METHOD OF ESTIMATING DEVELOPMENT OF HEPATITIS C VIRUS-RELATED HEPATOCELLULAR CANCER IN FUTURE, DECISION TREE TO BE USED IN THE METHOD AND METHOD OF CONSTRUCTING THE DECISION TREE

摘要

It is intended to provide a method whereby the development of HCV-related hepatocellular cancer can be estimated in the early stage at a high accuracy by using the decision tree technique before the clinical diagnosis of the development of hepatocellular cancer. The serum of a patient carrying hepatitis C virus whose conditions have already advanced into chronic hepatitis or hepatitis cirrhosis and the serum of a hepatocellular cancer patient are subjected to a protein expression analysis by using the SELDI protein chip system. Based on the results of the analysis, a decision tree for diagnosing hepatitis C virus-related hepatocellular cancer is constructed using a multivariate analysis software. When the effectiveness of this decision tree is examined with the use of the known sera of chronic hepatitis, hepatitis cirrhosis and hepatocellular cancer patients, the sensitivity and specificity between a hepatocellular cancer group and a non-hepatocellular cancer group are respectively 82% and 81%. When the sera of seven patients clinically diagnosed as suffering from hepatocellular cancer that have been stored for 1 year are examined by using this decision tree, six cases are decided as hepatocellular cancer. Thus, this decision tree is useful in the early diagnosis of hepatitis C virus-related hepatocellular cancer.