Clinicopathologic features, treatment, survival, and prognostic factors of combined hepatocellular and cholangiocarcinoma: A nomogram development based on SEER database and validation in multicenter study

Purpose The aim of the study was to comprehensively understand the combined hepatocellular and cholangiocarcinoma (CHC) and develop a nomogram for prognostic prediction of CHC. Methods Data were collected from the Surveillance, Epidemiology and End Results (SEER) database (year 2004–2014). Propensity-score matching (PSM) was used to match the demographic characteristic of the CHC versus hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC). A nomogram model was established to predict the prognosis in terms of cancer specific survival (CSS). The established nomogram was externally validated by a multicenter cohort. Results A total of 71,756 patients enrolled in our study including 62,877 HCC patients, 566 CHC patients, and 8303 ICC patients. The CHC, HCC, and ICC are not exactly similar in clinical characteristic. After PSM, the CSS of CHC was better than HCC but comparable to ICC. Tumor size, M stage, surgery, chemotherapy, and surgery were independently prognostic factors of CHC and were included in the establishment of novel nomogram. The c-index of the novel nomogram in SEER training set and multicenter validation was 0.779 and 0.780, respectively, which indicated that the model was with better discrimination power. In addition, decision curve analyses proved the favorable potential clinical effect of the predictive model. Lastly, a risk classification based on nomogram also verified the reliability of the model. Conclusion CHC had better survival than HCC but was comparable to ICC. The nomogram was established based on tumor size, M stage, chemotherapy, surgery, and radiotherapy and well validated by external multicenter cohort.