ARTICLE |
a Departments of Pediatrics
b Medicine
d Microbiology and Immunology, University of Rochester, Rochester, New York
c Center for Biologics Evaluation and Research, US Food and Drug Administration, Bethesda, Maryland
e National VZV Laboratory, Centers for Disease Control and Prevention, Atlanta, Georgia
| ABSTRACT |
|---|
|
|
|---|
METHODS. Immune status before immunization and immune response after immunization with measles-mumps-rubella and varicella vaccines at 15 months of age were compared in 32 infants, 16 of whom were preterm (<29 weeks' gestation) and 16 of whom were term (
37 weeks' gestation) at birth. Blood was drawn before vaccination and 3 to 6 weeks thereafter. Measles antibody was measured by plaque reduction neutralization assay. Mumps and rubella immunoglobulin G were measured in available sera by enzyme-linked fluorescent immunoassay. Varicella immunoglobulin G was measured in available sera by glycoprotein enzyme-linked immunosorbent assay. Values that were above or below the assay limits were assigned values double or half those limits, respectively. The primary outcome was the geometric mean antibody titer.
RESULTS. Preterm children had lower mumps and rubella geometric mean titers than did term children before vaccine, and nearly all children were seronegative for each of the 4 vaccine antigens before immunization. Measles, mumps, rubella, and varicella geometric mean titers were similar between groups after vaccine. All children were seropositive for measles after vaccine, whereas 13 of 14 preterm and 11 of 13 term children were seropositive for mumps, 13 of 14 preterm and 13 of 13 term children were seropositive for rubella, and 11 of 16 preterm and 9 of 15 term children were seropositive for varicella.
CONCLUSIONS. Preterm children mounted antibody responses that were similar to those of term children after measles-mumps-rubella and varicella vaccines at 15 months of age.
Key Words: vaccines premature infant measles-mumps-rubella vaccine varicella vaccine very low birth weight infant immunology humoral immunity immunization
Abbreviations: MMR—measles-mumps-rubella GMT—geometric mean titer IgG—immunoglobulin G ELISA—enzyme-linked immunosorbent assay RFV—relative fluorescence value OD—optical density
The American Academy of Pediatrics recommends that preterm infants receive most vaccines, including the measles-mumps-rubella (MMR) and varicella vaccines, at the same chronologic age as term infants.1 However, extremely preterm infants (born at <28–30 weeks' gestation) have lower antibody responses than term infants to several vaccines, including Haemophilus influenzae type b,2–6 hepatitis B,7–11 and polio,12,13 when these vaccines are given at the postnatal ages that are recommended for term infants. The immunogenicity of vaccine series that are begun during the first 6 months in preterm infants may remain diminished after the administration of toddler14,15 and school-age15,16 boosters. Inactivated influenza vaccine, which is administered at or after 6 months of age, also seems to be less immunogenic in preterm than term infants.17 These data raise the question of whether a relative deficit in humoral immunity persists beyond 6 months' postnatal age in preterm infants.
We evaluated the serologic response of extremely preterm infants to MMR and varicella vaccines that were given at 15 months' chronologic age. We hypothesized that these children would have lower geometric mean antibody titers (GMT) to varicella, mumps, measles, and rubella after vaccination than infants who were born at term but that a similar number of preterm as term infants would become seropositive.
| METHODS |
|---|
|
|
|---|
37 weeks' gestation, they were <16 months of age, they had not yet received MMR or varicella vaccines, and the parents had given permission for and the primary pediatrician had agreed to the study. Infants were excluded when they had known immunodeficiency, had contraindications to vaccination, or required systemic corticosteroids or oxygen therapy at the time of vaccination. Concurrent administration of other vaccines was permitted.
Preterm infants were recruited at the time of a scheduled visit to the regional perinatal center's neurodevelopmental follow-up clinic (at which all infants who were born at <30 weeks' gestation were evaluated) at
9 to 12 months of age. Term infants were recruited from either a hospital-based pediatric clinic or a private pediatric practice at the time of the 12-month visit. Term infants were matched to preterm infants on race and, whenever possible, on gender.
Study Visits
Written parental permission, demographic and historical information, and a blood specimen were obtained before immunization. MMR II (Merck Vaccines, Whitehouse Station, NJ) and varicella (Varivax; Merck) vaccines were administered by the primary pediatrician from his or her routine office stock according to the manufacturers' instructions and routine medical care practices. At a visit 3 to 6 weeks after immunization, interim medical history and a blood specimen were obtained.
Serology
Serum was separated and frozen, in 3 aliquots, at –80°C until the time of analysis. Measles antibody was measured by plaque reduction neutralization assay according to established protocols.18,19 Briefly, measles virus (25–30 plaques per well) and serial dilutions of heat-inactivated subject serum were incubated at 36°C, added to Vero cell monolayers in 24-well culture plates in duplicate, and covered with carboxymethylcellulose overlay media. The trays then were incubated for 5 days at 36°C, at which time the monolayers were stained and fixed and the plaques were counted. The end point for the test was calculated using the Kärber formula and reflected the dilution of serum that reduced the number of plaques by 50%. Dilution values were converted to mIU by comparison with results that were obtained using second International Standard (66/202) tested in parallel; in this assay, a titer of 1:8 was equivalent to 8 mIU/mL. A neutralizing antibody titer
120 mIU/mL was considered evidence of seroresponse.20
Mumps and rubella immunoglobulin G (IgG) titers were measured using a commercially available, automated, enzyme-linked fluorescent immunoassay (VIDAS Vitek ImmunoDiagnostic Assay System; bioMérieux, Inc, Hazelwood, MO). Titers were reported as relative fluorescence values (RFV). Mumps values <0.35 were considered negative, values 0.35 to 0.49 were considered equivocal (and treated as negative during analysis), and values
0.50 were considered positive. Rubella values <0.40 were considered negative, values 0.40 to 0.49 were considered equivocal (and treated as negative during analysis), and values
0.50 were considered positive.
Varicella IgG was measured using a glycoprotein enzyme-linked immunosorbent assay (ELISA) protocol that essentially was the same as that reported by Wasmuth and Miller.21 A twofold end point dilution series that began with a dilution of 1:20 was performed. Results for each dilution were scored as positive, negative, or equivocal on the basis of empirically determined cutoffs that were based on calculations of cumulative variance of test data from repeated testing of known varicella zoster virus–positive and –negative sera by several operators. Titration end points were defined as the highest dilution to produce a test result in the equivocal range. All tests were controlled internally using defined positive and negative sera. The reportable range of results for the glycoprotein ELISA was <0.100 optical density (OD) units = negative, 0.100 to 0.249 OD units = equivocal (and treated as negative during analysis), and
0.250 OD units = positive. In the event that the volume of serum was insufficient to complete all testing, assays were performed in the following priority order: (1) measles, (2) varicella, and (3) mumps and rubella.
Analysis
Antibody titers were normalized by logarithmic transformation and summarized as GMT. For the purposes of the logarithmic transformation, values above or below the limits for each assay were assigned values double or half those limits, respectively. Vaccine titers were compared using the Mann-Whitney U test. Birth weights and ages were summarized as medians. Categorical variables were compared using the
2 or Fisher's exact tests, as appropriate. The primary outcome was antibody GMT, and the secondary outcome was the proportion of children who were seropositive for each antibody.
A patient population of 16 patients per group was anticipated to be sufficient to measure a 1.5-fold elevation in term infants' GMT of antibody over preterm infants' GMT, assuming an SD of 0.5 times the preterm infants' GMT,
= .05, and a power of 0.80, or a twofold difference assuming an SD the same size as the preterm infants' GMT with the same
and power.
| RESULTS |
|---|
|
|
|---|
|
All patients in both groups were seronegative for measles-neutralizing antibody before vaccine, and all had measles titers >120 mIU/mL after vaccine (Table 2). Measles titers were similar between groups both before and after vaccine (Fig 1). Preterm infants had lower mumps and rubella titers than did term infants before vaccine; however, the groups had similar titers after vaccine (Fig 1). Similar proportions of infants had positive mumps and rubella antibody seroresponses after vaccine (Table 2). Varicella titers were similar between preterm and term infants both before and after vaccine (Fig 1). All prevaccine varicella titers in both groups fell below the limit of detection. Postvaccine varicella titers were lower than might have been expected, and several infants in both groups had titers below the defined seroresponse level (Table 2).
|
|
| DISCUSSION |
|---|
|
|
|---|
Identification of high-risk groups for poor MMR or varicella vaccine immunogenicity potentially is clinically important. Measles, mumps, and rubella still circulate to a limited degree in the US population, with continued importation from international sources.22–25 Measles and mumps cause occasional outbreaks,22–24,26 and these epidemics tend to cluster in populations with inadequate immunity.22,23 Measles remains a major killer of children worldwide.27 Varicella circulates freely in the US population and continues to cause an attenuated chickenpox illness in a significant proportion of vaccine recipients who are exposed to the virus.28–31 Individuals with lower levels of varicella immunity are more likely to experience a more severe form of varicella disease than those with higher levels of varicella immunity.28,30,32
Few data existed previously for measles, mumps, rubella, or varicella vaccines that were given to infants who were born preterm. Shortly after the introduction of rubella vaccine, a study of 7 term and 5 preterm (<2500 g birth weight) 12- to 25-month-old infants showed delayed cellular and humoral rubella vaccine responses in the preterm infants that became equivalent to those in term infants by 42 days after immunization.33
Many vaccines that first are administered under 6 months of age now have been studied fairly well in extremely preterm infants.2–13,34 Although antibody levels after vaccination in extremely preterm infants tend to be somewhat lower than those that are achieved by term infants, similar proportions of preterm and term infants mount antibody responses above the minimum levels that are considered to be protective against disease.
In contrast to our findings with MMR and varicella vaccines, limited experiences with other vaccines have suggested continued deficits beyond 6 months of age in preterm infants' vaccine responses. Inactivated influenza vaccine, given at 6 to 18 months of age, elicits lower antibody titers and diminished T-cell proliferative responses in preterm than in term children.17 We and others have reported diminished antibody responses after tetanus, diphtheria, polio, and Haemophilus influenzae type b booster vaccines in 4- to 7-year-old children who were born extremely preterm.14–16 The immunologic mechanisms of the decreased vaccine responses are unknown but seem to affect both the humoral14–17 and the cellular17 arms of the immune system.
The relatively robust vaccine responses in the preterm infants in this study may be explained in part by the specific vaccine antigens studied. Live viral vaccines, such as MMR and varicella, often are highly immunogenic. Early studies with the varicella vaccine suggested 94% to 100% seroconversion, 93% to 100% cellular immune response, and 90% to 96% efficacy, with immunity persisting unabated over 7 to 10 years.35–38 Comparisons between oral, live, attenuated polio vaccine and injectable, enhanced-potency, inactivated polio vaccine in preterm infants also have suggested that the live vaccine may be more immunogenic.12,13,15,39
This study has inherent limitations. Relatively few infants were studied. Although the numbers were calculated to be sufficient to test the hypothesis of difference between the groups, a lack of detected difference does not necessarily suggest equivalence of the 2 study groups. However, the narrow confidence intervals that were observed in this study (Fig 1) suggest that any true difference between groups is likely to be small and clinically insignificant. Vaccines, particularly live vaccines, stimulate both the cellular and the humoral components of the immune system, and cellular responses to the vaccines studied here have been described.17,33,35,37,40 This study measured only antibody titers. Differences in antibody affinity or cellular responses between groups would not be detected. In many cases, however, antibody levels alone provide a reasonable estimate of vaccine immunogenicity.5,41–43 In addition, our results cannot rule out the possibility that some persistent immune system alteration in preterm infants may be overcome by relatively highly immunogenic, live, attenuated vaccines such as MMR and varicella vaccines.
| CONCLUSION |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
We thank John Treanor, MD, Diane O'Brien, RN, and Doreen Francis, RN, for support of this project; Jason Roy, PhD, for statistical guidance; and the University of Rochester Neonatal Continuing Care Clinic, the Golisano Children's Hospital at Strong Pediatric Clinic, and Elmwood Pediatrics for help with recruitment. We also thank the infants and their families for participating.
| FOOTNOTES |
|---|
Address correspondence to Carl T. D'Angio, MD, Box 651, Neonatology, University of Rochester, 601 Elmwood Ave, Rochester, NY 14642. E-mail: carl_dangio{at}urmc.rochester.edu
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the funding agencies.
The authors have indicated they have no financial relationships relevant to this article to disclose.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||