We read with interest the article by Lin et al (1) on the prediction of hepatitis B surface antigen (HBsAg) loss. They concluded that rapid HBsAg decline >0.5 log10 IU/mL/yr alone can predict HBsAg loss in 1 year with a positive predictive value (PPV) of 0.517 and negative predictive value (NPV) of 0.972. Several points require clarification for further discussion.

First, Table 3 presented that none of the patients with HBsAg >200 IU/mL achieved HBsAg loss within 1 year, and all data of the 2 groups (<200 IU/mL + >0.5 log10 decline vs >0.5 log10 decline alone) were identical, suggesting that only patients with HBsAg <200 IU/mL achieved HBsAg decline >0.5 log10 IU/mL. With this, the evidence is insufficient to conclude that “HBsAg decline >0.5 log10 IU/mL alone among patients with various HBsAg levels” can predict HBsAg loss in 1 year. Second, the PPV of HBsAg level <200 IU/mL alone was only 10%, which increased to 52% if combined with >0.5 log10 IU/mL. By contrast, our earlier study showed that a cutoff level of <200 IU/mL had a PPV of 36% for the prediction of HBsAg in 1 year, which increased to 97% when combined with a precipitous decline of >1.0 log10 IU/mL HBsAg in the preceding 2 years (>0.5 log10 in 1 year) (2). A study on HBsAg kinetics after the cessation of entecavir therapy further showed that a precipitous HBsAg decline (>0.5 log10 IU/mL in 1 year) predicted HBsAg loss in 1 year with a PPV of 76.6% and in 3 years with an NPV of 100% while a steep decline (>0.76 log10 IU/mL in 1 year) showed an NPV of 66.7% and PPV of 92.9% for HBsAg loss in 3 years (3). A study from Hong Kong also found HBsAg of 200 IU/mL to be an optimal threshold level and combined with an annual HBsAg decline of >0.5 log10 IU/mL may predict HBsAg seroclearance in 3 years with a sensitivity of 74.1% and specificity of 89.4% (4). These remarkable differences need to be explained or discussed.

In conclusion, there are points that require further clarification, and more data have to be presented to verify or refute the current conclusions. The absolute HBsAg level of <200 IU/mL seems to be a prerequisite in the prediction of HBsAg seroclearance in 1–3 years.

CONFLICTS OF INTEREST

Guarantor of the article: Yun-Fan Liaw, MD.

Specific author contributions: Y.-C.C. and Y.-F.L.: contributed equally.

Financial support: The authors were supported by grants from Chang Gung Medical Research Fund (CMRPG1K0101-3, CMRPG1K0111-3) and the Prosperous Foundation, Taipei, Taiwan.

Potential competing interests: None to report.

REFERENCES 1. Lin HC, Liu J, Pan MH, et al. Rapid decline rather than absolute level of HBsAg predicts its seroclearance in untreated chronic hepatitis B patients from Taiwanese communities. Clin Transl Gastroenterol 2023;14:e00586. 2. Chen YC, Jeng WJ, Chu CM, et al. Decreasing levels of HBsAg predict HBsAg seroclearance in patients with inactive chronic hepatitis B virus infection. Clin Gastroenterol Hepatol 2012;10:297–302. 3. Seto WK, Wong DK, Fung J, et al. A large case-control study on the predictability of hepatitis B surface antigen levels three years before hepatitis B surface antigen seroclearnce. Hepatology 2012;56:812–9. 4. Jeng WJ, Chang ML, Liaw YF. Off-therapy precipitous HBsAg decline predicts HBsAg loss after finite entecavir therapy in HBeAg-negative patients. J Viral Hepat 2019;26:1019–26.