Lemondeestailleurs.ca

Long-Term Follow-up of Chronic Hepatitis B Virus Infection in Children
of Different Ethnic Origins
George Marx,1 Steven R. Martin,1
1Division of Gastroenterology, Hepatology and Nutrition, Jean-Francois Chicoine,2 and Fernando Alvarez1
Sainte-Justine Hospital; 2Department of Pediatrics, University The natural history of chronic hepatitis B in children is influenced by mode of transmission
and varies with regional endemicity. Seroconversion rates were studied in 174 hepatitis B e
antigen (HBeAg)–positive children who were of different ethnic origins and living in Canada.
Overall, 40.2% became anti-HBeAg positive, and 8.6% were hepatitis B surface-antigen positive
during a mean follow-up of 4.5 years. Spontaneous seroconversion rates were lower in Asian-
born, mainly vertically infected, children, versus those born either in Canada or where hor-
izontal transmission predominates (24% vs. 44%, P p .015). Kaplan-Meier analysis showed
that the cumulative persistence of HBeAg after 13 years was 25% in Asian-born children,
versus 6% in all others (P ! .05). Treatment of 27 children accelerated seroconversion by 3
years, without influencing the proportion seroconverting over time. Thus, although Asian-
born children seroconvert more slowly, a large proportion will seroconvert before adulthood.
Because treatment appears to accelerate anti-HBe seroconversion, longitudinal studies are
required in order to assess the long-term benefits of early treatment.
Universal vaccination against hepatitis B virus (HBV) infec- B e antigen (HBeAg) during a mean follow-up of 4 years, and tion was introduced into Taiwan in 1984 but not into most all children remained hepatitis B surface-antigen (HBsAg) pos- industrialized nations until 1991 [1]. The result has been a sig- itive [7]. In contrast, Evans et al. [8] found a higher serocon- nificant reduction in perinatal and horizontal transmission of version rate (13%) in 454 HBeAg-positive Asian-American car- HBV and in the incidence of hepatocellular carcinoma (HCC) riers (47% were children), who were followed a median of 25 [2, 3]. In the United States, a region considered to be of low months [8]. Of these patients, 95% were born in Asia but live endemicity, despite a reduction of ∼60% in the incidence of hepatitis B during 1984–1994, an estimated 185,000 new infec- In a long-term follow-up study by Bortolotti et al. [9], 185 tions still occur each year [4]. Furthermore, the increased in- Mediterranean children were followed an average of 13 years cidence of HCC in the United States is thought to be caused (some were treated). Of these children, 84% cleared HBeAg, by the increasing incidence of hepatitis C virus (HCV) infection whereas only 6% lost HBsAg. Five children developed liver and by increased immigration from regions where HBV is en- cirrhosis, which progressed to HCC in 2 of them [9]. A minority demic. Infections that occurred during the 1960s and 1970s now (14%) of these children were born to mothers who tested pos- are reaching 2–3 decades of evolution [5]. Published estimates itive for HBsAg; similar long-term studies for children infectedat birth are lacking.
of the risk for development of HCC in HBV carriers vary wide- Here we describe the long-term outcome of chronic HBV ly but are generally highest in Asian countries and Alaska infection in a heterogeneous group of children of different eth- (240–494/100,000 carrier-years) and lowest in western countries nic origins. All lived in a similar environment in the Province (0–13/100,000 carrier-years) [6]. A clear understanding of the natural history of HBV infection would permit appropriateintervention to prevent the development of complications.
The natural history of chronic HBV infection in children has Patients and Methods
been determined from limited data that vary by geographic Over 20 years (1980–2000), 174 children who were 1 region. In a Chinese study, only 7% of children cleared hepatitis month–18 years (mean, 4.7 ע 4.6 years) old and had hepatitis Binfection were consecutively observed. They were followed for a me- Received 23 October 2001; revised 6 February 2002; electronically pub- dian of 4.9 ע 3.1 years (range, 4 weeks–16 years) at the Department of Pediatric Gastroenterology and Hepatology of Hoˆpital Sainte- Reprints or correspondence: Dr. Fernando Alvarez, Sainte-Justine Hos- Justine in Montreal. Approximately two-thirds of the patients were pital, Div. of Pediatric Gastroenterology, Hepatology and Nutrition, 3175Coˆte-Sainte-Catherine, Montreal, Quebec H3T 1C5, Canada (fernando referred from the International Adoption Clinic at Hoˆpital Sainte- Justine, because of positive results of HBV serology screening. Thisclinic, in operation since 1989, sees about 50% of the 1000 children The Journal of Infectious Diseases
2002; 186:295–301
from foreign countries adopted in Quebec each year. The remaining 2002 by the Infectious Diseases Society of America. All rights reserved.
patients were referred for evaluation of elevated aminotransferases Demographic features of hepatitis B virus–infected children at presentation, by birth HBsAg, hepatitis B surface antigen; HCV, hepatitis C virus.
b These children were infected during the first year of life either by vertical transmission or by contact in and positive results of HBV serology. All children were HBsAg pos- HBsAg positive for у6 months, HBV DNA- and HBeAg-positive itive. The presence of chronic hepatitis was confirmed, on the basis twice, alanine aminotransferase (ALT) levels 12 times normal, and of liver histology, in 42 patients (23.2%).
chronic hepatitis on liver biopsy [10]. The second trial recruited The patients were divided into 4 groups, on the basis of region patients during 1999–2000, for lamivudine therapy. These children of birth. The Asian group included children born in China and were 2–17 years old; were HBsAg positive, HBV DNA positive, Vietnam. The eastern-European group comprised children born in and HBeAg positive; and had ALT levels 11.3 times normal [11].
Romania and Russia. The Canadian children were all born to All patients who met the entry criteria during each period partic- parents born in Canada; no native Canadian children were fol- ipated in the study. A response to therapy was defined as a loss of lowed. The fourth group (others) comprised children born in Haiti, HBeAg, development of anti-HBe, clearance of HBV DNA from Pakistan, Bangladesh, India, and Latin America.
serum, and normalization of ALT levels. During follow-up, patients Of 174 children, 27 (15.5%) were treated within the context of were usually seen in our outpatient clinic every 3–12 months, for clinical trials. The first trial, with interferon (IFN)–a, recruited physical examination and biochemical and serologic testing. Serum patients during 1992–1994. Inclusion criteria were as follows: samples were obtained and stored at Ϫ20ЊC. Beginning in 1992, Alanine aminotransferase (ALT), hepatitis B e antigen (HBeAg), hepatitis B surface antigen (HBsAg), hepatitis B virus (HBV) DNA, and a-fetoprotein levels in hepatitis B vi-rus–infected children at study entry, by birth region.
Anti-HBe, hepatitis B e antibody; anti-HBs, hepatitis B surface antibody.
a Canadians vs. all others.
b P ! .05 eastern Europeans vs. all others.
Histopathologic analysis of liver-biopsy samples of hepatitis B virus–infected children, by all patients were monitored, with yearly abdominal ultrasound and to evaluate the potential impact of the predictors, controlling for the age at diagnosis. Finally, we used multivariate Cox regression HBsAg; HBeAg; antibodies to HBe, HBs, and hep- survival analysis with stepwise selection of explanatory predictors, atitis D; and a-fetoprotein levels were measured by commercial to extract the best subset of predictors.
immunoassay kits (Abbott Laboratories). HBV DNA was analyzedsemiquantitatively by dot-blot hybridization, beginning in 1989(assay cutoff point, 5 pg/mL). In 27 of 41 patients, liver biopsy was performed, as per study protocols, prior to treatment; in therest, it was performed either because of persistently elevated ami- Table 1 lists the demographic features of notransferases or to confirm the suspicion of cirrhosis. The biopsy the 174 children at presentation to our clinic. Most of the chil- was performed with a Jamshidi needle under sedation and local dren from Asia and eastern Europe were adopted as young anesthesia, by a percutaneous intercostal or subcostal approach.
infants. The Asian children tended to have been adopted at a Statistical analysis.—For statistical analysis of data, we used the younger age, although the difference from the other groups was Pearson x2 test. We compared differences between groups by use not statistically significant. The higher proportion of girls in of unpaired Student t tests, adjusted for the inequality of variances the study population was mainly due to the large number of when within-group variances were statistically different according girls in the Asian group, a phenomenon directly related to the to the Brown-Forsythe test. For continuous variables, we used the adoption policies in the countries of origin. The mode of trans- Mann-Whitney nonparametric test. P ! .05 was considered statis- mission was classified as unknown when clear documentation was unavailable. Because of the high adoption rates in the For preliminary examination of the survival distribution across groups, we used Kaplan-Meier survival curves, which were com- Asian and eastern-European groups, the mode of transmission pared by log-rank test. Since patients were recruited into the study for the majority of children in these groups was not reliably at various ages, we then used regression survival analysis based on documented. However, most Chinese and Vietnamese infants the proportional-hazards model (Cox regression survival analysis), have been infected by vertical transmission, whereas eastern- Hepatitis B e antigen (HBeAg) seroconversion in hepatitis B virus–infected children, by birth Total seroconversions, no. (%) of total treated Ab, antibody; HBs, hepatitis B surface.
Figure 1.
Cumulative proportion of children from different ethnic groups maintaining hepatitis B e antigen (HBeAg) positivity during follow- up as calculated by the Kaplan-Meier test in the total series of HBeAg-positive cases and adjusted for age at diagnosis. P p .01, Asian vs. allother groups. ⅷ, Asian; ⅜, all other groups.
European adoptees tend to have been infected during the first Twenty-seven children (15%) were treated with either IFN- a (n p 19) or lamivudine (n p 8). Only 6.6% of Asian children All children were asymptomatic at presentation and remained were treated, compared with 35% of Canadian children, partly so throughout follow-up. Eight patients, distributed evenly because of their lower aminotransferase levels and partly be- among the groups, were coinfected with HCV. No patient was cause of our policy of not treating children with IFN who are coinfected with delta virus. One patient had glomerulonephritis.
No child presented with HCC, nor did any develop HCC during served in 70 (40.2%; table 4) of the children. Of 147 untreated Of the children studied, 96% were HBeAg positive at entry children, 51 (34.7%) seroconverted to anti-HBe–positive status, into the study; however, the 6 HBeAg-negative children were compared with 19 (70.4%) of the treated children; 12 patients found to be HBeAg positive in another medical center (table seroconverted during therapy, and 7 seroconverted a mean of 2). Elevated (12 times normal) serum ALT levels at presentation 14 months after therapy (range, 2–30 months). Asian children were less common in the Asian group (39.9%) than in the other showed significantly lower spontaneous seroconversion rates groups (eastern Europe, 66.7%, P p .04; Canadian, 58.8%, not than did the eastern-European children. When only children significant [NS]; other, 50%, NS). Of the 123 patients in whom with ALT levels 12 times normal were considered, there was HBV DNA testing was performed, 75% had detectable levels; no seroconversion-rate differences between groups.
no difference was observed among groups. Serum a-fetoprotein Figure 1 and figure 2 demonstrate the product-limit estimates testing was done in 90% of patients, and 90% of these had of the proportion of subjects who remained HBeAg positive normal levels. In the 9 patients with abnormal (19.1 mg/L) during follow-up, expressed in terms of time after diagnosis.
serum a-fetoprotein concentrations, the levels were only mar- Asian children (figure 1) seroconverted significantly more ginally above normal, and all became normal during the first slowly than did all other children. By 13 years after diagnosis, 6 months of observation, regardless of seroconversion status.
75% of Asian children, compared with 94% of all others, had Liver biopsy was done in 41 children, in most cases to de- seroconverted (P ! .05). Although treated children showed a termine their eligibility for treatment within clinical trials (table tendency to seroconvert faster, there was no statistical difference 3). Compared with the other groups, more patients in the Ca- between the two groups (figure 2). In this study, 75% of un- nadian group were biopsied, mainly because of the higher pro- treated children, versus 86% of treated children, had HBeAg portion of children with abnormal (85.3%, vs. 63%, 66.6%, and seroconversion by 13 years after diagnosis.
61.4%) serum ALT levels. Results of studies of liver histology The mean ALT level at presentation did not differ between showed minimal or mild inflammation in 80.5% of biopsy sam- children who showed spontaneous seroconversion (244 ע 397 ples and mild or no fibrosis in 90.3% of biopsy samples. Only IU/L) and those who showed treatment-induced seroconversion 1 Asian child had histologic evidence of cirrhosis.
(282 ע 444 IU/L). However, children who did not seroconvert Figure 2.
Cumulative proportion of children maintaining hepatitis B e antigen (HBeAg) positivity during follow-up as calculated by the Kaplan- Meier test in the total series of HBeAg-positive cases, in patients treated with either interferon or lamivudine (ⅷ) and in untreated patients (⅜).
No statistically significant difference was seen (P p presented with significantly lower ALT levels (44 ע 41 IU/L; failure or signs of portal hypertension during the follow-up The mean HBV DNA level was not significantly period of р16 years. Of the children with chronic HBV infec- different between the groups. The mean time to normalization tion, 90% had normalized ALT levels before reaching adult- of ALT (5.8 ע 2.5 months) in untreated patients who sero- hood. These data are similar to those that Bortolotti et al. [12] converted paralleled the time to disappearance of HBV DNA reported in children of Mediterranean origin who had been (5.38 ע 3.4 months) and was not significantly different from infected mainly by horizontal transmission.
the mean time to normalization of ALT in treated patients (4.4 Two main transmission patterns with geographic implica- tions have been described in the literature. Perinatal infection predominates in Asia, where most HBsAg-positive mothers served in 15 patients (8.6%). Treatment was associated with have circulating HBeAg, whereas in Africa, eastern Europe, seroconversion in 14.8% of children, whereas spontaneous se- and the Mediterranean basin, where the proportion of HBeAg- roconversion occurred in 7.5%. The mean time to HBsAg se- positive women is much lower, infection is transmitted during roconversion in treated children was 31 months, and that in infancy and childhood, mainly by chronically infected family untreated children was 4.2 ע 3.1 years.
members and playmates [13, 14] and by therapeutic injections Figure 3 shows the cumulative proportion of HBsAg per- via improperly sterilized needles [15–17]. Thus, although the sistence during follow-up, both in untreated children and in mode of transmission in many of the Asian-born children in children treated with IFN-a or lamivudine. We found no sta- the present study was not documented, it can be reasonably tistically significant difference in percentage of HBsAg sero- assumed that the disease in most of these children has been conversion, on the basis of either ethnic group or sex.
transmitted perinatally. This hypothesis is strengthened by our After follow-up for 1–8 years, children with proven vertically observation that Asian children had lower aminotransferase transmitted disease seemed to seroconvert HBsAg more slowly levels and seroconverted anti-HBe significantly more slowly than did children with other modes of transmission. However, than did all other patients, particularly those from eastern Eu- because of the small size of the sample, the results were not rope. Perinatal infection is associated with prolonged immune statistically significant (data not shown).
tolerance to HBV and is characterized by an absence of symp-toms, normal or slightly elevated ALT activity, HBeAg posi-tivity, and high levels of circulating HBV DNA [7].
Discussion
Although the Asian-born children seroconverted to anti-HBe This report has described the long-term outcome of chronic more slowly than did all other children, the spontaneous-se- hepatitis B in a heterogeneous group of children of different roconversion rate was 24% during the follow-up period, and ethnic origins who lived in the Province of Quebec. We observed the cumulative seroconversion to anti-HBe by 13 years after a survival rate of 100%, and no subject developed either liver diagnosis was as high as 75%. These data contrast with a se- Figure 3.
Cumulative proportion of children maintaining hepatitis B surface antigen (HBsAg) positivity during follow-up as calculated by the Kaplan-Meier test in the total series of HBsAg-positive cases, in patients treated with either interferon or lamivudine (ⅷ), and in untreated patients(⅜) (P p ) roconversion rate of only 7% over 4 years in Asians born and patients who were not treated. There was no statistically sig- living in Asia [7] but are comparable to those reported by Evans nificant difference in the cumulative proportion of treated chil- et al. [18], who found that 13% of Asians living in the United dren who seroconverted, compared with untreated children.
States seroconverted to anti-HBe during a median follow-up However, there was an acceleration of anti-HBe seroconversion period of 25 months and who predicted that 190% of HBV of at most 3 years in treated patients (figure 2); the population carriers infected at birth would become HBeAg negative by age that we studied comprises a large proportion of chronic hep- 50 years. Such differences in long-term outcome in patients with atitis B–infected children who were likely to have been vertically the same ethnic origin but living in different countries have infected. Similar findings have been reported in children in- been described for other aspects of hepatitis B. Thus, although fected mainly by horizontal transmission [19, 20].
long-term studies of Asian individuals suggest a high risk of These data, along with those of other pediatric studies in HCC development among patients infected at birth, Villeneuve which no difference, in the long-term HBeAg seroconversion et al. [6] reported a very low risk of development of HCC in rate, between treated and untreated children was observed [12, individuals living in the Montreal area, even among Asian pa- 19–21], raise important questions about the benefits and timing tients. These differences might be attributed to environmental of therapy in children with chronic hepatitis B. It remains to influences on either the host or the virus.
be shown whether acceleration of HBeAg seroconversion in- One possible environmental factor influencing the immune fluences the development of complications, such as liver cir- response to HBV is the nutritional status of the child. We ob- rhosis or HCC. Liver cirrhosis was observed in only 1 patient served that most Asian children adopted by Canadian families and was an early complication, rather than the end point of were initially malnourished but improved rapidly. There may long-lasting liver disease. Cirrhosis in children with hepatitis B also be a role for viral coinfection. For example, in Asian coun- usually affects young boys and follows early biochemical re- tries, the high population density might favor the transmission mission and clearance of HBeAg [22–24]. Only longitudinal of herpesviruses, including cytomegalovirus and human her- studies of liver histology can determine whether early treatment pesvirus 6, which are common and have immunosuppressive properties. Possibly, infection by such agents at an early age Biochemical remission of hepatitis was associated with loss may contribute to the prolonged immunotolerant phase of hep- of serum HBV DNA in most patients, independent of ethnic atitis B infection. Finally, the susceptibility to HCC in HBV- origin and type of treatment. However, in 8.5% of children who infected patients may be influenced by dietary aflatoxin ex- seroconverted to anti-HBe, HBV DNA remained elevated, even posure, which is particularly common in Asia.
though the semiquantitative analysis was by dot-blot hybridi- The natural history of hepatitis B was observed in the 147 zation, which is less sensitive than polymerase chain reaction (PCR). It has been observed that, with the use of the more B e antigen-positive chronic hepatitis B: implications for interferon ther- sensitive PCR technique, a minimal level of HBV replication apy. J Infect Dis 1997; 176:845–50.
9. Bortolotti F, Jara P, Crivellaro C, et al. Outcome of chronic hepatitis B in may persist in at least half of children for у10 years after anti- Caucasian children during a 20-year observation period. J Hepatol 1998;
HBe seroconversion and biochemical remission [21]. The long- term implication of this finding remains unclear.
10. Sokal EM, Conjeevaram H, Roberts E, et al. Interferon alpha therapy for Twenty years after the availability of hepatitis B vaccine, we chronic hepatitis B in children: a multinational randomized controlled are still far from worldwide eradication of the disease. Until trial. Gastroenterology 1998; 114:988–95.
11. Sokal EM, Kelly D, Mizerski J, et al. An international double blind placebo universal vaccination becomes a reality, the long-term compli- controlled trial of lamivudine in 286 children with chronic hepatitis B cations of HBV disease—particularly cirrhosis and HCC—re- [abstract 40]. J Pediatr Gastroenterol Nutr 2001; 32:390.
main significant problems. Although early treatment with either 12. Bortolotti F, Cadrobbi P, Crivellaro C, et al. Long-term outcome of chronic lamivudine or IFN seems to be an attractive approach to al- type B hepatitis in patients who acquire hepatitis B virus infection in teration of the course of chronic hepatitis B, the benefits of childhood. Gastroenterology 1990; 99:805–10.
13. Stevens CE, Neurath RA, Beasley RP, Szmuness W. HBe Ag and Anti-HBe universal therapy remain to be proven. Children with elevated detection by radioimmunoassay: correlation with vertical transmission of aminotransferases seroconvert spontaneously, in a period of hepatitis B virus in Taiwan. J Med Virol 1979; 3:237–41.
time that is similar that seen for children receiving treatment.
14. Tabor E, Bayley AC, Cairns J, Pollen L, Gerty RJ. Horizontal transmission Overall, treatment affords only an acceleration of 2–3 years in of hepatitis B virus among children and adults in five rural villages in HBeAg seroconversion and only a slight advantage in HBsAg Zambia. J Med Virol 1985; 15:113–20.
15. DeVoid DE, Pineiro-Carrero VM, Goodman Z, Latimer JS. Chronic active seroconversion. Long-term follow-up of children in randomized hepatitis B infection in Romanian adoptees. J Pediatr Gastroenterol Nutr clinical trials would answer this central question more reliably 1994; 19:431–6.
than would analysis of heterogeneous populations. The ulti- 16. Kane M. Unsafe injections. Bull World Health Organ 1998; 76:99–100.
mate impact of a 3-year acceleration in seroconversion may be 17. CDC. Frequency of vaccine-related and therapeutic injections—Romania, important but, as yet, remains unknown.
1998. MMWR Morb Mortal Wkly Rep 1999; 48:271–4.
18. Evans AA, Fine M, London WT. Spontaneous seroconversion in hepatitis B e antigen-positive chronic hepatitis B: implications for interferon ther- References
apy. J Infect Dis 1997; 176:845–50.
19. Ruiz-Moreno M, Camps T, Jimenez J, et al. Factors predictive of response to 1. Kane MA. Global status of hepatitis B immunization. Lancet 1996; 348:696.
interferon therapy in children with chronic hepatitis B. J Hepatol 1995; 22:
2. Chen HL, Chang MH, Ni YH, et al. Seroepidemiology of hepatitis B virus infection in children: ten years of mass vaccination in Taiwan. JAMA 20. Barbera C, Bortolotti F, Crivellaro C, et al. Recombinant interferon alpha 1996; 276:906–8.
2 a hastens the rate of HBeAg clearance in children with chronic hepatitis 3. Chang MH, Shau WY, Chen CJ, et al. Hepatitis B vaccination and hepa- B. Hepatology 1994; 20:287–90.
tocellular carcinoma rates in boys and girls. JAMA 2000; 284:3040–2.
21. Bortolotti F, Wirth S, Crivellaro C, Alberti A, Martine U, de Moliner L.
4. Lok AS, Heathcote EJ, Hoofnagle JH. Management of Hepatitis B: 2000.
Long-term persistence of hepatitis B virus DNA in the serum of children Summary of a workshop. Gastroenterology 2001; 120:1828–53.
with chronic hepatitis B after hepatitis B e antigen to antibody serocon- 5. El-Serag H, Mason AC. Rising incidence of hepatocellular carcinoma in the version. J Pediatr Gastroenterol Nutr 1996; 22:270–4.
United States. N Engl J Med 1999; 340:745–50.
22. Bortolotti F, Calzia R, Cadrobbi P, et al. Liver cirrhosis associated with 6. Villeneuve JP, Desrochers M, Infante-Rivard C, et al. A long-term follow-up chronic hepatitis B virus infection in childhood. J Pediatr 1986; 108:244–7.
study of asymptomatic hepatitis B surface-antigen-positive carriers in 23. Hsu HY, Chang MH, Lee CH, Hsu HC, Chen DS. Spontaneous loss of Montreal. Gastroenterology 1994; 106:1000–5.
HBsAg in children with chronic hepatitis B virus infection. Hepatology 7. Lok ASF, Lai CA. A longitudinal follow-up of asymptomatic hepatitis B 1992; 15:382–6.
surface antigen-positive Chinese children. Hepatology 1988; 8:1130–3.
24. McMahorn BJ, Alberts SR, Wainwright RB, Bulkow L, Lanier AP. Hepatitis 8. Evans AA, Fine M, London WT. Spontaneous seroconversion in hepatitis B related sequelae. Arch Intern Med 1990; 150:1051–4.

Source: http://www.lemondeestailleurs.ca/Notre_parole/documents/LongTermFollowupChronicHepatitisBVirusInfectionArticleUofMChicoineJul2002.pdf