Objective
This study aimed to determine the predictors for significant hepatic abnormality (SHA),
a treatment indication, by assessing demographic, laboratory, and radiological results
of chronic hepatitis B (CHB) patients who underwent liver biopsy.
Materials and Methods
In this retrospective study, individuals with untreated hepatitis B e-antigen
(HBeAg)-negative CHB infection were enrolled. Multivariate analysis modeling was
conducted with parameters identified as predictors for SHA in univariate analysis.
Optimal threshold levels for variables to predict SHA in patients with chronic hepatitis
B were determined based on receiver operating characteristic (ROC) curve analysis.
Results
A total of 566 patients with untreated chronic hepatitis B were included in the cohort;
61% (345/566) were male, and the median age was 41 years (interquartile range
[IQR]=34-50). Notably, 36.9% (209/566) had SHA. In the multivariate analysis, utilizing
different models, age, gender, HBV-DNA, LDL, ALT, and platelet count were identified as
the most reliable predictors for SHA in CHB patients. For predicting SHA, the area under
the ROC curve values of HBV-DNA, AST, and ALT were 0.704 (sensitivity=62.8%,
specificity=76.2%;
p
<0.0001), 0.747 (sensitivity=51.9%,
specificity=88.9%;
p
<0.0001), and 0.737 (sensitivity=68.6%,
specificity=68.4%;
p
<0.0001), respectively.
Conclusion
In our study, age, male gender, ALT, AST, HBV-DNA, LDL cholesterol, platelet count, and
FIB-4 score were independent predictors of SHA in HBeAg-negative chronic hepatitis B.
The most sensitive parameters for SHA were LDL and ALT. The most specific parameters
were age, AST, and APRI score. SHA may occur in patients with high HBV-DNA levels, even
if ALT values are normal in HBeAg-negative patients.