Chronic HBV infection can progress to persistent or intermittent liver inflammation, and further to cirrhosis, HCC and end-stage liver disease [1, 2], which is the leading cause of most viral hepatitis related mortality. Therefore, early and reasonable antiviral therapy can significantly reduce the necro-inflammatory activity, and reduce the occurrence of HCC and other related complications [16, 17]).
In recent years, serum ALT levels reflect the severity of liver damage , but ALT levels are not consistently elevated in all patients with chronic HBV-infected. Studies have shown sever liver damage in some patients with CHB with normal ALT levels [18, 19]. Liver biopsy in 28–37% of CHB patients with normal ALT or mildly elevated ALT levels suggests moderate or severe inflammation or significant fibrosis, requiring antiviral therapy [4–6]. In addition, several large cohort studies have reported that patients with CHB with normal ALT levels are risk for developing cirrhosis and HCC [20, 21]. Other studies have shown that 18% of CHB patients with normal ALT levels have significant liver fibrosis [4, 5]. Therefore, normal ALT levels do not rule out the absence of liver damage. Given the fact that early and long-time antiviral therapy can reverse liver fibrosis and cirrhosis, accurate assessment of disease progression is critical.
Although liver biopsy remains the gold standard for evaluating liver fibrosis. However, due to the occurrence of related complications and risk associated with liver biopsy, it’s not appropriate for all patients . Considering the shortcoming of liver biopsy, most noninvasive methods had been developed for the evaluation of liver fibrosis, such as transient elastography (TE), serum markers and LSM. Among these noninvasive methods, APRI and FIB-4 had received high scores in several major guidelines for the management of CHB. However, their diagnostic accuracy remains unclear [1, 23–25].
Both APRI and FIB-4 used routine laboratory tests to predict the severity of liver fibrosis, and were validated in patients with HBV-infected with mild elevated ALT. In this study, we compared the diagnostic performance of two models to predict significant and advanced fibrosis in patients with CHB with normal ALT, mildly and significantly elevated ALT. Based on our data, 65(41.7%) of patients with normal ALT had significant fibrosis. Among patients with mildly elevated ALT, 35(41.2%) had significant fibrosis. Of the patients with significantly elevated ALT, 28(63.6%) had significant fibrosis.
In a retrospective study, the AUROC of APRI in patients with F2, F3 and F4 fibrosis in CHB was 0.72, 0.812 and 0.707, respectively . Another study found that the AUROC of APRI was 0.65, 0.659 and 0.735 in CHB patients with normal ALT, mildly and significantly elevated ALT . Based on our data, the AUROC of APRI in CHB patients with normal ALT, mildly and significantly elevated ALT was 0.608, 0.634, 0.708 and 0.636, 0.751, 0.708 in significant and advanced fibrosis, respectively. Therefore, we believe that APRI is more accurate in the diagnostic of advanced fibrosis than significant fibrosis, especially in the patients with CHB with mildly elevated ALT.
Based on this study, FIB-4 is better than APRI in the diagnostic of both significant and advanced fibrosis. A previous study showed that FIB-4 had a high diagnostic accuracy for CHB, with AUROC in differentiating between significant and advanced fibrosis was 0.730 and 0.748 in all patients . In our study, the AUROC of FIB-4 in CHB patients with normal ALT, mildly and significantly elevated ALT was 0.638, 0.679, 0.734 and 0.652, 0.763, 0.734 in significant and advanced fibrosis, respectively. Surprisingly, FIB-4 was superior to APRI in the diagnostic both significant and advanced fibrosis.
Our study also has some limitations. First of all, the patients included in this study were all from the same medical institution, the absence and deviation of clinical data may have a certain impact on the accuracy of models. Secondly, we defined normal ALT levels based on our own laboratory values. The AST levels of most patients have a great influence for liver fibrosis, so we can conduct stratified study on the AST level in the further study. Finally, we should analyze and verify many noninvasive models to better evaluate liver fibrosis.