Mak and colleagues [7] examined the serum hepatitis B core-related antigen (HBcrAg) and HBV pre-genomic RNA (pgRNA) in the first 48 weeks of NA treatment in 64 CHB patients. Analyses were performed separately in 28 HBeAg-positive and 36 HBeAg-negative patients. These patients had participated in previous phase III trials, during which they received lamivudine, adefovir dipivoxil, telbivudine, clevudine, or entecavir with serum samples collected at weeks 0, 4, 12, 24, 36, and 48 of treatment and paired liver biopsies at weeks 0 and 48. HBcrAg was measured with a lower limit of detection (LLOD) of 2 log U/mL and a reliable quantification at >3 log U/mL. HBV RNA was measured using an investigational assay with a LLOD of 1 log copies/mL. Patients with undetectable serum HBV pgRNA or HBcrAg at week 0 were excluded from the analysis of early on-treatment changes of viral markers. Due to censoring and the difference in follow-up durations between patients with and without FHR at the EOFU, time-dependent area under receiver operating curve was utilized to assess the discriminatory ability of HBcrAg and HBV pg RNA.

At a median follow-up duration of 17 years, 22/64 patients achieved FHR, including eight cases of HBsAg seroclearance. At the start of NA therapy, HBeAg-positive patients had higher HBV pgRNA compared to HBeAg-negative patients (4.9 vs. 3.3 log copies/mL). The median HBV pgRNA reduced to the LLOD at week 48 and week 12 for HBeAg-positive and HBeAg-negative patients, respectively. The difference in median reduction of HBV pgRNA between FHR and non-FHR patients was prominent among HBeAg-positive patients, with over 1 log difference sustained between week 4 and week 48. Patients with FHR had a higher HBV pgRNA in the first 12 weeks than those without FHR, while the level was comparable after week 24. The decline of HBV pgRNA at weeks 4 and 48 were the most discriminative of FHR. In contrast, among HBeAg-negative patients, the HBV pgRNA level and its reduction did not significantly differ throughout the first 48 weeks between patients with and without FHR.

Serum HBV RNA is a novel biomarker that can measure the circulating HBV pgRNA present in virus-like particles [8,9]. Serum HBV RNA is a mixture of intact, pre-genomic and subgenomic, spliced, truncated, and polyA-free species [10], which reflects the transcriptional activity of intrahepatic covalently closed circular DNA (cccDNA) [11]. HBV RNA level at NA treatment cessation also correlates with post-treatment viral relapse [11]. Notably, while there was a significant reduction of HBV pgRNA, no significant differences in the reduction of intrahepatic total HBV DNA and cccDNA level were observed at week 48 between HBeAg-positive patients with and without FHR [7]. One may speculate that cccDNA transcriptional activity is more predictive of FHR than its absolute amount. Nevertheless, the difference in the role of pgRNA in HBeAg-positive and HBeAg-negative patients remain unclear. Also, assays for serum HBV RNA should be standardized and validated to better define its clinical utility [11].

Mak and colleagues [7] also demonstrated the importance of early decline in serum HBcrAg at week 4 during antiviral treatment, which was associated with FHR in HBeAg-negative patients. HBcrAg is a novel biomarker in CHB patients and consists of the hepatitis B core antigen, HBeAg, and the 22-kDa precore protein [12]. It gradually decreased after antiviral therapy. Patients with severe alanine transaminase flares had increased HBcrAg levels after antiviral therapy cessation, and the concentration declined after recommencing antiviral therapy [13].

The decline of HBcrAg during antiviral therapy reflects intrahepatic cccDNA reduction and suppression of viral replication activity. HBcrAg decline was positively correlated with HBsAg reduction or HBeAg seroconversion [14]. While the HBeAg expression outnumbers the HBcrAg expression in HBeAg-positive patients, the predictive performance of HBcrAg on FHR may be affected. After antiviral treatment cessation, patients with lower HBcrAg levels are more likely to have HBsAg loss (Table 1) [15]. It may be appropriate to monitor HBcrAg to assess clinical outcomes and treatment effects. HBcrAg would be a substitute marker for predicting the course of the disease.

In addition to being a useful biomarker for monitoring viral replication activity and cccDNA, HBcrAg is of high value in predicting the development of HCC in CHB patients. For treatment-naïve CHB patients, high serum HBcrAg level was associated with the development of HCC [16]. Patients with decreased HBcrAg had lower HCC risk than those with persistently high HBcrAg levels [17]. Serum HBcrAg also had a superior prediction value for predicting the HCC risk than other HBV markers, such as HBV DNA level (Table 1).

HCC risk was not eliminated in antiviral-treated HBeAgnegative patients with high HBcrAg levels. For antiviral-treated CHB patients, among whom HBV DNA and HBsAg levels may not perform well in HCC risk prediction, HBcrAg can predict the incidence of HCC accurately in HBeAg-negative patients with high sensitivity and negative predictive value [4]. It also works in cirrhotic CHB patients. Consistently high on-treatment serum HBcrAg level was associated with a higher HCC incidence, despite prolonged antiviral treatment, in both HBeAg-positive and HBeAg-negative CHB patients (Table 1) [17]. The predictive value for post-treatment HCC recurrence of HBcrAg was also demonstrated. Further studies are needed to explore the underlying mechanism of correlation between high HBcrAg and HCC.