015 were retrieved from the Surveillance, Epidemiology, and End Results (SEER) database, and these patients were included as the training cohort. Least absolute shrinkage and selection operator (LASSO) and Cox proportional hazards regression analyses were used to identify the prognostic risk factors and construct a nomogram. The predictive accuracy and discriminative ability of the nomogram were determined using concordance index (C-index), calibration curve, and receiver operating characteristic (ROC) curve. Decision curve analysis (DCA) was used to compare the clinical benefits of the prognostic evaluation model (PEM) with that of the American Joint Committee on Cancer (AJCC) staging system. The results were validated with an external validation cohort. Results In total, 116 patients with SHC were included in the training cohort. Multivariate Cox analysis revealed M stage (distant metastasis), primary tumor surgery, and chemotherapy to be associated with CSS, and along with tumor size, an integrated PEM was constructed. A calibration curve for the probability of survival showed good agreement between the nomogram and actual observation. The C-index value of the nomogram for predicting CSS and AJCC was 0.853 and 0.649, respectively. In the validation cohort, the C-index value of the PEM discrimination was better than that of the Barcelona Clinic Liver Cancer (BCLC) staging system, CLIP score, and Okuda staging system, and no statistical difference was observed with eighth edition of the AJCC staging system and Izumi staging system. Conclusion The proposed four-factor nomogram of PEM could accurately predict the prognosis of SHC and could be used in clinical practice.