Serum alpha-fetoprotein concentrations in urban and rural Southern African blacks with hepatocellular carcinoma.

The reason why only some hepatocellular carcinomas synthesize alpha-fetoprotein is not known. Both the frequency with which this foetal globulin is produced and the major aetiological associations of hepatocellular carcinoma vary between populations with high and low incidences of the tumour, raising the possibility that re-expression of the gene for alpha-fetoprotein is determined, or influenced by, the molecular genetic events that occur during hepatocellular carcinogenesis. This hypothesis could be tested by comparing serum alpha-fetoprotein concentrations in populations in which the major risk factors for hepatocellular carcinoma differ. Two such populations are urban and rural southern African blacks. We measured serum alpha-fetoprotein concentrations by radioimmunoassay in 234 southern African blacks with hepatocellular carcinoma: 78 of the patients were urban and they were age-matched with 156 patients born in rural areas, one-half of whom had remained in a rural environment (rural), whereas the others had migrated to the cities in adulthood (rural-urban). Urban patients were more likely than rural-born patients to have a normal serum alpha-fetoprotein value [23.1% (18/78) compared with 10.2% (16/156); p = 0.02]. There was no significant difference between the concentrations in rural and rural-urban patients. The absolute values of the raised serum alpha-fetoprotein values did not differ between urban (69,558 +/- 176,737 ng/ml; and rural-born patients (53,998 +/- 125,681 ng/ml), or between rural (69,207 +/- 159,975 ng/ml) and urban-rural patients (40,434 +/- 83,028 ng/ml). These findings are compatible with the hypothesis that re-expression of the alpha-fetoprotein gene in hepatocellular carcinoma is related to the aetiology or pathogenesis of the tumour.

The pathogenesis of raised serum alpha-fetoprotein levels in amoebic and pyogenic hepatic abscesses.

Serum alpha-fetoprotein (AFP) concentrations may be slightly raised in patients with amoebic hepatic abscesses. In an attempt to learn more about the pathogenesis of the raised levels, we studied 74 patients with amoebic and six with pyogenic hepatic abscesses. Serum (AFP) levels were slightly elevated (24-72 ng/ml) on admission in four patients and markedly raised (2000 ng/ml) in one, who had hepatocellular carcinoma in addition to a pyogenic hepatic abscess. The pattern of the early rise in AFP levels could not be determined because these four patients were lost to follow-up. However, serial serum AFP estimations were obtained in 29 patients with a normal value on admission and in none of these did the concentration rise during recovery. Our findings do not support the prevailing hypothesis that regenerating hepatocytes are responsible for the raised serum AFP levels in non-neoplastic hepatic disorders, including hepatic abscesses.