M. Cottone et al., SCREENING FOR HEPATOCELLULAR-CARCINOMA IN PATIENTS WITH CHILDS-A-CIRRHOSIS - AN 8-YEAR PROSPECTIVE-STUDY BY ULTRASOUND AND ALPHA-FETOPROTEIN, Journal of hepatology, 21(6), 1994, pp. 1029-1034
One hundred and forty-seven patients with Child's A cirrhosis and no e
vidence of hepatocellular carcinoma were followed up in an X-year pros
pective surveillance program with testing by ultrasound and alphafetop
rotein every 6 months. Eighteen of 147 patients were HBsAg positive. A
nti-hepatitis C virus antibodies were found in 103 out of 133 cases te
sted. Sixteen patients had a history of heavy drinking. Thirty hepatoc
ellular carcinomas were detected during follow up. At the time of diag
nosis, ultrasound detected focal lesions in all the patients whereas a
lphafetoprotein was below diagnostic levels. The hepatocellular carcin
oma was single in 26 patients and multiple in four. The overall 8-year
cumulative tumor-free rate was 69% (95% confidence interval=58-73). T
he yearly hepatocellular carcinoma incidence from 1985 to 1992 was res
pectively 2%, 1.5%, 2%, 3%, 5%, 4.8%, 7% and 10%. The initial value of
AFP>50 ng/ml and <400 ng/ml was significantly related to the developm
ent of hepatocellular carcinoma. This series shows that the cumulative
incidence of hepatocellular carcinoma in cirrhosis in Italy is higher
than previously reported, but lower than that observed in Asiatic are
as. A 6-month interval for ultrasound is reasonable to detect treatabl
e tumors. Alphafetoprotein has no value for early diagnosis, although
its intermediate values (>50 and <400 ng/ml) may indicate the presence
of undetectable cancer which will appear during the follow up, and su
ggests that ultrasound should be employed more frequently in patients
with these values. (C) Journal of Hepatology.