LAUSR

We read with interest the recent paper by Choi et al.(1) on the impact of biopsy-proven nonalcoholic steatohepatitis (NASH) on chronic hepatitis B (CHB) outcomes. The study concluded that compared to patients with CHB alone, patients with CHB and concomitant NASH were more likely to have advanced fibrosis (F3/F4) and shorter time to liver complications and that superimposed NASH predicted poorer outcomes in those with advanced fibrosis. However, prior studies have also suggested that patients with CHB and coexisting hepatic steatosis may not be at higher risk for disease progression; and in fact, more favorable long-term outcomes were even reported by some.(2) In the current study, the major driver of poor outcome appears to be advanced fibrosis as this was significantly associated with poor outcomes whether NASH was present or not, and there was no significant association between NASH and outcomes if there was no advanced fibrosis. Therefore, the current results should be interpreted with caution. First, two major biases can contribute to the higher risk of advanced disease and complications observed in this study. One is referral bias introduced by inclusion of only patients from tertiary care centers. Two is the selection bias introduced by inclusion of only CHB patients who had undergone liver biopsy, who may have been selected for biopsy due to higher perceived risk. Second, antiviral treatment was notably associated with higher risk of poor clinical outcomes in univariable analysis (hazard ratio, 1.99; 95% confidence interval, 1.13-3.50; P = 0.017), but this factor was not included in the multivariable analysis or discussed. Antiviral therapy can improve hepatic dysfunction even with nonalcoholic fatty liver disease (NAFLD) patients,(3) and neither NAFLD nor related metabolic features influence antiviral treatment outcomes.(4) Therefore, the impact of hepatic steatosis as well as NASH on the natural history of CHB requires further investigation, and specifically the effect of antiviral therapy should be accounted for. Finally, we congratulate Choi and colleagues’ laudable effort to address this very clinically relevant medical condition because both NAFLD and CHB are common, and they frequently coexist.(5)