Nomogram Prediction of Overall and Cancer-specific Survival in Patients with Hepatocellular Carcinoma and Pulmonary Metastasis:A Retrospective Cohort Study / 华中科技大学学报(医学版)

ABSTRACT

Objective To study the influencing factors of overall survival(OS)and cancer-specific survival(CSS)in patients with hepatocellular carcinoma and pulmonary metastasis and establish nomograms to predict survival.Methods The study pop-ulation consisted of 2242 cases with a first primary hepatocellular carcinoma who presented with pulmonary metastasis at the time of diagnosis in the Surveillance,Epidemiology,and End Results(SEER)database of the National Cancer Institute from 2010 to 2016.The influencing factors of OS and CSS were evaluated by using multivariable Cox proportional hazards regression mod-els.Nomograms predicting 1-year OS and CSS were constructed.Data analysis and construction of nomograms were performed with Cox proportional hazards regression models,the Kaplan-Meier curves(log-rank test)and C-index.Results The 1-year OS and CSS rates in the cohort were 10.5％(95％CI:8.7％-12.7％)and 11.8％(95％CI:9.8％-14.2％),respectively.In multi-variable survival analysis,insurance status,small tumor,tumor stage 1-2,negative AFP,chemotherapy treatment,and surgical treatment were associated with OS.Sex,insurance status,tumor staging,AFP status,chemotherapy and surgery treatment were incorporated into the nomogram for CSS prediction.The bootstrap-corrected concordance indexes(C-indexes)predicted by nomo-gram were 0.72(95％CI:0.70-0.74)and 0.71(95％CI:0.69-0.73),which could be used to predict OS and CSS.The models were internally validated and shown to have good calibration.Conclusion The nomograms are established based on the associat-ed factors,which shows good performance in predicting survival in patients with hepatocellular carcinoma and pulmonary metas-tasis.