905

CONCISE COMMUNICATION

Effectiveness of Hepatitis B Vaccination in Babies Born to Hepatitis B

Surface AntigenPositive Mothers in Italy
Alfonso Mele,1 Francesco Tancredi,3 Luisa Romano`,5
Anna Giuseppone,3 Mario Colucci,6 Aldo Sangiuolo,7
Rosina Lecce,2 Brunella Adamo,4 Maria Elena Tosti,1
Gloria Taliani,8 and Alessandro R. Zanetti5

1
Clinical Epidemiology Unit, Istituto Superiore di Sanita`,
and 2Infectious and Tropical Diseases Unit, Policlinico Umberto I,
Rome, 3S. S. Annunziata Hospital and 4Local Health Unit 1,
Epidemiology Service, Naples, 5Institute of Virology, University
of Milan, Milan, 6Infectious Diseases Unit, S. Giovanni di Dio
Hospital, Salerno, 7G. Rummo Hospital, Benevento, and 8Institute
of Infectious Diseases, University of Florence, Florence, Italy

Viral hepatitis type B is a major health problem worldwide.

Infection may progress to chronic liver disease, including cirrhosis and hepatocellular carcinoma. Age at acquisition of
hepatitis B virus (HBV) infection is a key determinant of chronicity, since the rate of development of the hepatitis B surface
antigen (HBsAg) carrier state is extremely high when infection
occurs in newborns. This declines progressively with increasing
age. Thus, universal vaccination of children during the first year
of life is the most effective strategy to reduce the size of the
carrier reservoir and to control the spread of HBV infection.
Italy is one of the first countries to initiate a policy of universal
and selective high-riskoriented vaccination against hepatitis
B. Since 1983, babies born to HBsAg carrier mothers have been
given hepatitis B immune globulin (HBIG) and vaccine at birth.
In 1991, vaccination became mandatory for all infants and 12year-old adolescents [1].
Hepatitis B vaccination is safe and effective, although breakthrough infections occasionally occur in vaccinees [2, 3]. In
Received 2 April 2001; revised 22 June 2001; electronically published 14
August 2001.
Informed consent was obtained from parents or guardians of all children
enrolled. Human experimentation guidelines were followed in the conduct
of clinical research.
Financial support: Viral Hepatitis Project, Istituto Superiore di Sanita`
(D.leg.vo 30/12/1992 n. 502).
Reprints or correspondence: Dr. Alfonso Mele, Reparto di Epidemiologia
Clinica, Istituto Superiore di Sanita`, Viale Regina Elena, 299, 00161 Rome,
Italy (amele@iss.it).
The Journal of Infectious Diseases 2001; 184:9058
2001 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2001/18407-0013$02.00

addition, HBV mutants with amino acid substitutions, within

the common a determinant of HBsAg, have been identified,
which can potentially escape vaccine-induced immunity [4]. The
most frequent HBV mutant, the so-called G145R, has a single
amino acid substitution of glycinerarginine at position 145 of
the S gene. This was first identified in Naples 110 years ago in
a successfully immunized child who became an HBsAg carrier,
despite the presence of protective levels of antibodies to HBsAg
(anti-HBs) [5, 6].
Whether mutants of the HBV a determinant may pose a
potential threat to immunization programs is still controversial
[4, 69]. To evaluate the outcome of hepatitis B vaccination in
a population at high risk of infection and to investigate further
the need for booster vaccination to sustain immunity, we studied the persistence of antiHBs antibodies and the occurrence
of infections due to wild-type (wt) HBV or to HBV a epitope
mutants in babies born to HBsAg carrier mothers, who, at
birth, were given postexposure prophylaxis with HBIG and
hepatitis B vaccine.

Materials and Methods

Vaccination program. Screening of pregnant women for HBsAg
in the last trimester of pregnancy, to identify babies in need of
passive plus active prophylaxis, was recommended in Italy in 1983
and became mandatory in 1991. In accordance with the affirmation
by Italian health authorities of the essential equivalence of different
vaccine brands, babies born to HBsAg carrier mothers were given
commercial products: HBIG (Biagini) supplemented by plasmaderived vaccines (HEVAC-B, Pasteur; HB-Vax, Merck Sharp &

Downloaded from http://jid.oxfordjournals.org/ by guest on December 26, 2015

This study examined 522 children born to hepatitis B surface antigen (HBsAg)positive mothers from 1985 through 1994 and evaluated the protection provided by antihepatitis B virus
(HBV) immunization at birth. Babies were given hepatitis B immunoglobulin and hepatitis B
vaccine at birth. At 514 years after immunization, 17 children (3.3%) were antiHB core antigen
positive, and 3 also were HBsAg positive. One carrier child had a double mutation, with substitution of prolinerserine at codons 120 (P120S) and 127 (P127S) within the a determinant of
HBsAg. Of the 522 children, 400 (79.2%) of 505 still had protective anti-HBsAg titers 10 mIU/
mL. Thus, HBV vaccination of children born to HBsAg-positive mothers is effective and confers
long-term immunity. There is no evidence that the emergence of HBV escape mutants secondary
to the immune pressure against wild-type HBV is of concern.

906

Mele et al.

JID 2001;184 (1 October)

Table 1. Serologic features of hepatitis B surface antigen (HBsAg) carrier children at study
enrollment who became HBsAg carriers and of their corresponding mothers.
a

Mother-child
pair

Age,
years

ALT,
IU/L

HBsAg

Anti-HBs,
mIU/mL

HBeAg/
anti-HBe

Subtype

HBV strain

1
2
3

27/6
40/14
34/13

30/22
46/29
40/64

/
/
/

/27
/0
/610

///
///
///

ayw / ayw
ayw / ayw
ayw / ayw

wt/wt
wt/wt
wt/mutant (P120S
and P127S)

NOTE. Data are mother/child. All 3 children in this study were boys. Anti-HBe, antibodies to HBeAg;
anti-HBs, antibodies to HBsAg; HBeAg, hepatitis B e antigen; HBV, hepatitis B virus.
a
Normal value of alanine aminotransferase (ALT), 40 IU/L.

Results
The HBe status at delivery was known for 72 mothers: 6
(8.3%) were HBeAg positive (95% confidence interval, 2.3
17.3). Overall, 17 (3.3%) of 522 immunized children acquired
infection, as shown by the presence of anti-HBc alone (n p
14) or with contemporaneous concomitant presence of HBsAg

(n p 3). All 3 carrier children were born to mothers with unknown HBeAg/anti-HBe status. None of the 17 immunized
infected children developed signs or symptoms of acute hepatitis, and none of the 14 children who were anti-HBc positive
but HBsAg negative had detectable HBV DNA.
The 3 HBsAg carrier children were seropositive for HBV
DNA. Two children harbored wt HBsAg of subtype ayw, similar to their mothers, whereas the S gene of the remaining third
child presented a double mutation, with substitution of prolinerserine at codons 120 (P120S) and 127 (P127S) within the
a determinant (table 1). This child was HBeAg positive and
had a high titer of anti-HBs (610 mIU/mL) and slightly altered
alanine aminotransferase (ALT) levels (64 IU/L); his mother
carried wt HBsAg of subtype ayw and was positive for antiHBe, with normal ALT levels.
As shown in table 2, 400 (79.2%) of the 505 uninfected children had anti-HBs antibodies at levels considered to be protective (10 mIU/mL). The anti-HBs titer levels decreased with
age (GMTs were 124.1 mIU/mL in children 510 years old and
77.2 mIU/mL in the older group; P p .0191), whereas the percentage of antibody-negative children increased from 17.4% to
25.1% (P p .0336). All 14 children who seroconverted to antiHBc without becoming HBsAg carriers had somewhat higher
anti-HBs titers but were not statistically different from uninfected children (GMTs, 200.1 vs. 100.6 mIU/mL; P p .325).

Discussion
In this study, 97% of children born to HBsAg carrier mothers who were immunized with HBIG plus vaccine at birth
avoided HBV infection. Fourteen children (2.7%) seroconverted
to anti-HBc without becoming carriers or developing disease,
and none had detectable levels of HBV DNA. This finding
concurs with other reports of occasional breakthrough infections characterized by transient elevations of ALT and/or appearance of antiHBc antibodies in vaccinees exposed to HBV
[3]. Notwithstanding the combination of passive and active immunization, 3 children (0.6%) became HBsAg carriers. The
emergence of point mutations within the common a determinant of the S gene has been reported worldwide as a possible
cause of vaccine failure. This genomic region is considered to

Downloaded from http://jid.oxfordjournals.org/ by guest on December 26, 2015

Dohme) until 1987 and with recombinant DNA vaccines (ENGERIX B, Smith Kline Biologicals; Recombivax-HB, Pasteur
Merieux; MSD, Merck Sharp & Dohme) thereafter.
Study population. In 19981999, we traced 522 children born
to HBsAg-positive mothers who subsequently were immunized
against hepatitis B in 3 public hospitals in the Campania Region
(southern Italy) during 19851994. For the present study, we obtained a blood sample from each child and mother. Information
regarding the immunization schedule, type of vaccine administered
(plasma derived or DNA recombinant), and hepatitis e antigen
(HBeAg)/antibody (anti-HBe) status of the HBsAg carrier mother
at delivery was collected by use of a precoded questionnaire for
each mother-child pair in the study.
Laboratory tests. All mother-child pairs were tested at enrollment for the presence of HBsAg, anti-HBs, antiHB core antigen
(anti-HBc), HBeAg, and anti-HBe (MEIA or IMx; Abbott Laboratories). In addition, anti-HBs titers were measured in the children and were expressed in milliinternational units per milliliter
(mIU/mL) by comparison with a curve of calibrators standardized
against the World Health Organization reference standard.
HBV DNA was detected in serum samples from children who
seroconverted to anti-HBc by nested polymerase chain reaction
(PCR), using primers derived from both the S and C regions of
the HBV genome [10]. To assess whether children who became
HBsAg carriers were infected with wt or with surface gene mutants
of HBV, we extracted HBV DNA from 200-mL serum samples and
amplified the entire surface antigen gene by PCR [10]. The PCR
products were sequenced by the Sanger dideoxy-chain termination
method. The nucleotide sequences and their deduced amino acid
derived from infected children were compared with those of their
carrier mothers and with published consensus sequences.
Statistical analysis. Differences in frequency were tested by x2
test. Anti-HBs geometric mean titers (GMTs) in children were compared by using nonparametric Mann-Whitney U tests. In the case
of undetectable antibody levels, we assigned an arbitrary value of
5 mIU/mL to allow for calculation.

JID 2001;184 (1 October)

Efficacy of Hepatitis B Vaccination in Newborns

Table 2. Percentage of antibody to hepatitis B surface antigen (antiHBs) positivity and antibody geometric mean titer (GMT) in 505 uninfected children subdivided by age.
a

Anti-HBs

Age, mean
years (range)

Total no.
of children

!10 mIU/mL

8.5 (510)
12.7 (1114)
Total

282
223
505

49 (17.4)
b
56 (25.1)
105 (20.8)

a
b
c

10 mIU/mL

GMT,
mIU/mL

233 (82.6)
167 (74.9)
400 (79.2)

124.1
c
77.2
100.6

Data are no. (%) of children.

P p .0336, x2 test.
P p .0191, Mann-Whitney U test.

of children treated at birth with HBIG and vaccine still have

anti-HBs antibodies at levels considered to be protective (10
mIU/mL). It is reasonable to speculate that the majority of
children with undetectable antibodies may be protected against
HBV, since the immunologic memory for HBsAg is thought to
outlast the presence of circulating antibodies [14, 15]. Thus,
routine administration of booster doses of vaccine to children
may not be necessary, but additional information is needed to
assess whether the immunologic memory in children vaccinated
as infants persists into adolescence and adulthood, when the
risk of infection, either by lifestyle or HBV professional exposure, becomes higher.
In conclusion, our findings indicate that a hepatitis B vaccine
administered at birth in association with HBIG provides immediate and long-term protection against HBV in children born
to HBsAg carrier mothers. The emergence of HBV escape mutants does not raise concern about the efficacy of universal
vaccination programs in Italy, where the G145R prototype of
such variants was first identified 110 years ago [5, 6].
Acknowledgments
Antonella Marzolini handled data, and Maria DOnofrio and Nunzia Coppola collected clinical data.

References
1. Zanetti AR, Tanzi E, Romano` L, Grappasonni I. Vaccination against hepatitis B: the Italian strategy. Vaccine 1993; 11:5214.
2. Whittle HC, Maine N, Pilkington J, et al. Long-term efficacy of continuing
hepatitis B vaccination in infancy in two Gambian villages. Lancet 1995;
345:108992.
3. Hadler SC, Francis DP, Maynard JE, et al. Long-term immunogenicity and
efficacy of hepatitis B vaccine in homosexual men. N Engl J Med 1986;
315:20914.
4. Zuckerman AJ. Effect of hepatitis B virus mutants on efficacy of vaccination.
Lancet 2000; 355:13824.
5. Zanetti AR, Tanzi E, Manzillo G, et al. Hepatitis B variant in Europe. Lancet
1988; 2:11323.
6. Carman WF, Zanetti AR, Karayiannis P, et al. Vaccine-induced escape mutant of hepatitis B virus. Lancet 1990; 336:3259.
7. Ogata N, Cote PJ, Zanetti AR, et al. Licensed recombinant hepatitis B vaccines protect chimpanzees against infection with the prototype surface
gene mutant of hepatitis B virus. Hepatology 1999; 30:77986.
8. Hsu HY, Chang MH, Liaw SH, Ni YH, Chen HL. Changes of hepatitis B
surface antigen variants in carrier children before and after universal vaccination in Taiwan. Hepatology 1999; 30:13127.
9. Purcell RH. Hepatitis B virus mutants and efficacy of vaccination. Lancet
2000; 356:769.
10. Coleman PF, Chen JYC, Mushahwar IK. Immunoassay detection of hepatitis
B surface antigen mutants. J Med Virol 1999; 59:1924.
11. Carman WF, Van Deursen FJ, Mimms LT, et al. The prevalence of surface
antigen variants of hepatitis B virus in Papua New Guinea, South Africa,
and Sardinia. Hepatology 1997; 26:165866.
12. Wallace LA, Echevarria JE, Echevarria JM, Carman WF. Molecular characterization of envelope antigenic variants of hepatitis B virus from Spain.
J Infect Dis 1994; 170:13003.
13. Norder H, Hammas B, Lofdhal S, Courouce AM, Magnius LO. Comparison

Downloaded from http://jid.oxfordjournals.org/ by guest on December 26, 2015

be the major target for the neutralizing antibodies produced

during natural infection or after vaccination [4, 6, 11]. The most
prevalent of such vaccine-escape mutants, which involves a glycinerarginine substitution at aa position 145 (G145R) of the
second loop of the a determinant, was first identified in Naples
in a child born to a carrier mother. The child developed HBV
chronic infection after receiving postexposure prophylaxis with
HBIG and hepatitis B vaccine [6].
This G145R mutation and other less prevalent mutations of
HBsAg a determinant epitopes that allow for infection in successfully vaccinated persons have been found in many countries. Recently, Hsu et al. [8] reported that the universal vaccination against hepatitis B in Taiwan (which began in 1984)
has promoted an accumulation of HBsAg a determinant mutants and suggested the need for including both wt and mutant
S proteins in future vaccines.
In our study, 2 carrier children had wt HBsAg with subtype
ayw, similar to their mothers. The third child had a double
mutation with an amino acid substitution of prolinerserine at
positions 120 (P120S) and 127 (P127S), respectively, which may
explain why this child became an HBsAg carrier, despite having
anti-HBs levels considered to be protective (610 mIU/mL).
Monoclonal mapping studies have shown that loss of proline
at aa 120 affects the antibody binding pattern through alteration of the 3-dimensional epitope presented on the second loop
(aa 139147) of the a determinant, whereas residue 127 is essential for variation of viral subtype specificity, and mutation
at this site may lower the reactivity to the anti-HBs induced
by wt HBsAg [12, 13]. These mutations were not detected in
the mothers by direct sequencing, suggesting either that they
emerged de novo or that they were present originally as a minor
quasi species in maternal serum and that after transmission to
the newborn, they were selected under the immune pressure
generated by the postexposure prophylaxis with HBIG and
hepatitis B vaccine. Whether this child and the 2 chronically
infected children who carried wt HBsAg similar to that of their
mothers were infected in utero or postnatally is hard to establish. No information was available regarding the HBeAg/antiHBe status of the mothers during pregnancy and delivery, and
the laboratory follow-up of these children is incomplete.
Our data indicate that 514 years after immunization, 79.2%

907

908

Mele et al.

of the amino acid sequences of nine different serotypes of hepatitis B

surface antigen and genomic classification of the corresponding hepatitis
B virus strain. J Gen Virol 1992; 73:12018.
14. West DJ, Calandra GB. Vaccine induced immunologic memory for hepatitis B

JID 2001;184 (1 October)

surface antigen: implication for policy on booster vaccination. Vaccine

1996;14:101927.
15. European Consensus Group on Hepatitis B Immunity. Are booster immunisations needed for lifelong hepatitis B immunity? Lancet 2000;355:5615.

Downloaded from http://jid.oxfordjournals.org/ by guest on December 26, 2015