The clinical course and pathological patterns of a group of 13 patients with both primary liver cell carcinoma and Hepatitis B surface antigen (HB,Ag) are described and contrasted with those of 43 patients with primary liver cell carcinoma but without HB.Ag. HB,Ag-positive carcinoma patients demonstrated a higher incidence of splenomegaly, transudative ascites, and the presence of alphafetoprotein, although none of these reached statistical significance. Serum bilirubin was significantly higher in patients with HB.Ag. HB,Ag-positive carcinoma patients most frequently originated from countries where the presence of HB,Ag is high in the general population. Survival time from the diagnosis of primary liver cell carcinoma was shorter in patients with HB,Ag.Cancer 38:901-905, 1976.
EPATITIS B SURFACE ANTIGEN ( H B a g ) HASH been related to the development of chronic active hepatitis and c i r r h o~i s . '~~'~~~' HB,Ag-positive cirrhosis has been described also occurring with hepatocellular carcinoma.'7*21 In some geographic areas, there is not only a high incidence of both HB,Ag and primary liver cell carcinoma occurring separately, but also a higher incidence of antigen in the patients with carcinoma.'"4~'7"9"22'23'24 It is important to define any differences between the HB,Ag-positive and negative patients with primary liver cell carcinoma, as this may affect earlier diagnosis, prognosis, and therapy.
MATERIAL AND METHODS PatientsFifty-six patients with primary liver cell carcinoma treated at the Koyal Free Hospital, Lon- don, between 1969 and 1974 were tested for the presence of Hepatitis B surface antigen. The clinical course, laboratory investigations, and serologic changes were reviewed. All patients had histologic evidence of primary liver cell carcinoma. Cases of intrahepatic bile duct carcinoma were not included.
TechniquesSera were examined for HB,Ag by a n Ouchterlony t e c h n i q u e , '' or by c o u n t e rimmunoelectrophoresis.'2 T h e presence of alpha-fetoprotein was tested for by a doublediffusion technique in agar," without quantitative measurement. Serum biochemical values were estimated by standard laboratory techniques.Histologic specimens were examined by staining with hematoxylin and eosin, and by impregnation with silver for reticulin fibers. Other stains were performed when necessary. Where wedge biopsies or postmortem specimens were available, sections were stained for the HB,Ag by a modified version6 of a n orcein stain described by Shikata et al." Liver scanning was performed with technetium-99m sulfide. Selective coeliac axis arteriography was performed in the standard manner. (Table 1) T h e ages of the 53 men and 3 women ranged from 17 to 76 years and 22 to 70 years, respectively. HB,Ag-positive patients were all male, with a n age range of 21-67 years and a mean age 90 1
RESULTS
Patients
