Among chronic HBV carriers who visited the Hepatology Center of Bundang Jesaeng General Hospital between March 2003 and September 2015, all patients with HBsAg seroclearance were recruited. The clinical and laboratory data were retrospectively recorded at baseline and during follow-up. Chronic carriers were defined as those with HBsAg positivity on two successive tests performed at least six months apart. Registration or entry was defined as the initial visit to the center. The baseline data included the data collected at the first visit. HBsAg seroclearance was defined as two consecutive negative HBsAg tests performed at least six months apart.

The patients underwent tests for biochemical liver function, hepatitis B virological markers, and alpha-fetoprotein, as well as liver ultrasonography (US). The virological markers assessed included HBsAg, anti-HBs, HBeAg, anti-HBe, and HBV DNA. The anti-HBs or anti-HBe test was performed as clinically indicated. The anti-HCV test was performed at baseline and was repeated if the alanine aminotransferase level was elevated to more than twice the upper limit of normal. The signal-to-cutoff ratio of the HBsAg (HBsAg-SCR) level was measured for each patient by qualitative assay.

The HBsAg, anti-HBs, HBeAg and anti-HBe levels were determined by ARCHITECT qualitative assay, which is a chemiluminescent microparticle immunoassay (CMIA) that measures the resulting chemiluminescent reaction as relative light units (RLUs) (Abbott Laboratories, North Chicago, IL, United States). There is a direct relationship between the amount of HBsAg in a sample and the RLUs detected by ARCHITECT assay (Abbott Laboratories). The anti-HCV levels were determined using a third-generation enzyme immunoassay (Abbott Laboratories). HBV DNA was tested using a commercially available real-time polymerase chain reaction (RT-PCR; Roche Molecular Systems, Pleasanton, CA, United States), according to the manufacturer’s instructions. The lower limit of detection for the Roche assays is 20 IU/mL. The viral load data obtained with the Roche assay were converted to international reference units for analysis using approximations for 1 pg (283000 copies) and 1 IU/mL (5.8 copies/mL).

The data on the HBsAg levels in 177 patients were separately collected to perform correlation analysis with the HBsAg-SCRs obtained with CMIA. The HBsAg quantification (HBsAg-QNT) was measured by electrochemiluminescence immunoassay using an Elecsys HBsAg II Quant Assay (Roche Diagnostics, Indianapolis, IN, United States), according to the manufacturer’s instructions. This approach quantifies HBsAg against an internal World Health Organization reference standard in IU/mL.

To compare the characteristics between groups, either the χ² test or Fisher’s exact test was used to analyze categorical variables, and Student’s t-test or Wilcoxon nonparametric test was used to analyze continuous variables. Statistical data were expressed as the mean ± SD. The correlation between the HBsAg-SCR and HBsAg-QNT was analyzed. The statistical procedures were performed with R-project version 3.2[15]. P values of < 0.05 were considered significant.

Negative seroconversion of HBsAg was observed in 90 of 1919 carriers (4.7%), with an estimated annual rate was 0.76%. The rate of HBsAg loss was higher in the carriers over the age of 40 than in the younger carriers (5.7% vs 2.5%, P = 0.001). Among the carriers with HBsAg seroclearance, male predominance was more significant in the 40-50 and 50-60 year age groups than in the 30-40 and > 60 year age groups (P < 0.0001). The period of time until HBsAg clearance after registration was 6.48 ± 3.76 years. This value did not significantly differ among the age groups (P = 0.148) (Table 1).

Among the 90 carriers with HBsAg seroclearance, 82.2% were HBeAg-negative at registration. Among these HBeAg-negative carriers, 95.3% had no history of anti-viral treatment and were considered to be in the inactive phase. The time interval to HBsAg clearance did not differ between the HBeAg-positive and negative carriers (P = 0.982). Among the carriers with HBsAg seroclearance who had a treatment history, the time interval to HBsAg clearance did not differ between the HBeAg-positive and negative carriers (8.13 ± 3.12 vs 7.95 ± 3.45 years, P = 0.4442). However, this time interval was longer for the carriers who had received treatment than for those who had not received treatment (7.95 ± 3.34 vs 5.52 ± 3.48 years, P = 0.017), suggesting that anti-viral therapy does not lead to HBsAg seroclearance[16,17]. On the other hand, the time interval from entry to HBsAg seroclearance was significantly longer for the HBeAg-positive carriers than for the HBeAg-negative carriers in the 40s and 50s age groups (9.24 ± 2.49 years vs 6.18 ± 3.63 years, P = 0.013) (Table 2). Likewise, the time interval was significantly longer for the carriers with an HBV DNA level of > 2000 IU/mL than for those with a level of < 20 IU/mL (8.31 ± 3.26 years vs 4.92 ± 3.36 years; P = 0.0005) (Table 3). Data on the BCP double mutations, A1762G and A1764T, and the e-minus mutation, G1896A, were available for 18 patients. Women were predominant in the wild-type BCP group, whereas men were predominant in the mutant group (male/female = 1/6 vs 9/2, respectively; P = 0.0014). The BCP mutation carriers tended to be older than the wild-type carriers (41.9 ± 10.8 years vs 48.5 ± 7.6 years old, P = 0.0725). However, the time interval from entry to HBsAg seroclearance did not differ between the groups.

A log-linear-shaped correlation was detected between the HBsAg-SCR and HBsAg-QNT (Figure 1A), and a good correlation with a linear drift curve was observed between an HBsAg-SCR < 1000 and an HBsAg-QNT < 100 IU/mL (y = 0.0674x + 0.9902, R² = 0.916) (Figure 1B). No correlation was observed between these two values in the patients with HBsAg-QNT levels of over 200 IU/mL (Figure 1C). An inverse correlation was observed in the patients with very high HBsAg titers of over 10000 IU/mL (Figure 1D), indicating a prozone effect on the HBsAg-SCR caused by excess antigen. Based on the HBsAg-SCR value of 1000, we arbitrarily divided the 90 carriers with HBsAg seroclearance into two groups. The period of time from entry to HBsAg seroclearance was significantly shorter in the < 1000 group than in the ≥ 1000 group (5.07 ± 3.74 years vs 7.47 ± 3.05 years, P = 0.0008). Moreover, a sequential decrease in the HBsAg-SCR was predictive of HBsAg-NC in every case (Figure 2). The HBeAg-positive rate was significantly higher in the ≥ 1,000 group than in the < 1000 group (25% vs 11.9%, P = 0.0057). Further, the mean age in the < 1000 group tended to be higher than that in the ≥ 1000 group (46.2 ± 11 years vs 50 ± 11.1 years old, P = 0.0593). However, there was no difference in the sex ratio or the prevalence of HCC or LC between the two groups.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Signal-to-cutoff ratio of hepatitis B surface antigen was measured by qualitative assay and partially reflects the quantity of hepatitis B surface antigen in the serum. A: There was a log-linear correlation between the HBsAg-SCR and HBsAg-QNT; B: There was a good correlation with a linear drift curve between an HBsAg-SCR < 1000 and an HBsAg-QNT < 100 IU/mL (y = 0.0674x + 0.9902, R² = 0.916); C: HBsAg-QNT levels of greater than 200 IU/mL were not correlated with the HBsAg-SCR; D: Very high HBsAg titers of more than 10000 IU/mL were inversely correlated with the HBsAg-SCR, which was caused by a prozone effect. HBsAg-SCR: Signal-to-cutoff ratio of HBsAg; HBsAg-QNT: Quantity of HBsAg; HBsAg: Hepatitis B surface antigen.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Sequential changes in the signal-to-cutoff ratio of Hepatitis B surface antigen in the patients with negative conversion of Hepatitis B surface antigen during the follow-up period after registration. A: HBsAg-SCRs in the patients with HBsAg-NC after 5 years post-entry; B: HBsAg-SCRs in the patients with HBsAg-NC before 5 years post-entry. HBsAg-SCR: Signal-to-cutoff ratio of HBsAg; HBsAg-NC: Negative conversion of HBsAg; HBsAg: Hepatitis B surface antigen.

We classified the patients with HBsAg seroclearance into the following three groups according to their US findings: normal, non-cirrhotic chronic liver disease (CLD) and cirrhotic. The period of time from study entry to HBsAg loss was significantly shorter in the normal US group than in the CLD and cirrhotic US groups (5.05 ± 3.92, 6.86 ± 3.51 years and 6.8 ± 2.98 years, respectively; normal US vs CLD, P = 0.0259; normal vs cirrhotic US, P = 0.0366) (Table 4). Among the 90 patients with HBsAg seroclearance, 31% had asymptomatic, inactive LC without HCC at registration. Men were predominant in the 30s (3/3) and 40s (9/10) age groups, but the proportion of females was increased in the 50s (3/5) and 60s (5/10) age groups. Five patients (17.9% of 28; male/female = 4/1) developed HCC during follow-up, including one before and four after HBsAg loss.

Among the 90 patients with HBsAg seroclearance, thirteen were diagnosed with HCC, including seven (7.8%) at registration (HCC group-1) and six (6.7%) during follow-up (HCC group-2). Among the patients in HCC group-2, two developed HCC prior to HBsAg loss, and four developed HCC at 0.66 ± 0.63 years after HBsAg loss, corresponding to a rate of 4.8% among the 83 carriers without HCC at entry. There was no difference in the time interval from entry to HBsAg loss between HCC group-1 and group-2 (6.78 ± 2.77 years and 5.77 ± 3.18 years, respectively). In addition to LC, the risk factors for HCC included excessive drinking in three men and vertical infection in one woman.

Chronic hepatitis B virus (HBV) infection follows a unique natural course, consisting of immune tolerance, eradication, and a recovery phase. Intricate changes in the balance between host and viral factors stimulate the phase transition. Prognosis of chronic carriers is determined by a complicated process with consideration of phase transition. Seroclearance of hepatitis B surface antigen (HBsAg) is the final step of the recovery phase. However, the clinical features of HBsAg seroclearance have not yet been completely elucidated.

The recovery phase usually overlaps with the occurrence of escape mutations in the HBV genome and the development of liver cirrhosis (LC) and hepatocellular carcinoma (HCC). Thus, it is difficult to determine when and how the negative conversion of HBsAg (HBsAg-NC) has occurred to a sufficient extent to have preventive effects for HCC. The research hotspot is clarification of the clinical features associated with spontaneous HBsAg seroclearance, including the outcome and predictive markers, to facilitate the development of a patient management strategy both during and after the recovery phase.

The results of the present study have suggested that the phase transition to recovery occurs spontaneously in most patients and that asymptomatic LC is prevalent among these patients (31%). Thus, a significant proportion of chronic carriers are at risk of HCC at the time of recovery. In fact, the data showed that HCC developed among the patients with LC during the recovery phase both pre- and post-HBsAg seroclearance. On the other hand, it is known that the HBsAg quantity, in combination with the HBV DNA titer, is a predictive marker for HBsAg seroclearance. The authors have demonstrated for the first time that a signal-to-cutoff ratio of the qualitative HBsAg level (HBsAg-SCR) of < 1000 and its sequential decrease are notable predictive markers of HBsAg loss. Our data support the value of sequential HBsAg-SCR data.

The results of our study may facilitate the design of a management strategy for HBV carriers both pre- and post-HBsAg seroclearance. In particular, patients with evidence of chronic liver disease or LC should be carefully monitored for the development of HCC, even after the spontaneous loss of HBsAg.

Serum HBsAg positivity lasting for six months is a serological indicator for chronic HBV infection. HBsAg seroclearance refers to the spontaneous HBsAg-NC in the serum. The HBsAg-SCR is the signal-to-cutoff ratio of HBsAg, as measured by qualitative assay, and the quantification of HBV DNA and HBsAg (HBsAg-QNT) is the quantity of HBsAg in the serum, as measured by quantitative assay. The HBsAg-SCR is determined according to the signal intensity of an antigen-antibody reaction. The authors found that carriers with a very high HBV DNA titer of > 7 logs IU/mL exhibited a prozone effect, in which the HBsAg-SCR was paradoxically reduced by excess antigen. In contrast, in the HBV DNA-negative patients, an HBsAg-SCR < 1000 was strongly correlated with an HBsAg-QNT < 200 IU/mL. Otherwise, the HBsAg-SCR was not correlated with the HBsAg-QNT.

This invited manuscript written by Park et al investigated the clinical features of HBsAg serological clearance among almost 2000 Korean HBV carriers. This is a valuable study that included large-scale clinical sample analyses; the significant conclusion of this study is that liver cancer surveillance should be continued after HBsAg seroclearance, particularly in patients with cirrhosis. The title describes the contents of the paper. The abstract is informative and completely self-explanatory; it briefly presents the topic, states the scope of the experiments, provides the significant data, and notes the major findings and conclusions. The purpose or purported significance of the article is explicitly stated. The research study methods are complete enough to enable the experiments to be reproduced. All figures and tables are necessary and appropriate. The discussion interprets the findings in view of the results obtained in this and in past studies on this topic.