SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Hepatitis B e antigen (HBeAg) seroconversion in chronic hepatitis B virus infection confers a favorable prognosis, but untoward outcomes may develop in some patients. The impact of the age of HBeAg seroconversion on prognosis is not clearly known. HBeAg-positive patients with biopsy-proven chronic hepatitis B were followed up long-term. Follow-up studies included liver biochemistry, alpha-fetoprotein, and ultrasonography every 3 to 6 months or more frequently if clinically indicated. Of the patients who underwent spontaneous HBeAg seroconversion, the incidences of HBeAg-negative hepatitis, cirrhosis, hepatocellular carcinoma (HCC), and hepatitis B surface antigen seroclearance were compared between patient groups with different ages at the time of HBeAg seroconversion using Kaplan–Meier survival analysis and Poisson regression model. Spontaneous HBeAg seroconversion was documented in 508 patients. Of the 483 patients who had no evidence of cirrhosis or HCC at the time of HBeAg seroconversion, HBeAg seroconversion occurred before age 30 in 218 patients (group A), between age 31 and 40 in 199 patients (group B), and after age 40 in 66 patients (group C). The 15-year cumulative incidences of HBeAg-negative hepatitis, cirrhosis, and HCC increased with increasing age of HBeAg seroconversion, the lowest being in group A (31.2%, 3.7%, and 2.1%, respectively) and highest being in group C (66.7% [P < 0.0001], 42.9% [P <0.0001], and 7.7% [P = 0.29], respectively). The hazard ratio of HBeAg-negative hepatitis, cirrhosis, and HCC was 2.95, 17.6, and 5.22, respectively, in group C compared with group A. Conclusion: Patients with HBeAg seroconversion before age 30 have excellent prognosis, whereas patients with delayed HBeAg seroconversion after age 40 have significantly higher incidences of HBeAg-negative hepatitis, cirrhosis, and HCC. (HEPATOLOGY 2010.)

It has been well demonstrated that age is an important factor in the natural course of chronic hepatitis B virus (HBV) infection. Studies have shown that the incidence of cirrhosis increases with increasing age at entry1, 2 and increasing age of hepatitis B e antigen (HBeAg) seroconversion, especially over age 40.3, 4 It has also been demonstrated that the incidence of hepatocellular carcinoma (HCC) increases and survival decreases in patients with increasing age at the time of cirrhosis development.5, 6 Finally, age was a significant factor associated with hepatitis B surface antigen (HBsAg) seroclearance.7–9

Spontaneous HBeAg seroconversion with loss of viral replication is a hallmark of transition from active to inactive phase with remission of disease activity and excellent prognosis, though untoward outcomes may develop in a certain proportion of patients thereafter.4, 10–12 Recent studies have shown that age over 40 years at the time of HBeAg seroconversion was a factor associated with the reactivation of hepatitis B.13, 14 In addition, there is a hint that HBeAg seroconversion before age 30 is associated with low incidence of hepatitis reactivation.13 However, the detailed impact of age of HBeAg seroconversion, in relation to other factors, on the long-term outcomes has not been well illustrated. We therefore conducted a cohort study to explore whether HBeAg seroconversion at different ages is associated with different outcomes.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Patients.

The long-term follow-up study of our patients with chronic hepatitis B, as described earlier,1, 10 has continued. Excluding patients with concurrent hepatitis C virus or hepatitis D virus (HCV/HDV) infection or other liver diseases (cirrhosis, HCC, alcoholism, autoimmune disease), 770 consecutive HBeAg-positive patients with histologically confirmed active chronic hepatitis B had been followed up for a minimum of 1 year by the end of 2000. At entry, the alanine aminotransferase (ALT) level was >2 times the upper limit of normal (ULN) in 543 (70.5%) patients and >5 × ULN in 371 (48.2%) patients. Of the 770 patients, 508 underwent HBeAg seroconversion. HBeAg reversion occurred in 14 patients who later regained stable HBeAg seroconvesion. In total, spontaneous HBeAg seroconversion was confirmed in 508 (66.0%) patients during a mean follow-up period of 3.9 ± 3.8 years (median, 2.5 years [range, 0.1-23.2 years]), with a calculated annual incidence of 16.9%. An additional 27 patients became HBeAg-seronegative but remained seronegative for HBeAg antibody (anti-HBe) until the last follow-up visit. Of the 508 patients with confirmed HBeAg seroconversion, 23 had developed liver cirrhosis and two had developed HCC before HBeAg seroconversion. The remaining 483 patients who had no evidence of cirrhosis or HCC at the time of seroconversion formed the basis of the present study (Fig. 1).

thumbnail image

Figure 1. Flow diagram of patient selection for the follow-up study following confirmed HBeAg seroconversion. Excluding 235 patients with persistent HBeAg seropositivity and 27 who lost HBeAg but remained seronegative for its antibody (anti-HBe), and 25 patients who had evidence of cirrhosis or hepatocellular carcinoma (HCC) before HBeAg seroconversion, 483 were selected and divided according to age at the time of confirmed HBeAg seroconversion. CHB, chronic hepatitis B; (+), seropositive; (−), seronegative.

Download figure to PowerPoint

Assessments.

Liver biochemical tests; virologic markers including HBsAg, anti-HBc immunoglobulin M, HBeAg, anti-HBe, anti–hepatitis A virus immunoglobulin M, anti-HCV, and anti-HDV; and alpha-fetoprotein were assayed and liver biopsy was performed to confirm the diagnosis of chronic hepatitis B at entry. Following HBeAg seroconversion, patients were followed up every 3 to 6 months or more frequently if clinically indicated. Follow-up studies included clinical assessment, liver biochemical tests, relevant hepatitis markers, alpha-fetoprotein, and ultrasonography. Anti-HCV and anti-HDV were assayed again after each event of ALT elevation to >2 × ULN. Serum HBV-DNA was assayed when serum ALT increased to >2 × ULN during follow-up. Retrospective assay for HBV-DNA was also performed if adequate serum specimens collected within 12 months following HBeAg seroconversion were available. Anti-HBs was tested when HBsAg seroclearance was detected. Ultrasonography and alpha-fetoprotein assay were performed every 3 to 6 months to detect the development of cirrhosis or HCC.

Methods.

The biochemical tests were measured by routine automated techniques. Serum hepatitis markers including HBsAg, anti-HBs, HBeAg, anti-HBe, and anti-HDV were assayed using a radioimmunoassay kit (Abbott, North Chicago, IL). Anti-HCV was tested using a commercially available enzyme immunoassay kit (HCV EIA II and III; Abbott). HBV genotypes were determined by way of polymerase chain reaction restriction fragment length polymorphism of the surface gene of HBV as described elsewhere.15 HBV-DNA was measured using quantitative polymerase chain reaction assay (Roche Cobas HBV monitor test, Roche Molecular Systems); the lowest detection limit was 300 copies/mL or 60 IU/mL. Ultrasonography was performed with a high-resolution, real-time scanner (model SSD-1200, SSD-630, SSD-650; Aloka, Tokyo, Japan) equipped with 3.5- and 5.0-MHz rectilinear array or convex scan probes.

HBeAg seroconversion was defined as HBeAg seroclearance with appearance of anti-HBe. HBeAg reversion was defined as reappearance of HBeAg seropositivity after HBeAg seroconversion. HBeAg seroconversion not followed by reversion or loss of anti-HBe was defined as confirmed HBeAg seroconversion. Time to HBeAg seroconversion was defined as the period from entry to the time of confirmed HBeAg seroconversion. HBeAg-negative hepatitis was defined as ALT elevation to >2 × ULN with serum HBV-DNA level >1.0 × 105 copies/mL (2 × 104 IU/mL) in patients who remained HBeAg-negative.4, 10 The duration of remission was defined as the time between confirmed HBeAg seroconversion and the onset of HBeAg-negative hepatitis, or the date of last follow-up if there was no ALT elevation. The diagnosis of cirrhosis was made by histological findings or ultrasonographic findings supplemented with clinical features such as thrombocytopenia, esophageal varices, ascites, or hepatic encephalopathy, as described elsewhere.1, 4, 10 The ultrasonographic findings were scored quantitatively for liver surface, liver parenchyma, hepatic vessel, and spleen size according to the scoring system,16 by which a correlation study revealed a close association between histological and ultrasonographic findings (r = 0.571; P < 0.001) in the diagnosis of cirrhosis.17 HCC was diagnosed by way of histology/cytology or imaging findings (ultrasonography, computed tomography, or magnetic resonance) with an alpha-fetoprotein level >400 ng/mL.10 HBsAg seroclearance was defined as persistent absence of HBsAg for at least 1 year and until the last visit, as described.18

Statistical analysis was performed with the chi-square test, Student t test, and one-way analysis of variance for categorical and continuous variables, respectively, between groups of patients with different ages at the time of HBeAg seroconversion. The difference and correlation between different age groups and the outcomes were analyzed using Poisson regression model. The Kaplan–Meier survival analysis with a log rank test was used in the comparison of outcomes (HBeAg-negative hepatitis, development of cirrhosis, HCC, and HBsAg seroclearance) after spontaneous HBeAg seroconversion. Univariate and multivariate analyses were performed by way of Cox proportional hazards regression models to assess the significant factors for the development of HBeAg-negative hepatitis, liver cirrhosis, and HCC. Statistic procedures were performed using SPSS software version 15.0 for Windows (SPSS, Chicago, IL). P < 0.05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Characteristics at Time of HBeAg Seroconversion.

Of the 483 patients with confirmed HBeAg seroconversion, the age range at entry was 16-69 years (mean, 29.1 ± 7.5 years; median, 28.1 years), and 390 (80.7%) were male. HBeAg seroconversion occurred at ≤30 years in 218 patients (group A), 31-40 years in 199 patients (group B), and >40 years in 66 patients (group C) (mean, 32.3 ± 8.0 years; median, 31.3 years [range, 16-71 years]). Stored serum specimens were available for HBV genotyping in 402 patients, and for HBV-DNA assay within 1 year after HBeAg seroconversion in 357 patients. The key characteristic of the patient groups with different ages at HBeAg seroconversion (groups A, B, and C) are shown in Table 1. Group C patients had a significantly older age at entry, had longer time to HBeAg seroconversion, and more often had an HBV-DNA level ≥104 copies/mL within 1 year following HBeAg seroconversion.

Table 1. Key Characteristics of Patients with Different Ages of HBeAg Seroconversion
 Years of Age at HBeAg SeroconversionP Value
Total (N = 483)Group A: ≤30 (n = 218)Group B: 31-40 (n = 199)Group C: <40 (n = 66)
  • Abbreviation: SD, standard deviation.

  • *

    HBV genotype data available in 402 patients.

  • HBV-DNA data within 1 year after HBeAg seroconversion available in 357 patients.

  • Alanine aminotransferase (ALT) remained abnormal >6 months after HBeAg seroconversion.

Years of age at entry, mean ± SD (range)29 ± 7.524 ± 3.731 ± 3.841 ± 8
(16-69)(16–30)(20–40)(27–69)
 ≤ 30, n (%) 218 (100)84 (42.2)5 (7.6)
 31-40, n (%) 0115 (57.8)21 (31.8)
 >40, n (%) 0040 (60.6)
Male sex, n (%)390 (80.7)179 (82.1)160 (80.4)51 (77.3)0.674
Genotype*    0.011
 B33215013547
 C70183913
Time to HBeAg seroconversion in years, mean ± SD3.2 ± 3.11.8 ± 1.73.9 ± 3.24.9 ± 4.6<0.0001
HBV- DNA, n (%)    0.001
 <104 copies/mL164 (45.9)77 (48.1)76 (51.7)11 (22.0)
 ≥104 copies/mL193 (54.1)83 (51.9)71 (48.3)39 (78.0)
Abnormal ALT, n (%)35 (7.2)14 (6.4)14 (7.0)7 (10.6)

Outcomes Following HBeAg Seroconversion.

The outcomes of these 483 patients in different age groups are compared in Table 2. The duration of remission was significantly shorter in patients with HBeAg seroconversion after age 40. ALT was still abnormal >6 months after HBeAg seroconversion, indicating direct transition from HBeAg-positive to HBeAg-negative hepatitis, in 14 (6.4%) of group A, 14 (7.0%) of group B, and 7 (10.6%) of group C patients (Fig. 1A). During a mean period of 11.7 years (median, 11.5 years [range, 1.0-27.0 years]) following HBeAg seroconversion in these 483 patients, HBeAg-negative hepatitis was documented in 165 (34.2%) patients at a mean age of 39 years (median, 39 years [range, 19-64 years]), cirrhosis developed in 50 (10.3%) patients at a mean age of 46 years (median, 45 years [range, 28-69 years]), HCC occurred in 12 (2.5%) patients at a mean age of 48 years (median, 47 years [range, 35-63 years]) and HBsAg seroclearance occurred in 40 (8.3%) patients at a mean age of 42 years (median, 44 years [range, 20-67 years]). Due to cost/reimbursement issues, anti-HBV therapy was instituted in only 42 patients who developed HBeAg-negative hepatitis. Their follow-up was censored on the day of starting therapy. The 15-year cumulative incidence of HBeAg-negative hepatitis, cirrhosis, and HCC development in 483 patients was 38.6%, 12.5%, and 3.1%, respectively. The patients who developed HBeAg-negative hepatitis, cirrhosis, or HCC were significantly older at the time of HBeAg seroconversion than those without these outcomes (34.6 ± 7.9 versus 31.0 ± 7.8 years [P < 0.0001], 39.6 ± 7.9 versus 31.4 ± 7.6 years [P < 0.0001], and 36.9 ± 7.7 versus 32.1 ± 8.0 years [P = 0.04], respectively).

Table 2. Comparison of Outcomes Among Patients with Different Ages of HBeAg Seroconversion
Events Following HBeAg SeroconversionYears of Age at HBeAg SeroconversionP Value
Total (N = 483)Group A: ≤30 (n = 218)Group B: 31-40 (n = 199)Group C: >40 (n = 66)
  • Abbreviation: SD, standard deviation.

  • *

    15-year cumulative incidence by way of Kaplan–Meier survival analysis with log rank test.

Follow-up period in years, mean ± SD (range)11.7 ± 6.9 (1–27)11.8 ± 7.4 (1–27)12.2 ± 6.7 (1.1–26.5)9.9 ± 5.5 (1.1–20.7)0.069
Duration of remission in years, mean ± SD9.0 ± 7.39.5 ± 7.59.4 ± 7.46.2 ± 5.20.004
Outcome*     
HBeAg-negative hepatitis38.6%31.2%37.3%66.7%<0.0001
Cirrhosis12.5%3.7%12.9%42.9%<0.0001
HCC3.1%2.1%3.2%7.7%0.583
HBsAg seroclearance10.3%8.9%13.2%5%0.518

Because the number of patients at risk in group C was very small at 20-year follow-up, the incidences were compared only up to 15 years following HBeAg seroconversion. The 15-year cumulative incidences of HBeAg-negative hepatitis and cirrhosis development were significantly higher in group C than in groups A and B (P < 0.0001). The 15-year cumulative incidence of HCC tended to be higher in patients >40 years at HBeAg seroconversion; however, the difference was not significant (P = 0.583), nor was the difference in 15-year cumulative incidence of HBsAg seroclearance (P = 0.518) (Table 2 and Fig. 2).

thumbnail image

Figure 2. Comparisons of post-HBeAg seroconversion cumulative incidence of (A) HBeAg-negative hepatitis, (B) liver cirrhosis, (C) HCC, and (D) HBsAg seroclearance among patients with HBeAg seroconversion at age ≤30, 31-40, and >40.

Download figure to PowerPoint

The number of patients, patient-years of follow-up, calculated annual incidence, and hazard ratio (HR) of HBeAg-negative hepatitis, cirrhosis, and HCC development and HBsAg seroclearance in the three groups are shown in Table 3. The development of cirrhosis increased significantly with increasing age of HBeAg seroconversion, especially in patients with age of HBeAg seroconversion over 40 (HR 17.6). The occurrence of HBeAg-negative hepatitis and HCC was significantly higher for the patients with age of HBeAg seroconversion over 40 (HR 2.95 and 5.22, respectively). The trend for HBsAg seroclearance was not significantly different between different age groups.

Table 3. Comparison of Calculated Annual Incidence of HBeAg(−) Hepatitis, Liver Cirrhosis, HCC, and HBsAg Seroclearance in Patients with Different Ages at HBeAg Seroconversion
HBeAg SeroconversionHBeAg(−) HepatitisLiver CirrhosisHCCHBsAg Seroclearance
Age in YearsNo. of Patientsn (Calculated Annual Incidence)*Person-YearsHR (95% CI)P Valuen (Calculated Annual Incidence)*Person-YearsHR (95% CI)P Valuen (Calculated Annual Incidence)*Person-YearsHR (95% CI)P Valuen (Calculated Annual Incidence)*Person-YearsHR (95% CI)P Value
  • Abbreviation: CI, confidence interval.

  • *

    Calculated annual incidence expressed as %/year.

≤3021860 (2.9)2,0511 6 (0.2)2,5181 3 (0.1)2,5521 21 (0.8)2,4711 
31-4019968 (3.8)1,7851.26 (0.89–1.78)0.19122 (1.0)2,2624.08 (1.66–10.1)0.0025 (0.2)2,4121.76 (0.42–7.38)0.43717 (0.7)2,3280.86 (0.45–1.63)0.642
>406637 (8.6)4282.95 (1.96–4.45)<0.000122 (4.2)52317.6 (7.15–43.5)<0.00014 (0.6)6515.22 (1.17–23.3)0.0302 (0.3)6470.36 (0.09–1.55)0.171
Total483165 (3.9)4,264  50 (0.9)5,303  12 (0.2)5,615  40 (0.7)5,446  

For retrospective HBV-DNA assay, stored serum specimens collected within 12 months after HBeAg seroconversion were available in 357 patients (229 within 6 months, 128 between 6 and 12 months). Of the serum specimens collected within 6 months and between 6 and 12 months, HBV-DNA was ≥104 copies/mL in 135 (59%) and 58 (45%), respectively. The proportion of assays ≥104 copies/mL tended to decrease with the time after HBeAg seroconversion (P = 0.018). Group C patients had a significantly higher rate of HBV-DNA level ≥104 copies/mL (P = 0.001) (Table 1). Patients with HBV-DNA ≥ 104 log copies/mL within 1 year after HBeAg seroconversion had significantly higher 15-year cumulative incidences of HBeAg-negative hepatitis (54.1% versus 30.2% [P < 0.0001]) and cirrhosis (17.3% versus 4.6% [P = 0.002]). The cumulative incidence of HCC increased two- to three-fold but was not significantly different (10.4% versus 3.9% [P = 0.594]), possibly due to the small number of patients who developed HCC.

In the univariate and multivariate analyses using Cox proportional hazards regression models, the possible associated factors included were sex, age at entry, HBV genotype, HBeAg reversion, age at HBeAg seroconversion, time to HBeAg seroconversion, HBV-DNA level within 1 year after HBeAg seroconversion, duration of remission after HBeAg seroconversion, and HBeAg-negative hepatitis and cirrhosis, where appropriate (Table 4). After excluding the patients with missing data, 314 (65%) patients were analyzed in the multivariate analysis. The age at entry was an independent factor for the ocurrence of HBeAg-negative hepatitis (P = 0.003) and cirrhosis development (P < 0.0001) in Cox regression analysis. Age at entry and time to HBeAg seroconversion determine the age at HBeAg seroconversion. Interference arises when all these interrelated factors are included in the multivariate analysis. Because the aim of the present study was to examine the impact of age at HBeAg seroconversion on the prognosis of chronic hepatitis B, we chose to exclude age at entry in the meta-analysis. The age of HBeAg seroconversion and HBV-DNA ≥ 104 copies/mL were two independent factors for the development of HBeAg-negative hepatitis following HBeAg seroconversion in both univariate and multivariate analysis. The age of HBeAg seroconversion (P < 0.0001), time to HBeAg seroconversion (P = 0.001), HBV-DNA ≥ 104 copies/mL (P = 0.004), duration of remission (P < 0.0001), and HBeAg-negative hepatitis (P < 0.0001) were factors for cirrhosis development after HBeAg seroconversion in univariate analysis, while the age at HBeAg seroconversion (P < 0.0001), HBeAg reversion (P = 0.007), and duration of remission (P = 0.039) were independent factors for cirrhosis development in multivariate analysis. The age at HBeAg seroconversion (P = 0.01), time to HBeAg seroconversion (P = 0.02), duration of remission (P = 0.016), and cirrhosis (P < 0.0001) were factors for HCC development in univariate analysis, but cirrhosis development after HBeAg seroconversion was the only independent factor for HCC occurrence (P = 0.004) in a multivariate model. Sex and HBV genotype were not factors associated with HBeAg-negative hepatitis, cirrhosis, or HCC development following spontaneous HBeAg seroconversion (Table 4). Although the HBeAg-negative hepatitis was not an independent factor of cirrhosis and HCC, the cumulative incidence of cirrhosis in patients with and those without HBeAg-negative hepatitis was 32.1% and 6.3%, respectively (P < 0.0001). The cumulative incidence of HCC in patients with and those without HBeAg-negative hepatitis was 8.8% and 4.7%, respectively (P = 0.133). If analyses were restricted to the 372 patients (335 had HBV genotype data and 283 had HBV-DNA data) with a follow-up period ≥5 years after HBeAg seroconversion, the results and conclusion were similar.

Table 4. Univariate and Multivariate Analyses of Factors for Outcomes After HBeAg Seroconversion
FactorsHBeAg(−) HepatitisLiver CirrhosisHCC
Univariate AnalysisMultivariate AnalysisUnivariate AnalysisMultivariate AnalysisUnivariate AnalysisMultivariate Analysis
HR (95% CI)P ValueHR (95% CI)P ValueHR (95% CI)P ValueHR (95% CI)P ValueHR (95% CI)P ValueHR (95% CI)P Value
  • Abbreviations: NA, not applicable; ND, not done.

  • *

    HBV genotype data available in 402 patients.

  • All 12 patients with HCC development were male, and no analysis was performed.

  • HBV DNA (copies/mL) within 1 year after HBeAg seroconversion were available in 357 patients.

Age (increase by year)1.04 (1.03–1.06)<0.00011.04 (1.01–1.06)0.0021.09 (1.07–1.12)<0.00011.09 (1.05–1.13)<0.00011.09 (1.02–1.16)0.0100.93 (0.83–1.05)0.240
Age (increase by decade)1.53 (1.30–1.82)<0.00011.45 (1.14–1.84)0.0022.52 (1.97–3.23)<0.00012.26 (1.55–3.30)<0.00012.29 (1.22–4.33)0.0100.51 (0.16–1.57)0.240
Sex            
 Female10.06810.12510.91410.232NDNDNDND
 Male1.51 (0.97–2.34) 1.56 (0.88–2.77) 0.96 (0.47–1.98) 0.57 (0.23–1.43)     
Genotype*            
 B10.44610.54910.08910.39610.39410.741
 C0.84 (0.54–1.31) 0.86 (0.52–1.42) 1.77 (0.92–3.40) 1.47 (0.61–3.57) 1.77 (0.48–6.52) 1.33 (0.24–7.35) 
Reversion1.49 (0.70–3.19)0.2991.78 (0.81–3.91)0.1492.42 (0.75–7.77)0.1396.03 (1.65–22.0)0.0072.62 (0.34–20.3)0.3570.37 (0.03–4.27)0.424
Time to HBeAg seroconversion (year)1.03 (0.98–1.08)0.2271.01 (0.95–1.06)0.8151.14 (1.06–1.23)0.0011.07 (0.98–1.17)0.1151.21 (1.03–1.41)0.0201.22 (0.97–1.52)0.082
HBV-DNA            
<1041<0.00011<0.000110.00410.15410.59610.532
≥1043.14 (2.08–4.73) 2.51 (1.62–3.89) 3.76 (1.54–9.20) 2.00 (0.77–5.19) 1.41 (0.40–4.99) 0.63 (0.15–2.65) 
Duration of remission in yearsNANANANA0.87 (0.83–0.91)<0.00010.90 (0.82–0.99)0.0390.91 (0.84–0.98)0.0160.83 (0.69–1.01)0.061
HBeAg(−) hepatitisNANANANA6.16 (3.22–11.8)<0.00011.91 (0.61–6.01)0.2712.35 (0.75–7.41)0.1450.16 (0.01–2.42)0.186
Liver cirrhosisNANANANANANANANA10.9 (3.45–34.3)<0.000114.4 (2.34–89.1)0.004

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This large-scale long-term cohort study included a consecutive series of patients with chronic HBV infection from immune clearance phase to inactive/residual phase and reactivation phase.19 The calculated annual incidence of HBeAg seroconversion was 16.9%, which was close to the rates in chronic active hepatitis B reported earlier.20, 21 Of note is that the older the patients were at the time of HBeAg seroconversion, the older their age at entry and the longer time to HBeAg seroconversion (Table 1). This trend is consistent with the observation of Evans et al.21 that Asian American patients aged 20-34 years at entry showed a 15-month HBeAg seroconversion rate of 23% in contrast to a rate of 17% in those aged 35-50 years at entry. These findings suggest that patients with HBeAg persistence over age 35 years are inherently more difficult to clear HBeAg. As shown in Table 1, higher proportions of the patients with older age at HBeAg seroconversion were infected with genotype C HBV, which is known to be associated with late HBeAg seroconversion.22 Because HBeAg seroconversion is a hallmark of transition from active to inactive phase of chronic HBV infection,3, 19 the onset of HBeAg seroconversion was used as the starting point in studies addressing long-term prognosis following spontaneous HBeAg seoconversion,4, 10 as we did in the present study.

The results of the present study have shown that patients with HBeAg seroconversion before age 30 have a very low cumulative incidence of cirrhosis and HCC, whereas HBeAg seroconversion after age 40 was associated with a much higher incidence of HBeAg-negative hepatitis, cirrhosis, and HCC during a mean follow-up period of 11.7 years (range, 1-27 years). Although there was no significant difference in the incidence of HBeAg-negative hepatitis between patients with HBeAg seroconversion at age 31-40 years and those before age 30 (Fig. 1A), the incidence of cirrhosis development was significantly higher in the former (Fig. 1B). It has been shown that HBeAg-positive phase of chronic HBV infection is usually associated with hepatitis activity and hepatitis flares19 and that persistence of HBeAg was associated with a higher likelihood of fibrosis progression23 and cirrhosis development.3, 4, 24 It is possible that patients with HBeAg seroconversion at older age had already accumulated more advanced fibrosis during their longer HBeAg-positive phase thus they were closer to and more apt to develop cirrhosis following the onset of HBeAg-negative hepatitis. The Poisson regression analysis showed the HR of HBeAg-negative hepatitis, cirrhosis, and HCC development in patients with HBeAg seroconversion over age 40 was 2.95, 17.6, and 5.22, respectively when compared with those with HBeAg seroconversion before age 30 (Table 3). Given the finding that HCC occurred at a median age of 47 years (range, 35-63 years), it is possible that the risk of HCC is lower in group A due to a younger population. However, all three HCC patients in group A developed HCC within 7 to 8 years following spontaneous HBeAg seroconversion (data not shown). In addition, group A patients also had a much lower incidence of cirrhosis, which is the only independent factor for HCC development (HR 14.4 versus 0.51 per decade of age) (Table 4). Our earlier studies have consistently shown that most cases of HBeAg-negative hepatitis occur within 10 years following HBeAg seroconversion or after entry.4, 10, 17 Group A patients had been followed up for a mean period of 11.8 years; the risk of disease progression is likely to be minimal upon longer follow-up.

The excellent prognosis in patients with earlier HBeAg seroconversion is consistent with the excellent prognosis following HBeAg seroconversion in Italian pediatric patients in that only four (6.3%) of their 63 inactive carriers developed HBeAg-negative hepatitis, none developed cirrhosis or HCC during a mean follow-up of 14.8 ± 5.7 years, and that the children who developed HBeAg-negative hepatitis had seroconverted at an older age (mean, 17.2 ± 3.4 versus 11.2 ± 4.9 years).11 Similarly, of the 58 Taiwanese pediatric patients who underwent HBeAg seroconversion at a mean age of 17.2 ± 5.8 years, none developed HBeAg-negative hepatitis, cirrhosis, or HCC during a mean follow-up period of 6.5 ± 5.0 years.25 The results are also consistent with an earlier finding in a study of 1,241 inactive HBsAg carriers whose age of HBeAg seroconversion was unknown in that those of age <30 years at entry had the lowest incidence of HBeAg-negative hepatitis and cirrhosis development.13 Of note is that an even lower annual incidence of HBeAg-negative hepatitis (0.89%) and cirrhosis (0.03%) was observed in that group of inactive carriers compared with the corresponding rates in the current study (3.9% and 0.9% respectively) (Table 3).13 Because their onset of HBeAg seroconversion was unknown and thus they conceivably presented at a later phase after HBeAg seroconversion, their real post-HBeAg seroconversion follow-up period was conceivably longer than the follow-up period reported in that study.13 Their low incidence is consistent with the findings that the vast majority of HBeAg-negative hepatitis occurred in the first 10 years after HBeAg seroconversion, as shown in Fig. 1A and in earlier studies.4, 10 In addition, age at the time of HBeAg seroconversion was an independent factor for the development of cirrhosis, but not for HCC, following HBeAg seroconversion in multivariate analysis. However, its significant role in cirrhosis development may have an indirect influence on HCC because liver cirrhosis was the only independent factor for HCC development (Table 4). Our results have not only confirmed that delayed HBeAg seroconversion is associated with higher risk of untoward outcomes,3, 4, 13, 14 but also further indicate that the younger the age of HBeAg seroconversion, the better the outcomes.

In addition to age at HBeAg seroconversion, serum HBV-DNA ≥ 104 copies/mL in the first year following HBeAg seroconversion was also an independent factor for the development of HBeAg-negative hepatitis (Table 4). This is consistent with earlier observations that serum HBV-DNA ≥ 104 copies/mL is associated with future development of HBeAg-negative hepatitis26, 27 and disease progression.28, 29 The duration of remission was also an independent factor for cirrhosis development, whereas HBeAg-negative hepatitis was not a significant factor in multivariate analysis (Table 4). This is conceivable because earlier occurrence of HBeAg-negative hepatitis (shorter duration of remission) may lead to a longer period of hepatitis activity and associated disease progression (such as accumulation of fibrosis) compared with late onset of HBeAg-negative hepatitis during same duration of follow-up.

Earlier studies reported that HBeAg reversion and HBV genotype were important factors for hepatitis B reactivation and cirrhosis development10, 14, 22, 30 and male sex was associated with HBeAg-negative hepatitis following HBeAg seroconversion.13, 14 However, except HBeAg reversion as a factor for cirrhosis development, these earlier findings have not been confirmed by the present study involving a much more patients presenting with active chronic hepatitis (Table 4). The different patient population recruited (active versus inactive), the larger number of patients (483 versus 133), and more factors included in the analyses may be the reasons for the discrepancies of the results.

In conclusion, untoward outcomes can develop after spontaneous HBeAg seroconversion. The cumulative incidence and HR of HBeAg-negative hepatitis B, cirrhosis, and HCC development were significantly higher in patients with HBeAg seroconversion after age 40. Age of HBeAg seroconversion is an independent factor for the development of HBeAg-negative hepatitis and cirrhosis. Patients with HBeAg seroconversion before age 30 have the best outcomes and have a very low risk of disease progression.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We acknowledge the excellent assistance of Su-Chiung Chu.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information
FilenameFormatSizeDescription
HEP_23348_sm_Chinese-Version.pdf605KChinese Version

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.