Spontaneous Loss of HBsAg in Children with Chronic Hepatitis B Virus Infection HONG-YUAN HSU,l MEI-HWEICHANG,~ CHIN-YUN LEE,' JUEI-SANC H E N ,HEY-CHI ~ H S U AND ~ DING-SHINN CHEN4 Departments of 'Pediatrics, 'Clinical Pathology and 3Pathology and 4Graduate Institute of Clinical Medicine, College of Medicine, National Taiwan University, Taipei, Taiwan, Republic of China

Spontaneous loss of HBsAg is infrequent in adult HBV carriers. Little is known about this serological change in children. In a prospective study of 420 hepatitis B virus-carrier children who were observed for 1 to 12 yr (mean = 4.3 yr), spontaneous loss of HBsAg occurred in 10 patients, with an average incidence of 0.6%/yr.The HBsAg clearance rate was significantly higher in children who had anti-HBe; children who were at an older age on entry; children whose mothers were HBsAg-; or children with severe liver histological changes detected while they were HBeAg+ Children who seroconverted from HBeAg to anti-HBe before the age of 6 or who had a peak serum ALT level above 100 IU/L were more likely to clear HBsAg. In all 10 patients who became HBsAg- ,serum hepatitis B virus DNA became undetectable by both spot hybridization and the polymerase chain reaction, suggesting a complete clearance of the v i r u s from serum. After the loss of HBsAg, the anti-HBslevelswere higher in the children born to carrier mothers than in those born to noncarrier mothers. These findings suggest that chronic hepatitis B virus-carrier children rarely lose HBsAg, especially if they have been infected during the perinatal period and have mild histological changes. The poor humoral immune response to HBsAg may be a contributing factor in the establishment of carrier status during horizontal infection but may not be primarily involved in the establishment of carrier status during perinatal infection. (HEPATOLOGY 1992;

.

15:382-386.)

Patients with chronic HBV infection with high levels of HBV and HBeAg in serum can spontaneously lose HBeAg and develop anti-HBe, a seroconversion that usually proceeds a clinical and histological remission in disease (1-3). After seroconversion, most patients remain HBsAg+ with a low level or no detectable HBV. In contrast, loss of HBeAg and HBsAg rarely occurs in chronic HBV carriers. The annual rate of clearance of

Received November 16, 1990; accepted August 5, 1991. This study was supported by Grant NSC 78-0419-B002-134 from the National Science Council, Republic of China. Address reprint requests to: Mei-Hwei Chang, M.D., Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan, R.O.C.

31/1/34686

HBsAg in adult HBV carriers was reported to be 1%to 1.7% in two studies from the United States (4,5 ) and 0.5% in chronic hepatitis and 0.8% in asymptomatic carriers in one recent study from Taiwan (6). In the few long-term follow-up studies involving smaller series of chronic HBV carrier children, spontaneous loss of HBsAg was noted in only one report from Italy (7-9).In that study, the patients primarily were children with CAH, and HBsAg clearance occurred more frequently in carrier children with previous acute hepatitis (9).In HBV carriers, anti-HBs can develop after the loss of HBsAg (4,61, suggesting that the original immunological nonresponsiveness to HBsAg was overcome. Seroconversion to anti-HBs after acute HBV infection generally indicates recovery and long-term immunity to HBV (10). Whether a similar clinical significance of this serological change is implicated in chronic HBV infection remains unclear. This study investigated the incidence, contributing factors and clinical implications of spontaneous loss of HBsAg in a large group of HBV carrier children who were undergoing a prospective, long-term serological, biochemical and clinical follow-up. PATIENTS AND METHODS A total of 420 children (251 boys and 169 girls) who had serum HBsAg for at least 1 yr were studied. Children were recruited through two sources: (a) 331 were identified during serological surveys conducted in two well-baby clinics, three kindergartens, two primary schools and one junior middle school, or by random screening of infants born to HBsAg carrier mothers or siblings of known carrier children. Their initial mean ALT level was 22 IU/L (range = 1 to 463 IUIL), and the peak mean ALT level during follow-up was 45 IU/L (range = 13 to 813 IU/L); (b) 89 children were referred for evaluation of liver disease and HBsAg positivity. Their initial mean ALT was 98 IU/L (range = 5 to 706 IU/L), and the peak mean ALT level was 176 IU/L (range = 9 to 1,074 IU/L). At enrollment, the 420 patients ranged from less than 1 mo to 17 yr of age. Eighty-nine children were younger than 12 mo old, 39 were between 1and 3 yr old, 90 were between 3 and 6 yr old, 122 were between 6 and 10 yr old and 80 were 10 yr old or older. None had received antiviral or immunosuppressive therapy. Patients were seen every 3 to 6 mo or at shorter intervals at the Department of Pediatrics, National Taiwan University

382

383

CLEARANCE OF HBsAg IN HBV CARRIER CHILDREN

Vol. 15, No. 3, 1992

TABLE 1. HBsAg clearance in HBV carrier children according to HBeAdanti-HBe status at entry HBeAg/Auti-HBe at entry

+I-I+ -ITOTAL

No. of cases

375b 42b 3 420

Age at entry mean S.D.(yr)

Follow-up duration mean S.D. (mo)

5.5 t 4.2' 9.1 t 4.4" 9.9 c 4.0 5.9 t 4.4

51.7 t 29.8 49.8 t 22.6 25.8 2 0.6 51.4 i 22.6

*

*

"Expressed as 95% confidence interval. bKaplan-Meier life table method and generalized Wilcoxon's rank sum test, p 'Kruskal-Wallis test, p < 0.0001.

=

Loss of HBaAg No. (%)"

7 (1.9 t 1.4)' 3 (7.1 t 7.8)h 0 (0) 10 (2.4 2 1.5)

Incidence (%/yr)

0.4 ? 0.3 1.7 ? 1.9 0 0.6 t 0.4

0.03.

Hospital. At each visit, patients underwent physical exami- a reaction mixture without HBV DNA served as negative nation and had blood drawn for conventional liver biochemical controls. The amplified DNA was detected by ethidium tests and serum HBV markers. Of the 420 patients, 85 bromide staining after 3% agarose gel electrophoresis and by underwent a liver biopsy (after informed consent was obtained Southern-blot analysis. All serum samples were repeatedly from the parents) primarily because of the findings of an tested by PCR from the extraction of serum HBV DNA to elevated ALT level (mean = 144 IU/L, range = 29 to 1,074 exclude the possibility of poor DNA preparation. For the 10 patients with HBsAg clearance, detection of IU/L) with or without symptoms such as being easily fatigued, poor appetite, jaundice or hepatosplenomegaly. Liver histo- anti-HBc IgM and antibody to hepatitis D antigen (anti-HD) pathological diagnosis was made according to the revised were performed as previously described (12). Antibody to criteria of the International Group (111, with slight modifi- hepatitis C virus (anti-C100-3) was detected by enzyme cation (12). These criteria included nonspecific reactive hepa- immunoassay (Abbott, anti-HCV EIA). The Kaplan-Meier life table method and the generalized titis (NSRH), chronic persistent hepatitis (CPH), CAH or cirrhosis. Serum ALT, AST and bilirubin levels were tested by Wilcoxon's rank sum test were used to estimate the HBsAg an autoanalyzer. Serum HBV markers, including HBsAg, clearance rate (17).A p value of less than 0.05 was considered anti-HBs, anti-HBc, HBeAg and anti-HBe were determined significant, and a p value between 0.05 and 0.1 was considered using commercial RIAs (Abbott Laboratories, North Chicago, as showing a trend. IL). Mothers of 370 carrier children were also tested for serum HBV markers and underwent conventional liver tests. The RESULTS anti-HBs levels were expressed in milli-international units per milliliter by the serial dilution method with reference to the Among the 420 children, 375 (89.3%)had HBeAg, 42 World Health Organization anti-HBs standard. (10%)had anti-HBe and 3 (0.7%)were negative for both In patients who lost HBsAg, serum levels of HBV DNA were markers at entry. All 420 patients were observed for 1to serially assayed by spot hybridization (13) with minor modi- 12 yr (mean = 51.4 mo). fications (14).The lower limit of sensitivity of measuring HBV During follow-up, 90 (24%) of 375 patients with DNA was 0.5 pg/sample. at entry became HBeAg - . HBeAg Polymerase chain reaction (PCR) was also used for detection At the end of follow-up, only 10 (2.4%) children were of serum HBV DNA at 6 , 1 2 , 1 8 and 24 mo after clearance of HBsAg. Serum samples (10 kl) were digested using proteinase HBsAg- . The overall HBsAg clearance rate in all 420 K. HBV DNA was extracted by phenol/chloroform followed by carrier children was 0.6%/yr. The annual HBsAg precipitation with ethanol in the presence of ammonium clearance rate in anti-HBe patients was significantly acetate. Two sets of oligonucleotide sequences specific for HBV higher than those with HBeAg. The difference in age at surface and core regions were synthesized and used as primers. entry between children who were initially HBeAg or Primer pair A was within the HBV surface antigen gene (15): anti-HBe was also significant (Table 1). primer (5 ' -TTAGGGTTTAAATGTATACCC-3') at map position Loss of HBsAg subsequent to HBeAg clearance oc824-844 and primer (5'-CATCTTCTTGTTGGTTCTTCTG-3') curred in 1 (20%) of 5 children who cleared HBeAg at map position 429-450 of the HBV genome. Primer pair B was specific for the core region (16): primer (5'- before 3 yr of age, in 4 (16.7%)of 24 children who cleared CTGGGAGGAGTTGGGGGAGGA-3') at map position 1730- HBeAg between 3 and 6 yr of age and in 2 (3.3%)of 61 at children who cleared HBeAg after 6 yr of age. Thus the 1750 and primer (5'-CTAACATTGAGATTCCCGAGA-3') map position 2438-2458. Each reaction tube contained DNA frequency of clearance of HBsAg among children who purified from 10 p1 of serum in 10 mmoUL Tris (pH 8.3), 50 cleared HBeAg before 6 y r of age was 17.2% (5 of 29) mmol/L potassium chloride, 1.5 mmol/L magnesium chloride, compared with 3.3% (2 of 61) among those who cleared 1 nmole of each primer, 12.5 nmole of each nucleotide HBeAg older than 6 yr of age (p = 0.075, Kaplan-Meier triphosphate and 2 units of thermostable Taq polymerase life table method and generalized Wilcoxon's rank sum (Perkin-Elmer Cetus, Perkin-Elmer Corp., Norwalk, CT). test). Forty cycles of amplification were performed in an automatic The effect of sex, age at entry, source of subject DNA thermal cycler (Perkin-Elmer Cetus). Each cycle conrecruitment, maximal serum ALT level while HBeAg+ , sisted of 1 min at 94" C, 2 min at 50" C and 2 min at 72" C, maternal HBsAg status and liver histopathological except the last cycle in which primer extension at 72" C was extended to 7 min. In each run, DNA extracted from HBeAg-t results on the loss of HBsAg are shown in Table 2. serum and cloned HBV DNA served as positive controls, and Carrier children who were older at entry or whose DNA extracted from serum negative for all HBV markers and mothers were HBsAg - had a significantly higher

+

+

384

HEPATOLOGY

HSU ET AL.

TABLE2. Predictive factors for the loss of HBsAg in HBV carrier children Factors

Age at entry (yr) 0-3 4-6 7-9 2

10

Sex Male Female Recruitment source Referred patients Screening Peak ALT level" 5 100 IUiL > 100 IUiL Maternal HBsAg Positive Negative Liver histological result

Loss of HBsAg

No. of cases

Follow-up duration mean & S.D. (mo)

154 100 86 80

61.2 i- 33.3 49.0 t 26.4 44.4 i- 25.2 41.3 rt 21.9

4 (2.6 i- 2.5) l ( 1 . 0 -t 2.0) 2 (2.3 t 3.2) 3 (3.8 i- 4.2)

< 0.01

251 169

50.9 i- 30.7 52.0 t 26.9

8(3.2 i- 2.2) 2 (1.2 2 1.6)

0.45

89 33 1

55.8 ? 27.4 49.7 i- 29.6

4 (4.5 i- 4.3) 6(1.8 i- 1.4)

0.21

309 66

50.4 t 29.3 55.9 2 28.2

3 (1.0 t 1.1) 4 (6.1 t 5.8)

0.07

244 126

54.4 ? 32.2 51.2 ? 51.2

2 ( 0 . 8 t 1.1) 7 (5.6 t 4.0)

< 0.01

11 67

55.4 t 17.0 77.8 f 41.5

3 (27.3 2 26.3) l(1.5 rt 2.9)

CAH CPH, NSRH or NH

No. (%)"

p valueb

< 0.0001

NH = normal histological result. "Expressed as 95% confidence interval. bKaplan-Meierlife table method and generalized Wilcoxon's rank sum test. 'While HBeAg+ . "Among 78 patients who underwent liver biopsy while HBeAg+ .

incidence of HBsAg clearance. Consistently, we found that maternal HBsAg positivity was significantly less among the carrier children who became HBsAg- than among those who did not; the number was 2 of 9 vs. 242 of 361 (p = 0.036, x2 square test with Yates' correction). Those children whose mothers were HBsAg- had a significantly higher mean age at entry than those whose mothers were HBsAg+ (8.3 ? 3.9 vs. 4.4 rtr 4.1, mean -+ S.D., p < 0.0001, Kruskal-Wallis test). When all 370 children whose maternal HBsAg status was known were stratified by initial age, the comparison of HBsAg clearance between those whose mothers were HBsAg- and those whose mothers were HBsAg+ resulted in the following: 4 of 36 vs. 1 of 160 for those children with an initial age under 6 yr old (p < 0.01) and 3 of 90 vs. 1 of 84 for those children with an initial age of more than 6 yr old (p = 0.53) (Kaplan-Meier life table method and generalized Wilcoxon's rank sum test). Among the 85 patients who had liver biopsies, 78 underwent a liver biopsy while HBeAg . A significant correlation was seen between the severity of histological changes and the probability of clearance of HBsAg; we found 3 (27%)of 11 patients with CAH vs. 1(1.5%)of 67 patients with mild histological changes (CPH in 31 patients, NSRH in 33 patients and normal histological results in 3 patients) became HBsAg- (Table 3). Only nine liver specimens were available from patients who were anti-HBe+: one of two patients with severe histological results as opposed to three of seven patients with mild histological results lost serum HBsAg. Too few cases were available to analyze these differences.

+

The ten patients who cleared HBsAg lost this marker between 4.5 to 21 yr of age (Table 3). Seven patients lost HBsAg after the age of 10. The interval between loss of HBeAg and HBsAg in seven patients who were initially HBeAg+ ranged from 6 to 103 mo. Follow-up on all 10 patients extended from 2 to 4 yr after clearance of HBsAg. By both spot hybridization and PCR, serum HBV DNA remained undetectable in all 10 patients at 6, 12,18 and 24 mo after the loss of HBsAg. Two years after HBsAg clearance, both patients born to carrier mothers developed anti-HBs levels above 70 mIU/ml. In contrast, of the seven patients born to noncarrier mothers, the level of anti-HBs attained was less than 30 mIU/ml in all patients and below 10 mIU/ml in five patients. All 10 patients who cleared HBsAg have remained well, with normal liver tests and no symptoms of liver disease. None have undergone a repeat liver biopsy. None of the 10 were seropositive for anti-HBc IgM, anti-HDV or anti-HCV. DISCUSSION

In HBV carriers who are infected during adulthood, especially in Western countries, clearance of the carrier state and loss of HBsAg is uncommon but does occur, either spontaneously (4) or as a result of antiviral therapy (18, 19). The findings in this study show that carrier children are even less likely to clear HBsAg in the absence of treatment. Among 420 chronic HBV carrier children, the average rate of loss of HBsAg was O.G%/yr. The 10 patients who cleared HBsAg during follow-up were known to be HBsAg+ for 3 to 9 yr. Thus none were

Vol. 15, No. 3, 1992

385

CLEARANCE OF HBsAg IN HBV CARRIER CHILDREN

TABLE 3. Characteristics of 10 HBsAg carrier children who lost serum HBsAg on follow-up At entry Patientno.

Sex ~

1 2 3 4 5 6 7 8 9 10

Age (yr)

HBeAg

_____

~

M F M M M M M F M M

7 9 4.1 2 0.2 3.8 2.3 10.6 12.7 10.7

Age at clearance

Peak ALT“ (IU/L)

+ + + + + + + -

Maternal HBsAg

~_____

300 198 173 15 55 244 15 6 29 13

Liver histological result _______

-

-

+ -

-

+ ND -

_______

CAH CAH CPH ND CPH CAH ND ND CIR ND

HBeAg

HBsAg

After HBsAg loss Anti-HBs (mIUim1)

~_______

10 9.3 5.7 3.3 3 5.8 4 ? ? ?

HBV DNA by PCR

__________

13 10.8 11.2 6.8 11.6 6.3 4.5 13.6 21.7 13.3

1 28 7 1 95 0 17 76 51 4

~~

-

-

ND = not done; CIR = cirrhosis of liver. “Normal = < 17 IUIL. bDuring HBeAg+ phase except for patients 5 and 9.

in the antigenemic phase of acute hepatitis B infection. This study also showed that the clearance of HBsAg was associated with four clinical variables: absence of carrier state in the mother, older age on initial evaluation, presence of anti-HBe on initial evaluation and the severity of liver disease (as judged from peak ALT activity or liver histological results) during the early phases of infection. Newborns of chronic carrier mothers may have a deficient cellular immunity and be exposed to a relatively large inoculum of virus transmitted from highly infective mothers; they therefore may be more prone to having the virus persist than carrier children who have exposure to HBV at a later time. In support of this hypothesis, we observed that carrier children born to carrier mothers cleared HBsAg much less frequently than those born to noncarrier mothers. The effect of maternal HBsAg status on HBsAg clearance became less evident for those older than 6 yr of age at entry because it was overshadowed by other factors, including increasing age and increasing probability of anti-HBe + status among these carrier children. CAH occurring while a patient is HBeAg+ usually indicates a vigorous immune reactivity and attempt at clearance of HBV-infected hepatocytes (18). Loss of HBeAg and HBsAg, therefore, would reasonably be more likely to occur in this situation, particularly in patients who clear HBsAg shortly after HBeAg. In contrast with the “hepatitis-like illness” noted in many adult carriers who cleared both HBeAg and HBsAg after treatment with interferon (191, the interval between the clearance of HBeAg and HBsAg was usually quite long in our carrier children who cleared HBsAg spontaneously. Of the four patients who cleared HBsAg but who had constantly normal ALT levels during the follow-up period, two had seroconverted to anti-HBe when they were tested. In most carrier children, seroconversion from HBeAg to anti-HBe is usually followed by a sustained biochemical and histological remission (3, 9). The explanation for the remaining two patients who

were initially HBeAg+ is more difficult. Perhaps a peak ALT level may have been missed at the time of sampling. We found that the loss of HBsAg was more likely to occur in carrier children who were older or who were anti-HBe+ at entry. Because the onset of the carrier state among Taiwanese children occurred perinatally or during early childhood (201, the duration of the carrier state was certainly longer in carrier children who were older or who were anti-HBe + at entry. The probability of clearing HBsAg, therefore, increased with the duration of carrier status. Children under 6 yr of age who seroconverted to anti-HBe were more likely to clear HBsAg. It is possible that the loss of HBeAg at a younger age, at a time when HBV DNA exists largely in a nonintegrated form, may eliminate almost all hepatocytes supporting a replicating virus and minimize the number of remaining hepatocytes with integrated HBV DNA that could express HBsAg. An age-related maturation in the host immune system may be more likely to mount a response to completely eradicate the few hepatocytes with HBV DNA integrations, resulting in complete clearance of HBsAg and HBV DNA from the serum. These results also suggest that the probability of clearing HBsAg may be greater if antiviral therapy is given at a younger age. Interestingly, we found that after the loss of HBsAg, higher levels of anti-HBs were found in children born to carrier mothers than in those children born to noncarrier mothers. Recently, the levels of anti-HBs attained after HBsAg loss in adult HBV carriers after interferon-a treatment were shown to be significantly lower than those patients in recovery from acute HBV infection or vaccine responders (21). Obviously, these adult carriers also acquired HBV through horizontal infection. It seems that, whether or not therapeutic induction exists, loss of HBsAg in HBV carriers who are horizontally infected is usually followed by a poor or suboptimal anti-HBs response. The suggestion that a continued low level of HBsAg production by the integrated HBV DNA in the liver exists but is undetectable

386

HSU ET AL.

by current assays is one explanation for the sustained low level of anti-HBs. The other possibility is that a poor humoral immune response to HBsAg may be a contributing factor to the establishment of the carrier state during horizontal infection, whereas this factor may not be involved primarily in perinatal infection. The appearance of appreciable titers of anti-HBs in these carriers implies that the clearance of the virus and the disappearance of HBsAg from the serum may gradually allow the immune system to become responsive to HBsAg.

Acknowledgments: We thank Dr, L.M. Huang, Dr. P.I. Lee and C.C. Chang, MS, for their statistical assistance, and t o Miss K.L. Wu for her assistance in the follow-up of the patients. REFERENCES 1. Realdi G, Alberti A, Gugge M, Bortolotti F, Rigoli AM, Tremolada F, Ruol A. Seroconversion from hepatitis B e antigen to anti-HBe in chronic hepatitis B virus infection. Gastroenterology 1980;79: 195-199. 2. Hoofnagle JH, Dusheiko GM, Seeff LB, Jones EA, Waggoner JG, Bales ZB. Seroconversion from hepatitis B e antigen to antibody in chronic type B hepatitis. Ann Intern Med 1981;94:744-748. 3. Chang MH, Sung JL, Lee CY, Chen CJ, Chen JS, Hsu HY, Lee PI, et al. Factors affecting clearance of hepatitis B e antigen in hepatitis B surface antigen carrier children. J Pediatr 1989;115: 391-396. 4. Sampliner RE, Hamilton FA, Iser OA, Tabor E, Boitnott J. The liver histology and frequency of clearance of the hepatitis B surface antigen (HBsAg) in chronic carriers. Am J Med Sci 1979;277: 17-22. 5. Alward WLM, McMahon BJ, Hall DB, Heyward WL, Francis DP, Bender TR. The long-term serological course of asymptomatic hepatitis B virus carriers and the development of primary hepatocellular carcinoma. J Infect Dis 1985;151:604-609. 6. Liaw YF, Sheen IS, Chen TJ, Chu CM, Pa0 CC. Incidence, determinants and significance of delayed clearance of serum HBsAg in chronic hepatitis B virus infection: a prospective study. HEPATOLOGY 1991;13:627-31. 7. Lok ASF, Lai CL. A longitudinal follow-up of asymptomatic hepatitis B surface antigen-positive Chinese children. HEPATOLOGY 1988;8:1130-1133.

HEPATOLOGY

8. Ruiz-Moreno M, Camps T , Aguado JG, Porres JC, Oliva H, BartolomB J, Carreno V. Serological and histological follow up of chronic hepatitis B infection. Arch Dis Child 1989;64:11651169. 9. Bortolotti F, Cadrobbi P, Crivellaro C, Guido M, Rugge M, Noventa F, Calzia R, et al. Long-term outcome of chronic type B hepatitis in patients who acquire hepatitis B virus infection in childhood. Gastroenterology 1990;99:805-810. 10. Krugman S. Viral hepatitis, type B: studies on natural history and prevention re-examined. N Engl J Med 1979;300:101-106. 11. International group. Acute and chronic hepatitis revisited. Lancet 1977;2:914-919. 12. Hsu HY,Chang MH, Chen DS, Lee CY. Hepatitis D virus infection in children with acute or chronic hepatitis B virus infection in Taiwan. J Pediatr 1988;112:888-892. 13. Scotto J, Hadchouel M, Hery C, Yvart J, Tiollais P, Brechot C. Detection of hepatitis B virus DNA in serum by a spot hybridization technique: comparison with results for other markers. HEPATOLOGY 1983;3:279-284. 14. Lee PI, Chang MH, Lee CY, Hsu HY,Chen JS, Chen DS. Changes of serum hepatitis B virus DNA and aminotransferase levels during the course of chronic hepatitis B virus infection in children. HEPATOLOGY 1990;12:657-660. 15. Thiers V, Nakajima E, Kremsdorf D, Mack D, Schellekens H, Driss F, Goudeau A, et al. Transmission of hepatitis B from hepatitisB-seronegative subjects. Lancet 1988;2:1273-1276. 16. Carman WF, Tacyna MR, Hadziyannis S, Karayiannis P, McGarvey MJ, Makris A, Thomas HC. Mutation preventing formation of hepatitis B e antigen in patients with chronic hepatitis B infection. Lancet 1989;2:588-591. 17. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:459-481. 18. Thomas HC, Lever AM, Scully MJ, PignatelliM. Approaches to the treatment of hepatitis B virus and delta-related liver disease. Semin Liver Dis 1986;6:34-41. 19. Alexander GJM, Brahm J, Fagan EA, Smith HM, Daniels HM, Eddleston ALWF, Williams R. Loss of HBsAg with interferon therapy in chronic hepatitis B virus infection. Lancet 1987;2: 66-69. 20. Hsu HY, Chang MH, Chen DS, Lee CY, Sung JL. Baseline seroepidemiology of hepatitis B virus infection in children in Taipei 1984 a study just before mass hepatitis B vaccination program in Taiwan. J Med Virol 1986;18:301-307. 21. Karayiannis P, Kanatakis S, Thomas HC. Anti-HBs response in seroconverting chronic HBV carriers following alpha-interferon treatment. J Hepatol 1990;10:350-352.