Background/aims: Dosage of serum AFP (alpha-fetoprotein) is widely used for HCC screening in patients with chronic liver disease. Virus-related chronic liver disease is the main cause of cirrhosis and HCC in Western and Far Eastern countries, but the relationship between viral etiology and AFP levels in HCC is still unclear. The aim of this study was to verify, in Western patients with post-viral chronic liver disease, the usefulness of AFP dosage for the detection of HCC, and the influence of viral etiology on AFP levels in HCC.
Methodology: The study population included 350 patients with post viral chronic liver disease that underwent liver biopsy, serum AFP determination and ultrasound liver evaluation. Seven patients had normal liver histology, 197 had chronic hepatitis, 72 had cirrhosis, and 74 had cirrhosis and HCC. ROC (receiver operating characteristic) analysis was used to assess the best diagnostic AFP threshold value for HCC detection. Logistic regression analysis was performed to individuate independent predictors of HCC diagnosis.
Results: No difference was observed in AFP levels between HCV- and HBV-positive patients, neither in the whole population nor in the HCC patients only. ROC area under curve for AFP was 0.801 (95% CI: 0.721-0.867). The analysis individuated a best accurate AFP threshold value for HCC diagnosis of 50 ng/mL. HCC was detected with specificity > or = 95% only for AFP > 100 ng/mL. The sensitivity however was poor (25%). Male sex, age > 60, and AFP were independent predictors of HCC diagnosis.
Conclusions: Serum AFP levels in HCC patients are not influenced by virus B or C hepatitis pattern. AFP dosage should not be used for HCC diagnosis in non-cirrhotic patients. Male patients with cirrhosis should be regarded with a more "aggressive" screening program compared to females.