Objective. To determine the immunogenicity of hepatitis B vaccine in preterm infants when the first dose of vaccine is delayed until hospital discharge.
Methods. One hundred two preterm infants (23 to 36 weeks' gestational age) born to hepatitis B surface antigen-negative mothers were enrolled. Immunization was initiated just before hospital discharge with subsequent doses 1 and 6 months later. Serum specimens were obtained before the administration of each vaccine dose and 3 months after the last dose and were tested for antibody to hepatitis B surface antigen (antiHBs).
Results. Eighty-seven infants (85%) completed the study. Ninety percent (n = 78) of infants who completed the study seroconverted (antiHBs ≥10 mIU/mL); 10% (n = 9) remained seronegative at study completion. The geometric mean antibody titer to hepatitis B surface antigen for infants who seroconverted was 200 mIU/mL. Nonresponders (NR) differed from responders (R) in birth weight (NR = 2090 g, R = 1560 g) gestational age (NR = 33 weeks, R = 31 weeks), and weight gain before vaccine initiation (NR = 244 g, R = 633 g). There were no differences in weight or age at vaccine initiation, Apgar scores, interval between vaccine doses, or bacterial infections, steroid use, or transfusions before vaccine initiation.
Conclusions. Ninety percent of preterm infants responded to hepatitis B vaccine when the first dose of vaccine was delayed until hospital discharge. Nonresponders were more likely to be preterm infants of higher birth weight and higher gestational age, and to have gained less weight before vaccine initiation.
The first hepatitis B (HB) vaccine in the United States was licensed in 1981. Initial efforts to vaccinate individuals at high risk of exposure to hepatitis B virus had little impact on incidence of disease.1 In 1991, the Centers for Disease Control announced its recommendation for universal hepatitis B immunization of infants.2 In 1992, the American Academy of Pediatrics (AAP) issued its recommendation for universal immunization, including the initiation of vaccination before nursery discharge.3Because previous authors have demonstrated suboptimal seroconversion in preterm infants vaccinated with HB vaccine shortly after birth,4 the AAP modified its recommendations in 1994, cautioning that immunization of preterm infants might be more effective if initiation of immunization were delayed until hospital discharge or 2 months of age.5
The purpose of this study was to determine the immune response to HB vaccine in preterm infants when vaccination is initiated just before hospital discharge.
One hundred two preterm infants (23 to 36 weeks' gestational age) born to hepatitis B surface antigen- (HBsAg) negative mothers were enrolled in the study between November 1992 and April 1994 with parental consent. Subjects were enrolled at the Children's Hospital of Philadelphia, the Hospital of the University of Pennsylvania, and Lankenau Hospital. The study was approved by the institutional review boards of all three hospitals. Exclusion criteria included known maternal human immunodeficiency virus infection and treatment of the infant with immunoglobulin before vaccine initiation. Infants were vaccinated with a series of three 2.5-μg doses of the recombinant HB vaccine, Recombivax HB (Merck & Co, Inc, Rahway, NJ). The first dose of vaccine was given just before hospital discharge. The mean (± SD) interval between the first and second doses of the vaccine was 41 ± 12 days, and the mean (± SD) interval between the second and third vaccine doses was 167 ± 36 days.
The subjects' parents were notified when immunization and phlebotomy were due. The subjects' primary care physicians were notified of participation in the study and were contacted after each vaccine dose. The primary care physicians' immunization records were examined at study completion to ensure that no additional vaccine doses were given.
Phlebotomy was performed before each vaccine dose (time 0, time 1 month, and time 6 months) and at a mean (± SD) of 115 ± 34 days after the last dose of vaccine (time 9 months). The blood was allowed to clot, and the serum was separated and stored at −20°C until assayed. Time 0- and 9-month samples were assayed for HBsAg, antibody to hepatitis B core antigen (antiHBc), and antibody to hepatitis B surface antigen (antiHBs) by enzyme immunoassay (Abbott Laboratories, Abbott Park, IL). Time 9-month specimens that were positive for antiHBs were quantified using an Abbott quantitation panel (AUSAB Quantitation Panel). Time 1-month and 6-month samples were assayed for antiHBs only. All assays were performed according to the manufacturers' written instructions. Individuals who remained seronegative for antiHBs (<10 mIU/mL) at study completion were asked to return for a fourth dose of vaccine and follow-up serology to ensure seroconversion.
Dichotomous variables were analyzed using the χ2 test and continuous variables were analyzed using the Student's ttest. The Mann-Whitney U test was used for continuous skewed variables. AntiHBs titers were converted to log scale for analysis.
Of the 102 preterm infants enrolled, 87 (85%) completed the study, 5 were excluded from the study, and 10 were lost to follow-up. One of the infants excluded received immunoglobulin before initial immunization, two subjects inadvertently received four doses of vaccine, and two subjects received their first dose of vaccine significantly before hospital discharge. Data were analyzed only for those infants who completed the study. Infants who completed the study ranged in gestational age from 23 to 36 weeks (mean, 31 weeks). Birth weights ranged from 645 g to 2910 g (mean, 1615 g). Thirty-seven of the infants were female (43%) and 50 were male (57%). Forty-two of the infants were African-American (48%) and 41 were white (47%). The first dose of vaccine was given at a mean chronologic age of 5 weeks and a mean weight of 2203 g. Infants who dropped out of the study did not differ significantly from those who completed the study in gestational age, birth weight, Apgar scores, sex, race, antecedent steroid use, number of transfusions, or bacterial infections before vaccine initiation.
Seventeen infants had positive antiHBs and/or antiHBc at time 0. The most likely explanation for positive time 0 antiHBc is prior maternal infection. Possible explanations for positive time 0 antiHBs include prior maternal infection, prior maternal immunization, or blood transfusions before vaccine initiation. All infants with initially positive antiHBc and/or antiHBs responded to the vaccine, and none developed serologic evidence of hepatitis B infection.
Ninety percent (n = 78) of the infants who completed the study responded to the HB vaccine (antiHBs ≥10 mIU/mL). Forty-nine (66%) of 74 infants tested had already seroconverted by 6 months after only two doses (Fig 1). AntiHBs titers ranged from 18 mIU/mL to 10876 mIU/mL. Ten percent (n = 9) of the subjects were seronegative at study completion (nonresponders [NR]). Two of these infants were twins. Nonresponders (NR) differed from responders (R) in birth weight, gestational age, and weight gain before vaccine initiation (Table). There was no significant difference in the corrected age or weight at vaccine initiation, chronological age at vaccine initiation, race, sex, Apgar scores, interval between vaccine doses, or in the percentage of infants who had bacterial infections, received steroids, or received transfusions before vaccine initiation.
The geometric mean antibody titer to hepatitis B surface antigen (GMT) at time 9 months for R was 200 mIU/mL (95% confidence interval (CI) = 144 mIU/mL, 276 mIU/mL). Sixty-five percent (n = 49) of the R had antiHBs titers ≥100 mIU/mL (Fig 2). GMTs were significantly higher in those infants who seroconverted after two doses than in those who remained seronegative until their third vaccine dose (290 mIU/mL vs 68 mIU/mL, P < .01). AntiHBs titers did not differ with respect to gestational age, birth weight, steroids, bacterial infections, or transfusions before vaccine initiation.
Three of the nine NR have returned for a fourth vaccine dose. All subsequently seroconverted with a GMT of 417 mIU/mL (95% CI = 50 mIU/mL, 3461 mIU/mL).
The most recent recommendations from the AAP Committee on Infectious Diseases caution that the optimal time to begin hepatitis B immunization in preterm infants is unknown.6 Considerable data exist concerning immunogenicity of HB vaccine in term infants. In a previous study involving full-term infants, a course of three 2.5-μg doses of Recombivax HB initiated within the first 2 months of life seroconverted 98% of the vaccinees with a GMT of 647 mIU/mL.7 Although two studies of HB vaccine immunogenicity in preterm infants have shown seroconversion rates similar to those of term infants, one of these studies excluded infants with infection or severe illness,8 and the subjects in the other study had a median gestational age of 36 weeks.9
Two other studies have suggested suboptimal seroconversion in preterm infants. In 1991, Chawareewong et al10 reported a seroconversion rate of 88% in preterm infants who received the first dose of HB vaccine within 10 days of birth. However, only 67% of infants who weighed ≤1500 g at birth seroconverted, compared with 95% of infants born at >1500 g (P = NS).10 In 1992, Lau et al4 demonstrated differences in immunogenicity based on weight at vaccine initiation. Preterm infants who received their first dose of vaccine when they reached 1000 g had a seroconversion rate of 78.9%, compared to a 90.5% seroconversion rate in preterm infants who received their first dose of vaccine when they reached 2000 g (P= NS).4
After consideration of the above studies, the AAP Committee on Infectious Diseases advised that HB vaccine in preterm infants might be more immunogenic if initiation of immunization were delayed until hospital discharge or 2 months of age.5 Our study was designed to evaluate immune response in preterm infants when the first dose of HB vaccine is delayed until hospital discharge, with subsequent doses 1 and 6 months later.
Ninety percent of our subjects seroconverted with a GMT of 200 mIU/mL, much less than that seen in term infants. NR were of higher gestational age and birth weight than R. Although this seems incongruous, studies in term infants have demonstrated that very young infants (up to 2 months) achieve lower GMTs after three doses of HB vaccine compared to infants 3 months or older at vaccine initiation.11 This suggests that immune response may be associated with postnatal development. In our study, R gained more weight before their first dose of vaccine. In addition, R were 5 weeks of age at vaccine initiation compared to 3 weeks for NR (P = NS). Perhaps a shortened postnatal period before vaccine initiation accounted for the lack of immune response in the NR.
One limitation of our study is the lack of a control group of term infants. Due to the abundance of literature describing adequate immune response in term infants, we believed it would be difficult to recruit a control group for immunization and phlebotomy. In addition, due to the short follow-up time, we have no information on the duration of seropositivity in R. Another limitation is the small number of NR in this study. There may have been differences in the nonresponder and responder groups that did not achieve statistical significance due to the small numbers (type II error). In addition, because a large number of variables were statistically evaluated, this study has an increased risk of finding statistical significance by chance alone (type I error).
The recommendations for universal HB immunization have been altered since their introduction. Current recommendations advise delaying vaccine initiation in preterm infants until hospital discharge or 2 months of age. Although this study shows a higher seroconversion rate when vaccine initiation is delayed until hospital discharge, 10% of our subjects were lost to follow-up and 10% were NR. Because we are unable to predetermine which individuals will be NR, options to reduce the nonresponse rate include delaying vaccine initiation beyond hospital discharge, providing all preterm infants with four doses of vaccine, or performing post-vaccine antiHBs serologies on all preterm infants and providing only NR with a fourth dose of vaccine. Further investigation is necessary to determine the most cost-effective option.
In addition, the GMTs achieved by our R were much lower than that seen in term infants. One might question whether the lower GMT achieved in preterm infants will translate into a shorter duration of protection. It has been shown that higher initial antiHBs titers are associated with longer duration of seropositivity. It is not known, however, whether this translates into a longer duration of protection from hepatitis B infection. It is known that vaccine-induced immunologic memory for HBsAg does persist after the level of antiHBs falls below 10 mIU/mL, so that protection may endure even when antibody is no longer present.11 In conclusion, 90% of preterm infants seroconverted when HB vaccine initiation was delayed until hospital discharge. Additional studies are needed to determine if further delay in vaccine initiation in this population would allow seroconversion rates to approach those of term infants. Until further information is available, it may be advisable to determine antiHBs serologies in all preterm infants immunized with HB vaccine.
This study was supported by Merck Research Laboratories and by US Public Health Service Research grants MO1-RR00040 and RR00240 from the National Institutes of Health.
- Received January 2, 1996.
- Accepted May 28, 1996.
Reprint requests to (S.C.K.) 3470 Plantation River Drive, Boise, ID 83703.
- HB =
- hepatitis B •
- AAP =
- American Academy of Pediatrics •
- HBsAg =
- hepatitis B surface antigen •
- antiHBc =
- antibody to hepatitis B core antigen •
- antiHBs =
- antibody to hepatitis B surface antigen •
- NR =
- nonresponders •
- R =
- responders •
- GMT =
- geometric mean antibody titer (to hepatitis B surface antigen) •
- CI =
- confidence interval
- Committee on Infectious Diseases, American Academy of Pediatrics
- Committee on Infectious Diseases, American Academy of Pediatrics
- Committee on Infectious Diseases, American Academy of Pediatrics
- ↵American Academy of Pediatrics, Committee on Infectious Diseases. In: Peter G, ed. 1994 Red Book. Report of the Committee on Infectious Diseases. 23rd ed. Elk Grove Village, IL: American Academy of Pediatrics; 1994:231–232
- Chirico G,
- Belloni C,
- Gasparoni A,
- et al.
- Copyright © 1997 American Academy of Pediatrics