Original article
Clinical Predictors for Hepatocellular Carcinoma in Patients With Primary Biliary Cirrhosis

https://doi.org/10.1016/j.cgh.2006.09.031Get rights and content

Background & Aims: Although hepatocellular carcinoma (HCC) occurs with increased frequency in patients with primary biliary cirrhosis (PBC), the exact frequency is relatively low. Optimal selection of PBC patients for HCC screening needs to be determined for effective screening. In this study, we aimed to explore clinical predictors of HCC in PBC patients. Methods: We performed a case-control study using 17 PBC patients with HCC identified from 1976 to 2002 at the Mayo Clinic. Control PBC patients who had no evidence of HCC were selected for each case by matching the first year of their visit to the Mayo Clinic. All medical information was collected within 2 years from when the cases were diagnosed with HCC. Logistic regression models were used for the analyses. Results: Age, sex, history of blood transfusion, current smoking, histologic stage at PBC diagnosis, any signs of portal hypertension, Mayo score, hemoglobin level, platelet count, aspartate aminotransferase level, and albumin level were associated with the presence of HCC (P < .05 for each). In multivariable analysis, older age (OR, 1.7; 95% confidence interval [CI], 1.1–2.5 for 5 years), male sex (OR, 9.7; 95% CI, 1.4–68.3), history of blood transfusion (OR, 5.0; 95% CI, 1.0–24.3), and any signs of portal hypertension (OR, 22.9; 95% CI, 3.4–155.3) were associated significantly with increased odds of HCC and yielded an excellent diagnostic performance (area under the receiver operating characteristics curve rate, 0.91). Conclusions: Older age, male sex, history of blood transfusion, and any signs of portal hypertension or cirrhosis indicate higher likelihood of HCC and should be considered for HCC screening. Further studies in larger patient cohorts are required to verify the diagnostic model.

Section snippets

Study Design and Population

We conducted a case-control study using patients with PBC who were seen at the Mayo Clinic from 1976 to 2002. Case subjects were patients who had a diagnosis of PBC and HCC. Control subjects were patients who had a diagnosis of PBC but did not have HCC during their follow-up examination. The diagnosis of PBC was established by the following: (1) laboratory abnormalities consistent with chronic cholestatic liver disease for at least 6 months, (2) serum alkaline phosphatase level of at least 1.5

Results

The follow-up time defined as the time difference between the PBC diagnosis at the Mayo Clinic and HCC diagnosis (or closest visit in controls) was not significantly different between the groups (93 ± 18 vs 66 ± 10 mo, P = .194). For controlling the unmatched difference in follow-up time, the follow-up time was taken into account in the analyses.

The clinical characteristics of the study population are summarized in Table 2. As shown in Table 2, age, sex, history of blood transfusion, current

Discussion

In this study, we conducted a case-control study using patients with PBC who were seen at the Mayo Clinic from 1976 to 2002 and showed that older age, male sex, history of blood transfusion, and the presence of portal hypertension were associated significantly with increased odds of HCC among patients with PBC even after controlling for the other factors and the length of follow-up time. The model with male sex, age, history of blood transfusion, current smoking, and the presence of ascites

References (26)

  • F.P. Sarasin et al.

    Cost-effectiveness of screening for detection of small hepatocellular carcinoma in western patients with Child-Pugh class A cirrhosis

    Am J Med

    (1996)
  • M.M. Kaplan

    Primary biliary cirrhosis

    N Engl J Med

    (1996)
  • L. Loof et al.

    Cancer risk in primary biliary cirrhosis: a population-based study from Sweden

    Hepatology

    (1994)
  • Cited by (0)

    Supported in part by the Miles and Shirley Fiterman Center for Digestive Diseases at the Mayo Clinic, Rochester, MN.

    View full text