Most patients in the immune clearance phase are asymptomatic and have mild to moderate elevation in ALT levels and hepatitis activity, so called HBe Ag positive chronic hepatitis B (CHB). However, the clinical course may be punctuated by spontaneous acute hepatitis flare, defined as an abrupt elevation of ALT >5 times the upper limit of normal (ULN). These acute hepatitis flares are considered to be the results of HLA-class I antigen-restricted, CTL mediated immune response against HBV antigen(s) and its downstream apoptotic mechanisms [32]. The reasons for spontaneous acute hepatitis flares are not clear but are likely explained by subtle changes in immunological controls of viral replication. Several studies have found that acute hepatitis flares are often preceded by a sudden increase in serum levels of HBV DNA [33, 34], HBeAg and HBeAg-specific immune complexes [34], and enhanced T-cell response to HBcAg and HBeAg [35]. These results suggest that increases in viral replication, accumulation of nucleocapsid proteins in serum and hepatocytes, and the subsequent immune res ponse play an important role in initiating acute hepatitis flares in chronic HBV infection [34]. Histological evidence of lobular hepatitis superimposed upon the changes of chronic viral hepatitis is frequently observed [36]. IgM anti-HBc is positive in 14.4 % of patients during acute flares, but generally in lower titers than in acute HBV infection [37]. As HBeAg seroconversion is often preceded or accompanied by a transient ALT flare, it is believed that hepatitis flares are the results of the host attempt to clear the virus by the immune response. However, not all acute hepatitis flares lead to HBeAg seroconversion and HBV DNA clearance from serum, a phenomenon termed as “ineffective or abortive immune clearance” [32]. In this context, the patients may experience repeated episodes of acute hepatitis flares, which can account for increased risk of HBV-related cirrhosis.

The annual rate of acute hepatitis flare in patients with HBeAg positive CHB was as high as 28.6 % in an early hospital-based study from Taiwan [38]. However, in another study that followed up asymptomatic patients beginning at the immune tolerance phase through HBeAg seroconversion, t he overall incidence of acute hepatitis flare was 28.8 % during immune clearance phase (mean 3.7 years), with a n annual rate of 7.8 % only [39]. Most acute hepatitis flares are asymptomatic, but around 20 % of patients present with symptoms of overt acute hepatitis [37], and approximately 2–3 % may be complicated with hepatic decompensation [40]. One recent report from Taiwan found that a serum HBV DNA level ≥1.55 × 10⁹ copies/mL at the onset of acute flare can predict hepatic decompensation [41]. In HBV high-prevalence areas, acute hepatitis flares of chronic HBV infection is the most important etiology of acute hepatitis and fulminant hepatitis in adults [42, 43].

The immune clearance phase has a variable duration and often lasts for many years until HBeAg seroconversion occurs. HBeAg seroconversion is frequently preceded by ALT elevation, followed by a marked reduction o f serum HBV DNA levels that can only be detected by sensitive polymerase chain reaction (PCR) assay, decline of serum HBsAg level, ALT normalization and resolution of liver necroinflammation [32, 36]. However, abnormal ALT levels and high-level HBV DNA persist at the time of HBeAg seroconversion in about 5 % of patients [44]. These patients progress directly from HBeAg positive chronic hepatitis to HBeAg negative chronic hepatitis.

The average annual incidence of HBeAg seroconversion is 10 % (range, 2–15 %), depending on factors such as ethnicity, mode of transmission, age, ALT levels, histological activities and HBV genotype. HBeAg seroconversion is m uch more delayed in children with HBeAg positive carrier mothers than in children with HBeAg negative carrier mothers or children with non-carrier mothers [45]. Different mode of HBV transmission accounts for the much lower HBeAg positivity rates in black Africans of childbearing age than in women in the Far East [5, 6]. A higher HBeAg seroconversion rate has been reported in non-Asian children with horizontal transmission than Asian children with vertical transmission [46]. In Taiwan, the annual rate of HBeAg seroconversion is <2 % in children ≤3 years of age and 4–5 % in older children, so that around 85 % of children still remain HBeAg positive by age 15 [47]. The likelihood of HBeAg seroclearance correlates positively with ALT levels: HBeAg seroclearance rates at 18-months of follow-up are 0, 3–8, 17, and 59–70 %, respectively, if baseline ALT levels increase over <1, 1–2.5, 2.5–5, and >5 times ULN [32]. In patients with acute hepatitis flare, 72 % undergo HBeAg seroclearance within 3 months if serum α-fetoprotein (AFP) levels >100 ng/mL, compared to only 18 % of those with AFP <100 ng/ml [48]. Serum HBV DNA levels ≤7 log₁₀ copies/mL during acute hepatitis flare also can predict HBeAg seroconversion within 6 months [49]. The likelihood of HBeAg seroconversion also correlates with histological activ ities: the 5-year cumulative probabilities of HBeAg seroconversion is >65 % in patients with high necroinflammatory (interface or lobular) activities, compared to <25 % in those with low necroinflammatory activities [48]. HBeAg seroclearance may occur within 3 months in two-thirds of the patients with bridging hepatic necrosis [48]. In Eastern countries, patients infected with genotype B HBV seroconvert earlier and more frequently than those with genotype C HBV [50–52]. In Western countries, HBeAg seroconversion is similar in genotypes A, B, D, and F HBV infection but much slower in genotype C HBV infection [53, 54]. In Alaska native carriers, the median age of HBeAg seroclearance is <20 years in patients with genotypes A, B, D, and F HBV, but >40 years in patients with genotype C HBV [53]. Interestingly, HBeAg seroconversion is more frequently preceded by ALT flares >5 times ULN in genotype C HBV infection than in genotype B HBV infection, suggesting that a more vigorous immune-mediated hepatocytolysis may be needed to achieve HBeAg seroconversion in genotype C HBV infection [52].

In Taiwan, HBeAg seroconversion occurs at a median (interquartile range) age of 32 (26–36) years, with 90 % before age of 40 [55]. In accordance with these data, the prevalence of serum HBeAg declines remarkably from 85 % in children (age < 15 years) [46] to 5–10 % in adults ove r 40 years of age [5]. These findings suggest that HBeAg seroconversion most often occurs between 15 and 40 years of age in perinatally acquired chronic HBV infection. Persist ence of HBeAg over 40 years of age is rare and can be considered as “delayed” HBeAg seroconversion [55].

In some patients, the immune clearance phase may last for many years without HBeAg seroconversion. A prolonged HBeAg positive phase is associated with increased risk of disease progression. A recent cohor t study from Taiwan demonstrated that the risk of progression to cirrhosis increased with increasing age of HBeAg seroconversion, with a hazard ratio of 3.8 per decade increase in age of HBeAg seroconversion [39]. In particular, patients with HBeAg seroconversion after 40 years of ages were associated with a remarkably high risk of progression to cirrhosis [55, 56].

Several other studies also showed that persistence of serum HBeAg was associated with an increased risk for progression to cirrhosis, HCC development and liver related mortality [57–59]. For instance, in one study from Taiwan that followed up 233 untreated patients with HBeAg positive CHB for a median of 6.8 years, the annual incidence of cirrhosis and HCC development was significantly higher in 147 patients with persistent HBeAg (3.7 and 1.6 %, respectively) than in 86 patients who underwent HBeAg seroconversion (1.8 and 0.4 %, respectively) [58].

In a study from Alaska, 109 (20 %) of 541 seroconverters developed HBeAg reversion, which was frequently accompanied by hepatitis flare, and HBeAg tended to fluctuate between seroconversion and reversion [66]. HBeAg reversion, however, is much uncommon in other studies. In two studies from Taiwan, one involving 283 patients with HBeAg positive hepatitis and another involving 240 HBeAg positive carriers with normal baseline ALT, HBeAg reversion following initial HBeAg seroconversion occurred in 12 (4.2 %) and 7 (2.9 %) patients during a mean follow-up of 8.6 years and 6.8 years, respectively [39, 44]. In another study from Italy, only one (1.6 %) of 61 seroconverters had HBeAg reversion during a mean follow-up of 22.8 years [59]. Despite the low frequency, HBeAg reversion is significantly associated with increased risk of progressi on to cirrhosis as well as development of HCC [44, 66].

The majority of patients with reactivation of hepatitis B are negative for HBeAg [39, 44] and have “HBeAg negative CHB .” Patients with HBeAg negative CHB are usually older than patients with HBeAg positive CHB and are more likely to have advanced fibrosis and cirrhosis at the time of their first presentation. Serum levels of HBsAg are lower in HBeAg negative CHB than in HBeAg positive CHB by about 0.5–1 log₁₀ IU/mL (3.0–3.5 vs. 3.5–4.5) [26–28]. Serum HBV DNA levels also tend to be lower (4–8 log₁₀ copies/mL), compared to HBeAg positive CHB (6–10 log₁₀ copies/mL). However, many patients with HBeAg negative CHB have wide fluctuations in both HBV DNA and serum ALT levels. Episodes of hepatitis flare are frequently seen, with a rate of about 1/3–1/2 of that in HBeAg positive counterparts [38]. Spontaneous sustained remission of disease activity is rare [67].

The prevalence of HBeAg negati ve case in patients with CHB varies widely in different geographical areas: 80–90 % in the Mediterranean basin, 30–50 % in Taiwan and Hong Kong, but less than 10 % in the USA and Northern Europe [68]. This difference may be in part attributed to the different HBV genotype distribution: precore mutant is frequently detected in genotype D (the main genotype in the Mediterranean basin) and genotypes B and C infection (the predominant genotypes in East Asia), but rarely detected in genotype A infection (the main genotype in the USA and Northern Euro pe). The prevalence of HBeAg negative CHB has been increasing over the last few decades as a result of aging of the HBV-infected population and the effective prevention measures restricting new HBV infections. HBeAg negative CHB has become much more common than HBeAg positive CHB in many countries of the world nowadays.

However, the incidence of HBeAg negative CHB among HBeAg negative carriers remains largely unknown. This issue has been addressed in a few prospective studies that followed up the natural course following spontaneous HBeAg seroconversion. In two studies from Taiwan, the annual rate of HBeAg negative CHB was 2–3 % with a cumulative incidence of 25 % at 16 years, but hepatitis B reactivation typically occurred within the first 5–10 years [39, 44]. However, in another study from Italy, only 9 (14.8 %) of 61 seroconverters developed HBeAg negative CHB during a mean follow-up of 22.8 years (annual rate of 0.6 %) [59]. In another Italian study involving pediatric patients, the rate is even lower: only 4 (6.3 %) of 64 patients developed HBeAg negative CHB during a mean period of 15 years [69]. These differences can be explained by the finding that age of HBeAg s eroconversion is an important factor for HBV reactivation [56].

The incidence of hepatitis B reactivation among incidentally identified inactive carriers also varies in different geographical areas. In a study of 1241 inactive carriers from Taiwan, 211 (17.0 %) developed HBeAg negative CHB during a mean follow-up of 12.3 years, with the annual incidence of 1.4 % and the cumulative incidence of 20.2 % at 20 years [70]. Reactivation of hepatitis B occurred much more commonly during the first 5–10 years and became extremely rare after 20 years [70]. In other studies that enrolled a relatively small number of inactive carriers, the annual incidence of reactivation of hepatitis B varied from 0.4 % in Italy [71] and Greece [72] to 2.1 % in Japan [73]. However, in a more recent study of 85 inactive carriers from Greece, the cumulative incidence of HBeAg negative CHB was 24 % at 4 years [74]. The reason for such a high rate of HBV reactivation remains unclear.

Factors predictive for hepatitis B reactivation following HBeAg seroconversion include male gender [75], genotype C HBV (>genotype B) [75], genotype D HBV (>genotype A) [54], HBV-DNA levels >2000 IU/mL [76] or ≥10⁵ copies/mL [73] and HBV DNA >10⁴ copies/mL at 1 year after HBeAg seroconversion [77]. Age of HBeAg seroconversion <30 years is associated with a particularly low incidence of HBV reactivation [75]. In addition, ALT levels >5 × ULN during the immune clearance phase and age of HBeAg seroconversion >40 years are also associated with increased risk of hepatitis B reactivation [56, 75]. The latter findings suggest that HBV is more likely to reactivate if more vigorous immune-mediated hepatocytolysis or a more prolonged immune clearance phase is needed to clear the virus.

Recent studies have shown improved diagnostic a ccuracy by combined HBsAg and HBV DNA measurements to predict hepatitis B reactivation in inactive carriers: HBsAg >1000 IU/mL and HBV-DNA >200 IU/mL [78], HBsAg levels >850 IU/mL and HBV DNA >850 IU/mL [79], or HBsAg levels >1000 IU/mL in HBeAg negative carriers with HBV DNA <2000 IU/mL [80]. The latter finding reported from Taiwan of genotypes B and C patients is in keeping with the results of an earlier study from Italy of genotype D patients, in which the combined single point quantification of HBsAg <1000 IU/mL and HBV-DNA ≤2000 IU/mL allows the identification of inactive carriers with a very high diagnostic accuracy (94.3 %) [81].

Patients with hepatitis B reactivation have a 20-fold increased risk of progression to cirrhosis as compared with those without [82]. The annual rates of progression to cirrhosis and HCC were 2–3 % and 0.5 %, respectively, in patients with hepatitis B reactivation, significantly higher than 0.1 % and 0.2 %, respectively, in those wit h sustained remission of hepatitis [44, 82]. Notably, among patients with hepatitis B r eactivation, the incidence of cirrhosis is significantly higher in males and in those with age of reactivation older than 40 years [82]

During the low replicative inactive phase, serum HBsAg may disappear (HBsAg seroclearance) spontaneously. Short-term studies showed that the annual incidence of HBsAg seroclearance was 1–2 % in Caucasian carriers, and even lower (0.1–0.8 %) in carriers from the high-prevalence areas [83, 84]. However, a recent long-term follow-up study from Taiwan showed that the incidence of HBsAg seroclearance was appreciably high with an overall annual incidence of 1.2 %, being higher (1.8 %) in those >50 years than in those <30 years (0.8 %), and a cumulative incidence of 8 % at 10 years, increasing disproportionately to 25 % at 20 years, and 45 % at 25 years of follow-up [85].

Factors significantly associated with HBsAg seroclearance include older age [54, 64, 66, 83, 85–87], normal ALT levels [85, 86], HBeAg negativity [66, 83, 86], low viral load (<300 copies/mL) [87], genotype A HBV (> genotype D) [54], or genotype B HBV (> genotype C) infection [88], sustained remission of hepatitis [85], presence of cirrhosis [83] and HCV superinfection [89]. Among these, advanced age is the most constant and important predictor for HBsAg seroclearance [90]. The annual incidence of HBsAg seroclearance varies among different series, but correlates significantly with the mean or median age of patients at enrollment of each cohort [90]. The median age of HBsAg seroclearce in three large cohorts from Taiwan [85], Hong Kong [91] and Japan [92] is approximately 50 years (range, 48–51). Given that the mean or median age of HBeAg seroconversion in Asian adult carriers ranges from 30 to 35 years, it can be expected that sustained remission of hepatitis for a mean of 15 years after HBeAg seroconversion is required to achieve subsequent HBsAg seroclearance. Of note, HBsAg seroclearance can occur sometimes in carrier children, albeit at a relative low rate (0.58 % per year during a mean follow-up of 20.6 years), usually after age 15 (mean, 17.7 ± 7.8; range, 4.1–33.0) and is more common in those with non-carrier mother [93]. Interestingly, in one case–control study, carriers with HBsAg seroclearance had significantly higher body mass index and higher degrees of fatty liver than those without [94]. Furthermore, the mean age of HBsAg seroclearance is significantly younger in patients with fatty liver than in those without (48.7 years vs. 53.4 years) [95]. Notably, in two large cohort studies, fatty liver [96] and obesity [87] were independent factors significantly associated with HBsAg seroclearance. The underlying mechanism by which fatty liver enhances HBsAg seroclearance remains unclear.

More recent studies have investigated both absolute and serial changes of serum HBsAg levels in predicting HBsAg seroclearance [78, 97–101], as summarized in Table 11.2. Of these, two Asian studies used an HBsAg level <100 IU/mL as a remote (6–10 years) predictor of HBsAg seroclearance [97, 98]. For short-term prediction, a study from Taiwan has shown that HBsAg level <200 IU/mL plus >1 log₁₀ IU/mL reduction in preceding 2 years can predict HBsAg seroclearance at 1 and 3 years [100]. Another study from Hong Kong has also shown that HBsAg <200 IU/mL or an annual reduction of >0.5 log₁₀ IU/mL is predictive for HBsAg seroclearance within 3 years [101].

Only 17 % have detectable antibody against HBsAg (anti-HBs) within 1 year after HBsAg seroclearance, but the rate of anti-HBs seroconversion increases to 56 % after 5 years and 76 % after 10 years. Virtually all test negative for HBV DNA by hybridization assays after HBsAg seroclearance, but in some HBV DNA still can be detected by PCR-based assays. The persistence of low-level viremia after HBsAg seroclearance might be a potential source of HBV transmission through blood transfusion or transplantation and account for HBV reactivation with chemotherapy or immunosuppression. In a recent investigation using commercially available, ultrasensitive real-time PCR assay, HBV viremia was detectable in 24 % within 1 year after HBsAg seroclearance, and low-level HBV viremia persisted in ∼15 % up to >10 years after HBsAg seroclearance [102]. Serum levels of HBV DNA all are below the sensitivity of hybridization assays (<100 IU/mL in 86 % and 121–2770 IU/mL in 14 %).

Despite the extremely low viremic states, 5–18 % of patients have abnormal ALT levels after HBsAg seroclearance. Non-HBV-related etiologies of abnormal ALT levels can be identified in 75–100 % of such cases, with fatty liver, alcoholism and herbal medicine administration being the most common [90]. In addition, HCV might have displaced HBV to cause continuing ALT elevation and hepatitis activities [103].

Most patients with liver histological assessment after HBsAg seroclearance have only mild necroinflammation and no significant fibrosis. Immunostaining for HBsAg and HBcAg in liver is negative in all patients; however, all patients tested still harbor HBV inside the liver, mainly in the form of cccDNA, up to 4 years after HBsAg seroclearance [88], albeit at a very low replicative level and in a transcriptio nally inactive phase.

In an early study in 55 patients with spontaneous HBsAg seroclearance from Taiwan by Huo et al. [104], 32.7 % developed serious complications, including HCC, cirrhosis , and hepatic failure during a mean follow-up of 23 months. This study probably overestimated the frequency with which complication occurs, as it included 20 patients who had hepatitis C virus (HCV) or hepatitis D virus (HDV) coinfection. In subsequent studies that enrolled a large series of patients from Taiwan and Japan, virtually none of non-cirrhotic patients without HCV or HDV superinfection developed HCC, hepatic decompensation, or liver related death during a mean follow-up of 5 years [92, 105], as summarized in Table 11.3. HBsAg seroclearance usually confers excellent long-term prognosis, provided that HBsAg loss occurred in the absence of concurrent viral infections, and preceded the development of cirrhosis. However, in patients who have preexisting cirrhosis or HCV or HDV superinfection, clinical outcomes of disease progression may still occur [92, 105–108]. A recent report from Hong Kong suggested that cumulative risk for HCC was higher in patients with HBsAg seroclearance at age ≥50 years compared with those with HBsAg seroclearance at age <50 [91]. However, the majority of their patients who developed HCC (6 out of 7) after HBsAg seroclearance had ultrasonographic evidence of cirrhosis before or at the time of HBsAg seroclearance. Of note, the mean age of HBsAg seroclearance in the series of Huo et al. [104] is also appreciably high (see Table 11.3). It is highly suspected that patients who achieved HBsAg seroclearance at older age may be more likely to have undiagnosed cirrhosis and hence remain at risk for HCC. Older age of HBsAg seroclearance per se cannot be considered as an independent risk factor for HCC development after HBsAg seroclearance.

A more recent prospective population-based cohort study in 1271 Alaska native persons with chronic HBV infection followed for an average of 19.6 years showed that the incidence of HCC after HBsAg seroclearance was 36.8 per 100,000 per year (95 % CI 13.5–80.0), which was significantly lower than that in those who remained HBsAg positive (195.7 per 100,000 per year [95 % CI 141.1–264.5]; P < 0.001) [109]. This study is the first to show a significant reduction in the risk of developing HCC after HBsAg seroclearance.

Worldwide, approximately 5–20 % of HBsAg carriers are found to be anti-HCV positive. In HBV endemic areas, HCV superinfection in the setting of chronic HBV infection is the most common scenario of HBV and HCV dual infection. Acute HCV infection in HBsAg carriers with serum HBeAg and HBV DNA may result in only transient HCV infection. In contrast, most acute HCV superinfection in HBsAg carriers without serum HBeAg and HBV DNA progress to persistent HCV infection [111]. These findings suggest that the presence of underlying active HBV replication may interfere with HCV replication and thereby inhibit the persistence of HCV infection.