Chronic HBV infection (CHB) is comprised of dynamic interactions between HBV, hepatocytes, and the host immune system . CHB patients may achieve several milestones during the natural history of chronic HBV infection, as shown in Fig. 12.2. These milestones of CHB progression can be divided into two groups; one involves clinical phases that patients may experience, and the other involves parenchymal disease progressions such as the occurrence of cirrhosis and hepatocellular carcinoma (HCC) .

Traditionally, the natural course of perinatally acquired chronic HBV infection involves three chronological phases ; the immune tolerance phase, the immune clearance phase, and the low-replicative residual integrated phase [8]. These phases can be defined based on a combination of serum markers such as HBeAg and its antibody, serum HBV DNA levels, HBsAg, and serum ALT levels. The immune tolerance phase is characterized by HBeAg-seropositivity, high HBV DNA levels, normal-ALT levels and no evidence of liver injury. Most of the liver injury occurs during the immune clearance phase as the host immune system tries to clear infected hepatocytes, which may result in the development of cirrhosis and HCC. This phase features inflammation of the liver, elevation of serum ALT levels, gradual reduction of circulating HBV DNA levels, and hopefully, seroconversion of HBeAg to its antibody (anti-HBe).

Finally, a proportion of infected persons are able to inactivate the replication of HBV and enter the residual phase. This phase is characterized by the continued presence of HBsAg , the presence of anti-HBe antibody, low/undetectable levels of serum HBV DNA, and normal ALT. A very small proportion of infected persons are then able to spontaneously clear HBsAg and resolve the infection. An inactive carrier state can be identified during the late stages of the immune clearance phase and in the residual phase, where HBeAg is seronegative, serum HBsAg remains detectable, but serum HBV DNA levels are lower than 2000 IU/mL with repeatedly normal (or minimally raised) alanine aminotransferase levels. After the seroclearance of HBsAg, a small proportion of patients may also be identified as occult HBV infection, in whom low levels of HBV DNA can be detected by sensitive PCR assays in the serum and/or liver samples, despite HBsAg seronegativity.

The idea of these classifications is based on the reciprocal relationship between age, viral replication, and histological activity [21, 22], with the host immune response further underlying the segmentation. However, the boundaries between phases are poorly defined. For example, it is difficult to identify the point at which a given patient enters the immune clearance phase , or the point at which the patient enters the residual phase. Unless there are robust immunologic markers that can reflect the interaction between the host immune response and the virus, it would be difficult to detect transitions between phases.

Another approach to classifying the clinical phases in the natural history of CHB is based on seromarker changes, including the seroclearance of HBeAg, HBV DNA, and HBsAg [23–26]. The advantage of this approach is that these milestones can be detected through repeated measurements of these seromarkers. These seromarkers are highly associated with the risk of subsequent disease progression, and are important indicators of the efficacy of antiviral treatment.

Serological milestones and traditional clinical phases of CHB may be correlated. HBeAg seroclearance/seroconversion mostly occurs during the immune clearance phase, while the seroclearance of HBV DNA might be regarded as the end of the immune clearance phase and the beginning of the residual phase . The seroclearance of HBsAg can be the end of the residual phase and also the beginning of occult HBV infection. The development of end-stage liver disease such as cirrhosis and HCC mostly occurs during the immune clearance phase, but may also be associated with the length of time spent in certain clinical phases.

One important question in the natural history of CHB is the chronological order of the serological milestones. The REVEAL-HBV study has demonstrated that HBeAg seroclearance occurs first, followed by the seroclearance of HBV DNA, and then the seroclearance of HBsAg [23]. Through repeated measurements of HBeAg, HBV DNA levels, and HBsAg serostatus, it has been shown that at HBeAg seroclearance , only 11 % of REVEAL participants had undetectable serum HBV DNA levels, while most participants still had high levels of HBV DNA (median, 10⁵ copies/mL). On the contrary, at HBV DNA seroclearance, almost all (98 %) participants had already cleared HBeAg prior to clearing HBV DNA from the serum. In addition, 96 % of participants had undetectable serum HBV DNA levels at the point of HBsAg seroclearance, leaving only 4 % patients with newly incident occult HBV infection [25]. This order of events is quite different than what occurs during antiviral the rapy, where patients’ HBV DNA levels are quickly suppressed to undetectable levels in HBeAg seropositive patients.

The courses of CHB that patients may experience are heterogeneous. The time spans of clinical phases and ages where milestone transitions occur can be very distinct among individual patients. For example, some patients are still HBeAg-seropositive at age 70, while others may develop HCC at only 30 years of age. Therefore, it is important to predict what will happen and when in individual infected persons.

The presence of hepatitis B e antigen (HBeAg) in the serum usually indicates active viral replication of HBV in hepatocytes. Loss of detectable HBeAg, together with the emergence of antibodies against HBeAg, has been a key end point in the development of new antiviral treatment. It has been shown that among chronic hepatitis B (CHB) carriers, the incidence of HCC was 3.6-fold higher in carriers who were HBeAg seropositive (1169.4 per 100,000 person-years) than in those who were HBeAg seronegative (324.3 per 100,000 person-years). The cumulative incidence of HCC was significantly higher among those who were HBeAg seropositive than those who were HBeAg seronegative (P < 0.001), and the relative risk of HCC was also higher in those with HBeAg seropositivity (60.2, 95 % CI = 35.5–102.1) than in those with HBeAg seronegativity (9.6, 95 % CI = 6.0–15.2) [11]. Moreover, in the natural history of CHB infection, the spontaneous or interferon alpha induced development of antibodies against HBeAg leads to improvement in clinical outcomes, such as lower risk for major liver complications [27] and a lower frequency of HCC development [28], suggesting that HBeAg is a useful marker for end stage liver diseases.

In children infected with HBV, the HBeAg seropositive rate in children under 15 years of age is 80–85 % [29]. Spontaneous HBeAg seroclearance rarely occurs before the age of 3. In most instances, HBeAg seroclearance occurs during adolescence and early adulthood [29]. In a long-term follow-up study from Taiwan, the annual HBeAg seroclearance rate was 4–5 % in children older than 3 years of age, and was <2 % in those under 3 years of age [30]. HBeAg seroclearance in children is determined by age and maternal HBsAg status. In another study, HBeAg seroclearance occurred in only 9.7 % of carrier infants under 3 years of age, and the HBeAg seroclearance rate was lower in infants whose mothers were seropositive for HBsAg than in those whose mothers had undetectable HBsAg (14.3 % vs. 35.3 %) [20]. Additionally, higher HBeAg seroclearance rates have been reported in children infected horizontally (44 %) than in those infected perinatally (24 %) [31].

The REVEAL-HBV study has shown an annual incidence rate of 61.6 per 1000 person-years for spontaneous HBeAg seroclearance (Table 12.1) [24]. Among individuals with high serum HBV DNA levels (≥10⁴ copies/mL) at study entry, the cumulative lifetime incidence of spontaneous HBeAg seroclearance at 40, 50, 60, 70, and 74 years of age was 38.8, 69.4, 81.9, 89.1, and 95.5 %, respectively (Table 12.1) [23].

In multivariate analyses adjusted for age, gender, serum ALT levels, smoking, alcohol consumption, precore mutation, basal core promoter, HBV genotype, serum HBsAg levels, and serum HBV DNA levels, serum HBV DNA levels remained as a significant determinant of HBeAg seroclearance. Compared to individuals with HBV DNA levels ≥10⁸ copies/mL, the multivariate-adjusted rate ratio (95 % CI) of HBeAg seroclearance was 1.89 (1.28–2.78) and 3.27 (2.01–5.32) for those with HBV DNA levels of 10⁶ to <10⁸, and 10⁴ to <10⁶ copies/mL, respectively (Table 12.1) [26]. In addition to serum HBV DNA levels, gender, serum ALT levels, precore mutation, and HBV genotype were also significantly associated with HBeAg seroclearance after multivariate adjustment [23, 26]. The multivariate-adjusted rate ratio (95 % CI) of spontaneous HBeAg seroclearance was 1.92 (1.29–2.85) for women compared to men; 2.11 (1.40–3.18) for baseline serum ALT levels of 45 or more compared to less than 45 U/L (45 U/L is the ULN for the REVEAL study); 1.66 (1.03–2.68) for the precore 1896 G/A mutant compared to wild type; and 3.06 (2.11–4.44) for HBV genotype B or B and C compared to genotype C (Table 12.1) [26].

A score-based prediction model and nomogram for HBeAg seroclearance was created by integrating the previously mentioned significant determinants. The prediction model’s total score ranged from 0 to 7 as shown in Fig. 12.3a [26]. The AUROCs (95 % CI) for predicting the 5- and 10-year probability of HBeAg seroclearance were 0.85 (0.80–0.90) and 0.78 (0.73–0.83), respectively. The 5- and 10-year probabilities of HBeAg seroclearance ranged from 0.08 to 0.72 and from 0.23 to 0.98, respectively, providing a well-performing and clinically applicable tool for clinicians.

Serum HBV DNA level is a marker of viral replication and antiviral treatment efficacy in CHB patients [32]. Previous studies have reported the association of serum HBV DNA levels with the development of liver cirrhosis and HCC. Follow-up studies have shown that among CHB carriers, the incidence of HCC increased with serum HBV DNA levels at study entry in a dose–response relationship , with rates ranging from 108 per 100,000 person-years for serum HBV DNA levels of less than 300 copies/mL to 1152 per 100,000 person-years for serum HBV DNA levels of 10⁶ copies/mL or greater [12]. The same dose–response relationship was also observed in the cumulative incidence of HCC, ranging from 1.3 % for serum HBV DNA levels of less than 300 copies/mL to 14.9 % for serum HBV DNA levels of 10⁶ copies/mL or greater. The same pattern was observed in liver cirrhosis , with cumulative incidence ranging from 4.5 % for carriers with serum HBV DNA levels of less than 300 copies/mL to 36.2 % for carriers with serum HBV DNA levels of 10⁶ copies/mL or more [13]. Compared to carriers with serum HBV DNA levels of <300 copies/mL, hazard ratios of developing HCC (adjusted for HBeAg, serum ALT levels, and liver cirrhosis) increased with serum HBV DNA levels [12]. A similar situation was also observed for the relative risk of liver cirrhosis progression, after adjustment for HBeAg and serum ALT levels [13]. These findings suggest that elevated serum HBV DNA levels are a prominent risk predictor of HCC independent of HBeAg status, serum ALT levels, and the presence of liver cirrhosis, and that progression to liver cirrhosis is strongly correlated with increasing serum HBV DNA levels, independent of HBeAg status and serum ALT level. Thus, the seroclearance of HBV DNA is an important milestone that signals an improved prognosis and lower rates of end-stage liver disease.

The REVEAL-HBV study showed an annual incidence rate of 30.1 per 1000 person-years for spontaneous HBV DNA seroclearance (Table 12.2) [26]. Among individuals with high serum HBV DNA levels (≥10⁴ copies/mL) at study entry, the cumulative lifetime incidence of spontaneous HBV DNA seroclearance at 40, 50, 60, 70, and 77 years of age was 10.1, 25.0, 38.8, 54.2, and 82.8 %, respectively (Table 12.2) [23].

In multivariate analyses adjusted for age, gender, serum ALT levels, smoking, alcohol consumption, precore mutation, basal core promoter, HBV genotype, serum HBsAg levels, and serum HBV DNA levels, serum HBsAg level was a significant determinant of HBV DNA seroclearance. Compared to individuals with serum HBsAg levels ≥10⁴ IU/mL at study entry, the multivariate-adjusted rate ratio (95 % CI) of HBV DNA seroclearance was 1.18 (0.61–2.27), 2.49 (1.31–4.74), and 6.18 (3.24–11.79) for carriers with serum HBsAg levels of 10³ to <10⁴, 10² to <10³, and <10² IU/mL, respectively (Table 12.2) [26]. In addition to serum HBsAg levels, age and gender were also significantly associated with HBV DNA seroclearance after multivariate adjustment. The multivariate-adjusted rate ratio (95 % CI) for HBV DNA seroclearance was 1.35 (1.00–1.82) for those ≥60 years compared to <60 years, and 1.37 (1.10–1.72) for women compared to men (Table 12.2) [26]. In a subset of carriers with high serum HBV DNA levels (≥10⁴ copies/mL) at study entry regardless serum HBsAg level, serum HBV DNA level and precore mutation were also significant determinants of HBV DNA seroclearance. Multivariate-adjusted rate ratios (95 % CI) were 3.45 (1.73–6.91) for carriers with serum HBV DNA levels of 10⁴ to <10⁵ compared to those with serum HBV DNA levels ≥10⁶ copies/mL, and 0.55 (0.36–0.85) for precore 1896 G/A mutant compared to wild type (Table 12.2) [23]. HBV genotype was associated with HBV DNA seroclearance in univariate analyses with a rate ratio (95 % CI) of 1.52 (1.21–1.91) for genotype C compared to HBV genotype B or B and C (Table 12.2) [26].

According to the multivariate analyses from previous study including serum HBsAg levels, the addition of HBV genotype, serum HBV DNA level, and precore mutation did not improve the predictability of HBV DNA seroclearance [26]. Therefore, a score-based prediction model and nomogram for HBV DNA seroclearance was created by integrating only age, gender, and serum HBsAg levels, as shown in Fig. 12.3b. The total score of the prediction model for HBV DNA seroclearance ranged from 0 to 8. Predictive accuracy of the model was measured by the AUROC. The AUROCs (95 % CI) for predicting the 5- and 10-year probability of HBV DNA seroclearance were 0.77 (0.72–0.82) and 0.73 (0.70–0.76), respectively. The 5- and 10-year probabilities of HBV DNA seroclearance ranged from 0.04 to 0.36 and from 0.14 to 0.80, respectively [26].

Serum HBV DNA levels play a critical role during the transition between milestones of CHB progression . Although previous studies identified HBV DNA as the most important predictor of HBeAg, HBV DNA, and HBsAg seroclearance, recent studies have shown that serum HBV DNA level was no longer a significant predictor of HBV DNA seroclearance after taking serum HBsAg levels into consideration [26], indicating that serum HBsAg levels are the strongest predictor of HBV DNA seroclearance.

Previous studies have shown that HBeAg-seropositive patients are at increased risk for clinical endpoints such as hepatocellular carcinoma , and during clinical management of these patients, HBeAg seroconversion is an important milestone [11, 33]. For HBeAg-seronegative patients, HBsAg seroclearance has been well-documented as the most important clinical and treatment end point, as it leads to an improved prognosis, and confers lower rates of HCC and other clinical consequences [24, 33–35]. Previous community-based studies in Taiwan and Alaska examined prognoses of those who spontaneously cleared HBsAg. In a study among Alaskan natives, the incidence rates of HCC were significantly decreased in those with HBsAg seroclearance (36.8 per 100,000 person-years), when compared to those who remained HBsAg-positive (195.7 per 100,000 person-years) [35]. In another study from Taiwan using repeated measurements of seromarkers , reaching HBsAg seroclearance during follow-up was indicative of significantly decreased risk for developing HCC in the future [24]. Therefore, elucidating the determinants of HBsAg seroclearance is crucial to the clinical management of individuals with chronic hepatitis B infection.

However, the spontaneous seroclearance of HBsAg is quite rare. In highly endemic Taiwan, the annual incidence rate of spontaneous HBsAg seroclearance among untreated individuals in the community was 2.26 % per year, or 22.6 per 1000 person-years [25]. In a study of 3087 community-based individuals from the REVEAL-HBV cohort , HBsAg seroclearance was associated with female gender, increasing age, increasing body mass index (BMI), ethnicity of mainland Chinese (versus Fukkienese), and decreasing serum HBV DNA levels [25]. After the introduction of quantitative HBsAg (qHBsAg) as a potential marker for immune response, the determinants of HBsAg seroclearance were reanalyzed [36].

In this study, which was further limited to 2491 HBeAg-seronegative individuals >30 years old, the cumulative lifetime incidence of spontaneous HBsAg seroclearance at ages 40, 50, 60, 70, and 77 among those with detectable serum HBV DNA (≥57 IU/mL) was 3.0, 14.4, 26.5, 42.6, and 62.1 %, respectively. On the other hand, the cumulative lifetime incidence of HBsAg seroclearance at ages 40, 50, 60, 70, and 77 among those with undetectable (<57 IU/mL) HBV DNA was 31.5, 56.7, 74.2, 89.1, and 98.8, respectively. In multivariate analyses , serum HBsAg levels were the strongest predictor of spontaneous HBsAg seroclearance. While serum HBV DNA level was still significant, its rate ratios decreased after adjustment for serum HBsAg levels. Compared to baseline serum HBsAg levels ≥1000 IU/mL, the multivariate-adjusted rate ratio (95 % CI) of spontaneous HBsAg seroclearance was 3.55 (2.51–5.02) and 10.96 (7.92–15.16), respectively, for those with serum HBsAg levels of 100–999 and <100 IU/mL. These results suggested that both serum HBsAg and HBVDNA levels should be considered during monitoring of chronic hepatitis B, as they provide complementary information. Other significant predictors included increasing age and BMI (Table 12.3) [36].

Using this model, a score-based prediction model and nomogram were developed, assigning integer scores to each predictor as shown in Fig. 12.3c. Using each individual’s combined score, a 5 and 10 year probability of HBsAg seroclearance was calculated using equations derived from Cox Proportional Hazards models. This 30-point scale combining age, BMI, HBV DNA levels, and HBsAg levels, was able to predict 5 and 10 year probabilities of spontaneous seroclearance with AUROC’s of 0.89 and 0.84, respectively. Therefore, this model was able to accurately estimate the probability of HBsAg seroclearance for different clinical profiles in the community, and showed that the addition of serum HBsAg levels to current HBV DNA-based models significantly improves the predictability of HBsAg seroclearance among genotype B and C HBeAg-seronegative individuals (Table 12.3) [36]. These results have also recently been externally validated among a hospital based cohort of 1934 untreated patients, in which the model still performed adequately well, and was well calibrated, even among patients with more severe disease [37].

Additional studies among a cohort of children followed-up into adolescence also found that children with serum HBsAg levels <1000 IU/mL had a much greater chance of clearing HBsAg (HR [95 % CI] = 5.23 [2.77–9.85]) [38]. In addition, there was a significant association between HBsAg seroclearance and maternal serostatus of HBsAg and HBeAg. In conclusion, the determinants of HBsAg seroclearance have been well established in the community, and will provide important direction and information for the clarification of prediction of HBsAg seroclearance among treated individuals.

In the natural course of chronic hepatitis B infection, HBeAg-seronegative carriers represent a large majority of infected individuals. The severity of disease among HBeAg-seronegative individuals varies widely, and can include those who are either inactive, or active carriers [6, 33]. Inactive carriers are defined as HBeAg-seronegative individuals with serum HBV DNA levels <10,000 copies/mL (2000 IU/mL) and persistently normal ALT for 1 year, while active carriers are individuals with serum HBV DNA levels ≥10,000 copies/mL (2000 IU/mL) with persistently or intermittently abnormal ALT [33]. Previous studies have shown inactive carriers to have significantly improved survival, which is comparable with that of noninfected individuals [39]. In addition, in previous studies from the REVEAL-HBV cohort, inactive carriers also have significantly decreased risk for hard outcomes such as liver cirrhosis and hepatocellular carcinoma [12, 13, 40]. Thus, differentiating between active chronic hepatitis and inactive carriers is clinically meaningful, as it would allow the identification of a lower-risk population in need of less stringent follow-up, while, on the other hand, an earlier diagnosis of active hepatitis could lead to earlier initiation of antiviral therapy. However, accurate identification of active carriers is difficult, as ALT levels fluctuate and can often be affected by environmental exposures or nonviral factors. Moreover, studies examining inactive or active hepatitis among the community are rare, and the factors that can accurately differentiate the two are still relatively unknown.

Tohme et al. examined determinants and risk factors for reactivation of hepatitis B among a chronically infected community-based cohort of 414 Alaskan Native Persons who already had inactive hepatitis, with viral loads <2000 IU/mL, and persistently normal ALT for 1 year [41]. This study included individuals with viral genotypes A, B, C, D, and F, and reactivation was defined as HBV DNA ≥2000 IU/mL and ALT ≥40 U/L (Table 12.4). Over 2984 person-years of follow-up, 36 cases of reactivation occurred, for an annual incidence rate of 12 per 1000 person-years. In multivariate analyses, compared to individuals between 18 and 29 years old, those who were 30–39 or 40–49 years old had adjusted hazard ratios (HR [95 % CI]) of reactivation of 0.34 (0.12–90) and 0.20 (0.05–0.70), respectively. In addition, males, those with HBV DNA levels of 1000–1999 IU/mL (compared to HBV DNA <29 IU/mL), and genotype (compared to genotype) were significant predictors of hepatitis B reactivation, with adjusted hazard ratios (95 % CI) of 2.41 (1.17–4.96), 7.61 (2.81–20.6), and 6.08 (1.32–28.0), respectively [41].