Search

The hypothesis that adult infection with the hepatitis B virus in the United States leads to a carrier state with a high risk of primary liver cancer was tested in two ways: (a) a cohort mortality study of U.S. Army veterans given yellow fever vaccine contaminated with hepatitis B virus in 1942 and controls and (b) a case-control study comparing veterans with hepatocellular carcinoma in Veterans Affairs hospitals with matched controls with respect to receipt of contaminated vaccine in 1942. Three groups totaling 69,988 men were the subjects of the cohort study: group 1 comprised men hospitalized with hepatitis in 1942, group 2 comprised men subclinically infected in 1942 and group 3 comprised controls who entered service after the contaminated vaccine was discontinued. Hepatocellular carcinoma cases (n = 24) and control subjects (n = 63) derived from Veterans Affairs hospital discharge files were the subjects of the case-control study. Group comparisons of death rates from liver cancer were refined by expert review of records to select hepatocellular carcinoma from among all causes of death so diagnosed in the cohort study. Slightly excess mortality was found for hepatocellular carcinoma in group 2 (subclinical hepatitis B) but not for group 1 (overt hepatitis B) compared with group 3 (controls) (p = 0.08). Mortality from nonalcoholic chronic liver disease was less in group 2 than in group 3. In the case-control study, the relative risk for hepatocellular carcinoma conferred by receipt of contaminated vaccine was estimated as 3.3 (p = 0.06). We conclude from the cohort study that immunocompetent adult males rarely become carriers after hepatitis B virus infection, probably far less often than the frequently assumed rate of 5% to 10%. The small excess liver cancer mortality seen in the cohort study and the results of the case-control study are consistent, nevertheless, with the now well-established etiological role of hepatitis B virus infection in liver cancer.