Background: For patients with hepatocellular carcinoma (HCC), the selection of effective therapeutic options depends on a reliable prognostic assessment of both the tumor characteristics and liver impairment. This study sought to provide a modified Japan Integrated Stage (m-JIS) score to predict more accurately the survival of patients with HCC.
Methods: We analyzed the records of 42,269 patients diagnosed with HCC that were registered between 1992 and 1999 in a nationwide Japanese database. The m-JIS score was calculated from the tumor-node-metastasis stage and the grade of liver damage as defined by the Liver Cancer Study Group of Japan. The predictive accuracy for patient survival based on the m-JIS score was compared with that determined by the modified Cancer of the Liver Italian Program (m-CLIP) score using the cross-validation method.
Results: Patients were divided randomly into two groups, a training sample for construction of prediction models (n = 21,127 patients with 8,458 deaths) and a validation sample for assessment of those prediction models (n = 21,142 patients with 8,434 deaths). Both the m-JIS score and the m-CLIP score showed good discriminatory ability in the training sample. In the validation sample, the residuals of prediction based on the m-JIS score were smaller than those of the m-CLIP score (P < 0.0001).
Conclusions: The m-JIS scoring system had better predictive accuracy than the m-CLIP score system for survival of Japanese patients with HCC.