Chronic hepatitis B in children after e antigen seroclearance: Final report of a 29-year longitudinal study


  • Potential conflict of interest: Nothing to report.


Chronic hepatitis B is usually a benign disease in Caucasian children; however, the long-term prognosis remains unsettled. This report describes the results of a 29-year longitudinal study including 99 white children with chronic hepatitis B, mainly acquired horizontally: 91 were hepatitis B e antigen (HBeAg) positive (4 had cirrhosis), and 8 were HBeAg negative at presentation. Of the 91 HBeAg-positive children, 89 underwent HBeAg seroconversion after a mean period of 5.2 ± 4.0 years and were included in the study. Of the 85 children without cirrhosis, one had HBeAg-negative hepatitis and the other 84 became inactive carriers. During a mean follow-up of 14.5 ± 6.1 years after HBeAg seroclearance, 4 carriers experienced reactivation, and 3 of them had HBeAg-negative hepatitis at the last follow-up. Of the 8 initially HBeAg-negative children, 2 had HBeAg-negative hepatitis, and 6 were inactive carriers. Of the 4 children with cirrhosis, 2 had hepatocellular carcinoma (HCC) and remained alive and 2 lost the histological features of cirrhosis in adulthood. Two patients with HBeAg-negative hepatitis and 1 with cirrhosis had experienced drug abuse. At the end of follow-up, 15 of the 89 initially HBeAg-positive patients and 2 of 8 initially HBeAg-negative children had cleared hepatitis B surface antigen. In conclusion, the overall prognosis for chronic hepatitis B in horizontally infected Caucasian children is favorable; however, some patients progress to HCC and HBeAg-negative hepatitis. Long-term monitoring is important, as is counseling on cofactors of liver damage, such as alcohol and drug abuse. (HEPATOLOGY 2006;43:556–562.)

Chronic hepatitis B virus (HBV) infection is a leading cause of cirrhosis and hepatocellular carcinoma (HCC) worldwide.1–4 In endemic areas, infection is often acquired perinatally (Asia, Pacific region)5, 6 or in early childhood (Mediterranean area, Alaska, Africa).7, 8 Chronic HBV infection acquired at birth is characterized by an initial “tolerance phase,” with minimal liver damage and active virus replication, followed by an immune clearance phase with increased alanine aminotransferase (ALT) levels. After hepatitis B e antigen (HBeAg) to antibody (anti-HBe) seroconversion, most patients enter a phase of low replication, also referred to as the “inactive hepatitis B surface antigen (HBsAg) carrier state.” Epidemiological and clinical studies in children and adults suggest that HBeAg to anti-HBe seroconversion occurs in the third or fourth decade of life in a substantial proportion of perinatally infected HBV carriers.9–12 Delayed seroconversion seems to be a risk factor for liver disease reactivation and virus replication, which in turn could favor cirrhosis and HCC.11

In the Mediterranean area, horizontally acquired HBV infection runs a more accelerated course, with features of immune clearance at first observation.13, 14 A longitudinal study begun in Padua (Northern Italy) in the mid-1970s showed that spontaneous HBeAg to anti-HBe seroconversion occurred at a mean annual rate of 16% to 19% during the first 7 years of observation, and most children had become asymptomatic HBsAg carriers by the time they were 20 years old.15, 16 This condition remained stable in adolescence, but the final outcome in adulthood remained doubtful for various reasons. Using sensitive serological techniques, low viremia levels could be detected in adult carriers several years after anti-HBe seroconversion and ALT normalization.2, 17–19 The significance of this finding in relation to the outcome of infection and disease is still uncertain. HBeAg-negative hepatitis B is the most frequent type of chronic hepatitis in Italian adults20; it is usually associated with mutations in the precore region of HBV in subjects with genotype D, but the dynamics of mutant selection have still to be clarified.2–23 Adult lifestyle can also expose HBV carriers to different types of behavioral risks, such as alcohol24 and drug use, or comorbid conditions such as overweight25, 26 or co-infections with hepatotropic viruses,27 that could affect the stability of disease remission. The aim of this study was to provide further information on the anti-HBe positive phase of chronic hepatitis B acquired early in life. For this purpose, we extended the monitoring of our pediatric cohort up to the third and fourth decades of life.


HBV, hepatitis B virus; HCC, hepatocellular carcinoma; ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; anti-HBe, antibody to HBeAg; HBsAg, hepatitis B surface antigen; anti-HBs, antibody to HBsAg; IFN, interferon; PCR, polymerase chain reaction; HAI, histological activity index.

Patients and Methods


Between 1975 and 1985, 109 consecutive children, aged 1 to 16 years, were referred to the Department of Infectious Diseases at Padova Hospital with a diagnosis of chronic HBV infection, based on the presence of HBsAg in the serum for 6 months or longer. This pediatric population was homogeneous and consisted of 98 HBeAg-positive and 11 HBeAg-negative, anti-HBe–positive well-nourished Caucasian children belonging almost entirely to the population of the Veneto region, and living with their family of origin. None had any underlying systemic diseases or were co-infected with other hepatotropic viruses.

Of the 98 HBeAg-positive children, 7 were lost to follow-up within a year of their diagnosis, including 1 girl with normal ALT and no liver biopsy and 6 children with ALT abnormalities and histological evidence of mild-to-moderate hepatitis (3 of whom had cleared HBeAg before they dropped out). Ninety-one children entered the long-term follow-up: at their last visit, 2 teenagers were still HBeAg positive with slightly abnormal ALT, whereas 89 patients had seroconverted to anti-HBe and formed the basis of this study. As previously reported, all of the children enrolled in the study underwent liver biopsy, showing features of cirrhosis in 4 patients, and of minimal to severe hepatitis in the others. Twenty-one children were treated between 1976 and 1981: 4 with prednisone for 12 to 24 months, 9 with a 12-month course of levamisole, and one with a thymic derivative; 7 other patients who were still HBeAg positive in the late 1980s were enrolled in therapeutic trials with recombinant interferon (IFN) alpha at a dose of 3 to 5 MU/m2 thrice weekly for 6 months. The IFN-treated children had low ALT values at baseline and failed to respond to the treatment; they remained HBeAg positive for an average period of 7.8 years after stopping IFN.

The anti-HBe–positive series included 3 children with persistently normal ALT, who did not join the long-term follow-up, and 8 patients with abnormal ALT and biopsy-proven mild-to-moderate hepatitis, two of them born of a mother known to be HBsAg positive at delivery.

The natural history of infection and liver disease in HBeAg-positive and negative children has been investigated over the years,16 the last update being published in 1998 as part of a multicenter Italian and Spanish study.28 The current study was approved by the Padova Hospital Ethics Committee.

Definition of Clinical Events After Anti-HBe Seroconversion.

After anti-HBe seroconversion, patients were defined as inactive HBsAg carriers when ALT was persistently normal, former histology and ultrasound excluded cirrhosis, and HBV DNA was persistently negative according to non–polymerase chain reaction (PCR)-based tests (cutoff, 105-106 copies/mL). HBeAg-negative hepatitis was diagnosed when abnormal ALT and HBV DNA seropositivity by unamplified assays persisted continuously or recurred intermittently after HBeAg clearance. Occasional rises in ALT up to thrice the upper normal value, with no changes in viremia levels, in inactive HBsAg carriers were classified as sporadic ALT abnormalities. Reactivation was defined as an abrupt increase in ALT (more than twice the norm) with the reappearance of HBV DNA by unamplified assay, with or without seroreversion to HBeAg.

Management of the Cohort.

Inactive HBsAg carriers were invited to a check-up every 12 months, whereas patients who maintained or developed abnormal ALT were seen more frequently (every 4-6 months) as clinically required. Adult patients continued their follow-up at the outpatient clinic of the 5th Medical Clinic at Padova Hospital. At each routine visit, the clinical history was updated, particularly as regards self-reported alcohol intake in teenagers and adults. Weekly or daily ethanol consumption was calculated from the drinking frequency and the dose of alcohol in each kind of alcoholic beverages (one alcohol unit = approximately 12.5 g = 1 standard glass of wine, beer, or spirits). Patients who consumed less than 1 unit/week were classified as non-drinkers, and drinkers were considered as moderate–heavy when the mean number of units/week was greater than 21.29 During routine visits, height and weight were recorded and the body mass index (BMI) was calculated using the standard charts for adults developed by the Centers of Disease Control: overweight was defined as a BMI between 25 and 29.9 and obesity as a BMI higher than 30. Alpha-fetoprotein screening was included in the routine liver biochemical tests. Hepatic ultrasonography was initially recommended every year in all cases; then the lack of significant lesions over a long time span in cases with normal ALT levels suggested that a check-up every 2 years would suffice in inactive carriers. Post-seroconversion liver biopsies were routinely performed early in the study. When it became clear that ALT normalization was usually associated with an improvement in liver histology, biopsies were only required for patients with sustained biochemical abnormalities.

In recent years, as soon as the children became adolescents and young adults, counseling on the potential role of alcohol in maintaining or exacerbating liver disease was offered during routine visits.

Laboratory Procedures.

Liver biochemistry and viral markers (HBsAg/anti-HBs and HBeAg/anti-HBe, hepatitis C and D virus antibodies) were monitored by conventional serological assays. Serum HBV DNA was first investigated using home-made dot-blot hybridization and a commercial hybridization test (Murex Diagnostics, Dartford, UK) with cutoff values of 1 pg and 5 pg/mL, respectively. Then the Digene Hybrid Capture System assay (Digene, Beltsville, MD) was used, with a lower detection limit of 0.5 pg (1.4 × 105 copies/mL). For research purposes, sera obtained during the last 1 or 2 visits were tested with a commercial PCR assay (Cobas Amplicor Monitor, Roche, Basel, Switzerland) with a cutoff of 200 copies/mL. HBV genotyping was done by direct sequencing of PCR products of the Pre-S region.30 The PCR products were sequenced using an ABI 373 automated sequencer and PRISM Dye terminator cycle sequencing (Perkin Elmer; Foster City, CA), according to the manufacturer's instructions. Finally, HBV was genotyped by phylogenetic comparison with database sequences representing genotypes A through H.

Liver biopsy was done using the Menghini technique subject to the informed consent of parents or the patient. Liver histology was reviewed by the same pathologist (M.G.), using Ishak's scoring system.31 The final grade of activity was assigned as follows: minimal when the histological activity index (HAI) was ≤3; mild for HAI 4 through 6; moderate for HAI 7 through 12, and severe for HAI >12. Fibrosis was staged as mild (score 1 or 2), moderate (3 or 4), and severe (5 or 6).

Statistical Analysis.

The difference between categorical variables was assessed using the chi-squared and Fisher's exact test. The difference between the means of continuous variables was tested by ANOVA and the post-hoc Scheffe test.


Figure 1 shows the main steps in the clinical history of our cohort from enrollment in the longitudinal study up to the last follow-up visit. Table 1 summarizes the essential epidemiological and biochemical features at enrollment in the 89 HBeAg-positive and 8 HBeAg-negative children. The 7 children subsequently treated with IFN for sustained HBeAg positivity were significantly younger than the untreated HBeAg-positive patients. Only 10 HBeAg-positive and 2 HBeAg-negative children were born of mothers known to be infected at delivery. Table 2 summarizes the histological findings at entry and during the follow-up. A few months or years after HBeAg clearance, the liver histology in inactive HBsAg carriers was usually characterized by minimal or mild lesions. Mild and moderate hepatitis, respectively, were found in the twins with HBeAg-negative liver disease 9 years after diagnosis.

Figure 1.

Study flow diagram.

Table 1. Epidemiological, Clinical, and Biochemical Features in HBeAg-Positive and HBeAg-Negative Children at Enrollment in the Longitudinal Study
VariablesHBeAg + (89)Anti-HBe+ (8)
Untreated (68)Treated (21)
IFN (7)Other (14)Untreated (8)
  • *

    Mean ± SD years.

  • P < .001 vs.

  • treated with IFN.

  • §Mean ± SD times × upper normal value.

Male/female (N)39/294/35/96/2
Age at entry*7.4 ± 4.72.2 ± 1.15.7 ± 2.74.8 ± 3.2
Age at treatment*7.7 ± 2.77.7 ± 2.6
Source N (%)    
 Mother at delivery7 (10)3 (43)02
 Transfusions5 (7)1 (14)00
 Family contact36 (53)09 (64)6
 Unknown20 (30)3 (43)5 (36)2
ALT × normal**3.3 ± 3.13.6 ± 3.53.7 ± 3.43.8 ± 2.5
Table 2. Histological Features in the 89 HBeAg-Positive and in the 8 HBeAg-Negative Children at Enrollment and During Follow-Up
Subgroups of PatientsLiver Histology
Minimal/Mild HepatitisModerate HepatitisSevere HepatitisCirrhosis
  • *

    Follow-up histology in the girl who underwent transient reactivation is reported in the text and in Table 4.

  • Follow-up histology is reported in the text.

HBeAg+ at Enrollment    
 Initial biopsy in 89 cases:37 (41.5%)39 (44%)9(10%)4 (4.5%)
 Follow-up biopsy:    
  – in 32 inactive carriers within 5 years after seroconversion*26 (81%)6 (19%)00
  – in 3 HBeAg- hepatitis1200
HBeAg− at Enrollment    
 Initial biopsy in 8 cases7(87.5%)1 (12.5%)00
 After 9 years of follow-up in 2 children with HBeAg− hepatitis1100

HBV genotypes were investigated in 75 HBeAg-positive cases: genotype D was detected in 19 treated children, 52 untreated children without cirrhosis, and 3 children with cirrhosis; genotype A was found in 1 untreated child without cirrhosis.

Outcome After HBeAg Seroclearance in the 85 Children Without Cirrhosis.

The outcome in the treated and untreated patients after HBeAg clearance and anti-HBe seroconversion is summarized in Fig. 2 and illustrated in detail in Table 3.

Figure 2.

Flow diagram of the outcome in 85 children without cirrhosis.

Table 3. Outcome of Infection and Disease After HBeAg Seroclearance in 85 Children Without Cirrhosis
 Untreated (64)Treated (21)
At SeroclearanceAt Last VisitAt SeroclearanceAt Last Visit
  • *

    Mean ± SD years.

  • For measurement see text.

Duration of the HBeAg-positive phase*4.55 ± 3.219.3 ± 5.28.5 ± 5.521.9 ± 4.8
Follow-up after HBeAg seroclearance*14.8 ± 5.713.3 ± 7.2
Age*11.4 ± 5.126.43 ± 5.512.1 ± 4.825.7 ± 4.8
Category N (%):    
Inactive HBsAg carriers63 (98.5)60 (94)21 (100)21 (100)
HBeAg- hepatitis1 (1.5)4 (6)00
HBV DNA+ by PCR    
 N cases (%)Not done35 (55)Not done12 (57)
 Median copies/mL 4,000 3,600
HBsAg clearance N (%)09 (14)04 (19%)
Overweight012 (19)02 (10)
 Drinker09 (16)03 (14)
 Nondrinker64 (100)55 (84)21 (100)18 (86)

Within 6 months of anti-HBe seroconversion, all but 1 untreated and all treated patients had normalized ALT without further evidence of HBV DNA by non-PCR tests, and were considered inactive carriers. Only 1 untreated, 15-year-old boy with ALT fluctuating up to 4-fold the normal level and intermittent HBV DNA positivity by conventional tests met the criteria for HBeAg-negative hepatitis. His biochemical and virological profile remained the same for 15 years; then the ALT and viremia levels tended to rise. At 36 years of age, liver biopsy showed moderate activity and mild fibrosis, and the patient was considered for treatment.

Of the 84 inactive carriers, 13 (15.4%) lost their HBsAg and developed the related antibody. The average interval between anti-HBe seroconversion and HBsAg clearance was 3.9 ± 2.9 years in the 8 boys, and 13.8 ± 7.6 years in the 5 females (P = .08). At the last visit, after a mean 18.1 ± 4.3 years, HBV DNA was undetectable by PCR in all 13 patients. Conversely, 10 untreated carriers developed ALT abnormalities up to 3 times the upper normal limit once or twice during a follow-up lasting 15.0 ± 4.9 years. Co-infections, metabolic diseases, and autoimmunity were ruled out in this subgroup, whereas HBV DNA was detectable by PCR in 7 cases.

Reactivation, with or without seroreversion, was seen in 4 (6.3%) inactive HBsAg carriers, 3 of whom ultimately progressed to HBeAg-negative hepatitis. Table 4 shows the most significant events in the history of these patients. Case 1 had two subsequent ALT flares, the second during a second pregnancy, then retained abnormal ALT levels and a fluctuating viremia by non-PCR tests. Reactivation coincided with drug addiction in cases 2 and 3, and with puberty in case 4. In this last case, the girl soon converted to inactive HBsAg carrier status, and a liver biopsy taken 8 years after the transient ALT flare showed mild to moderate activity and mild fibrosis.

Table 4. Demographic, Clinical, Serological, and Histological Features in the Four Inactive HBsAg Carriers Who Experienced Reactivation During Follow-Up
 Case 1Case 2Case 3Case 4
  • *


Age at seroconversion*1622144
Age at reactivation*25251513
ALT peak (times × N)1225411
HBeAg seroreversion (transient)YesYesNoYes
Liver histology (after reactivation)Mild hepatitisNot doneModerate hepatitisMild hepatitis
Age at last visit*32272223
Clinical outcomeHBeAg- hepatitisHBeAg- hepatitisHBeAg- hepatitisInactive HBsAg carrier
Final eventsLost at follow-upDrug abuse, died at 27 yOn care for drug abuseOn regular follow-up

In short, 4 children had HBeAg-negative hepatitis at the last visit. In this small group, the mean (±SD) age at HBeAg seroclearance (17.2 ± 3.4 years) was significantly higher than in the inactive carriers (11.2 ± 4.9 years; P = .02) and even more so than in the carriers cleared of HBsAg (8.2 ± 5.2 years; P = .03). As a result, the follow-up after seroconversion was significantly longer among the children who cleared HBsAg than in the persistently HBsAg-positive cases (18.1 ± 4.3 vs. 14.2 ± 5.7 years; P = .03). No correlation was found between outcome of hepatitis and gender, source of infection, ALT peak before seroconversion, or duration of the HBeAg-positive phase (data not shown). At the last visit, none of the 85 children without any features of cirrhosis at seroconversion had developed lesions compatible with cirrhosis or HCC.

Table 3 also shows the prevalence of potential cofactors of liver damage. A minority of teenagers or young adults drank alcoholic beverages, mainly on weekends. The prevalence of alcohol drinkers was significantly higher among patients with sporadic ALT abnormalities than among inactive HBsAg carriers or those who had lost HBsAg (50% vs. 7.4%; P = .006). Alcohol consumption was generally mild to moderate.

Outcome in Children With Cirrhosis.

The 4 boys with cirrhosis seroconverted at a mean age of 4.5 ± 4.2 years. They were asymptomatic, with normal ALT. The diagnosis was confirmed in 1 or 2 further biopsy specimens in childhood. Two patients remained asymptomatic, with a normal liver function and HBV DNA negative according to non-PCR tests throughout a follow-up of 16 and 17 years, respectively; one also lost HBsAg. As recently reported in detail,32 liver biopsy in adulthood showed regression of the lesions of cirrhosis in both cases.

Two children with cirrhosis developed HCC 9 and 16 years, respectively, after anti-HBe seroconversion. One patient had also cleared HBsAg. Resection of a single lesion and chemotherapy were successful in both patients, but the final outcome differed: 9 years after surgery, the HBsAg-negative case had a normal liver function, a viremia below 200 copies/mL, and no focal lesions at ultrasonography; the other, younger patient abused drugs and alcohol and developed acute hepatitis C. After abstaining, at 24 years of age, the patient was invited to a check-up: he was well-developed and asymptomatic with serum HBV DNA below 105 copies/mL, but hepatitis C virus RNA positive and had high α-fetoprotein levels. A 40-mm hepatic lesion was detected in the right lobe, compatible with HCC in a liver with cirrhosis. The patient was referred to a transplantation unit.

Outcome in Children Who Were HBeAg Negative at Enrollment.

The 8 initially HBeAg-negative children (2 females and 6 males) were followed for an average 17.0 ± 5.1 years. Six cases slowly returned to normal ALT within 2 years of joining the study and became inactive HBsAg carriers. At the last visit, all 6 were HBV DNA negative by PCR, and 2 of them had also cleared HBsAg. A pair of 9-year-old male twins, whose parents were both HBsAg and anti-HBe positive at first observation, maintained biochemical features of chronic hepatitis during a 9-year follow-up period, after which liver histology showed mild chronic hepatitis in 1 and moderate hepatitis in the other. HBV DNA was detectable by PCR, and only on 1 occasion was the hybridization test positive.


The final outcome of this 29-year longitudinal study contributes information on the late stages of chronic HBV infection acquired in childhood, mainly by horizontal transmission. Our homogeneous study population, almost exclusively infected with genotype D, can be considered as a model of the dynamic interactions between host and HBV in a particular epidemiological setting. None of our patients died of liver-related causes during a mean 20-year follow-up and, at the final evaluation, 91% of untreated and all treated HBeAg-positive patients without cirrhosis, plus 6 of 8 initially anti-HBe–positive children, had become inactive HBsAg carriers. In a recent follow-up study involving 296 former blood donors with chronic HBV infection, Manno et al.33 found that inactive HBsAg carrier status was substantially stable over a 30-year period: these findings, recorded in the same geographical area, offer encouraging prospects for the future of our cohort. Nevertheless, the occurrence of late reactivation, the persistence of low viremia after HBeAg seroclearance and the recently-described occult infection with HBV DNA detectable only in the liver34, 35 cast doubts on the long-term prognosis for our patients and suggest ongoing surveillance.

Reactivation after anti-HBe seroconversion and the development of HBeAg-negative hepatitis were infrequent events in this cohort, though chronic HBeAg-negative hepatitis is thought to accumulate with time in Mediterranean HBV carriers, following the selection of precore-HBV mutants.21–23 We did not investigate e-minus mutants in the 6 children with HBeAg-negative hepatitis at the last visit; however, the clinical history of these patients revealed a slow disease progression, their liver histology showing mild-to-moderate fibrosis even after several years of follow-up.

Low levels of viremia can be detected in a sizable proportion of inactive HBsAg carriers, and our data emphasize the close relationship between loss of HBsAg and clearance of circulating HBV DNA by PCR. This virological pattern is likely to ensure further stability to the host–virus interaction. Whether low viremia levels are responsible for the mild ALT abnormalities detected in a subgroup of carriers remains to be seen. Given the low monitoring frequency, transient increases of HBV DNA levels may have been missed in these cases. Interestingly, alcohol consumption was significantly more frequent in HBsAg carriers with abnormal than in those with normal ALT, a finding that should stimulate further investigations into the pathogenic role of environmental factors.

As previously reported,15 cirrhosis was an early complication in young males with a short-lived HBeAg-positive phase. Why early HBeAg seroconversion led to regression of cirrhosis in 2 cases33 but failed to prevent the asymptomatic onset of HCC in 2 other cases remains without explanation. Closely monitoring patients with cirrhosis enabled an early diagnosis, which facilitated successful treatment and long-term survival.

In conclusion, this study shows that chronic infection with HBV genotype D can produce a wide spectrum of liver lesions and serological patterns in Caucasian children, emphasizing the pathogenic role of the host's immune response. Most patients become inactive HBsAg carriers as young adults, and this condition seems to be stable. Nevertheless, the development of HCC in 2% of cases over a period of 20 years and the progression to HBeAg-negative hepatitis in 6% of cases indicate that chronic hepatitis B acquired in childhood remains a potentially severe disease. Given the long life expectancy, a variety of viral, host, and environmental factors might change the subsequent prognosis of the disease in adulthood. This scenario suggests that an efficient treatment of chronic hepatitis B in non-tolerant children should shorten the highly replicative phase, ultimately increasing the rate of HBsAg clearance. Counseling could prevent comorbidity in adulthood.