[Evaluation of prognostic factors in hepatocellular carcinoma using multiple regression analysis].

We analysed the relationship between 19 prognostic factors and the life span of patients with hepatocellular carcinoma. Among 52 patients who had been admitted to the Gifu Prefectural Gifu Hospital, in 46 the pathological diagnosis was confirmed. We selected 9 of 19 variables by the stepwise forward regression method. These were Arels Classification, ascites, HBsAg, performance status, method of therapy, sex, GOT, albumin and bilirubin. Results of multiple regression analysis demonstrated that the presence of ascites is the most important factor; other significant factors are Ariel's Classification, performance status, positivity of HBsAg, method of therapy, and GOT. These parameters were paralleled with the survival curves, most of which had been categorized. Although intra-arterial infusion of ADM was considered the most favourable treatment method, the survival curves suggested that other methods were equally effective.