Deciency of the Humoral Immune Response to Measles Vaccine in Infants Immunized at Age 6 Months

Hayley A. Gans, MD; Ann M. Arvin, MD; Jill Galinus; Linda Logan; Ross DeHovitz, MD; Yvonne Maldonado, MD

Context.Measles causes serious morbidity in infants, with the highest risk among those who are 6 to 12 months of age. In the United States, measles vaccine has been given at age 12 to 15 months to minimize interference by passive antibodies and to achieve the high seroprevalence required for herd immunity. Infants of mothers with vaccine-induced immunity may lose passively acquired antibodies before 12 months, leaving them susceptible to measles infection. Objective.To assess the immunogenicity of measles vaccine in infants younger than 12 months. Design.Cohort study conducted before and after measles immunization. Setting.Pediatric clinic in Palo Alto, Calif. Participants.Infants 6 (n = 27), 9 (n = 26), and 12 (n = 34) months of age were enrolled; 72 provided both initial and follow-up samples. Main Outcome Measures.Evaluation of immunogenicity before and 12 weeks after measles vaccination, including measles neutralizing antibody titers, measlesspecic T-cell proliferation, and cytokine proles. Results.Measles neutralizing antibodies were present before vaccination in 52% (12/23), 35% (7/20), and 0% (0/22) of 6-, 9-, and 12-month-old infants, respectively. In the absence of detectable passive antibodies, geometric mean titers after vaccination were signicantly lower in 6-month-old infants compared with 9-month-old infants (27 vs 578, P = .01) and 12-month-old infants (27 vs 972, P = .001). The seroconversion rate, dened as a 4-fold rise in antibody titer, in these 6-month-old infants was only 67%, and only 36% of these infants achieved seroprotective neutralizing antibody titers of 120 or higher after vaccination compared with 100% of 9- and 12-month-old infants lacking detectable passive antibody prior to vaccination. T-cell proliferation and cytokine responses to measles did not differ with age. Conclusions.Humoral immunity was decient in 6-month-old infants given measles vaccine, even in the absence of detectable passively acquired neutralizing antibodies. Comparison of their responses with those of 9- and 12-month-old infants indicates that a developmental maturation of the immune response to measles may occur during the rst year of life, which affects the immunogenicity of measles vaccine.
JAMA. 1998;280:527-532

THE SUSCEPTIBILITY of infants to serious disease caused by viruses is recognized during the immediate neonatal period but it also extends through the first year of life, suggesting that immu-

From the Department of Pediatrics, Stanford University School of Medicine, Stanford, Calif (Drs Gans, Arvin, and Maldonado and Mss Galinus and Logan), and the Department of Pediatrics, Palo Alto Medical Foundation, Palo Alto, Calif (Dr DeHovitz). Reprints: Hayley Gans, MD, Department of Pediatrics, Stanford University School of Medicine, G-312, Stanford, CA 94305-5208 (e-mail: drhayley@leland .stanford.edu).

nocompetence of the host develops gradually over this time interval. In the case of measles, clinical experience demonstrates that a critical maturation of the host response occurs between 6 and 12 months, based on the subsequent decline in measles mortality.1-5 The high rate of infant morbidity and mortality observed in developing countries and during recent outbreaks in the United States has renewed interest in evaluating measles immunization of infants at the youngest possible age.1,6,7 Past studies showed a high failure rate of measles

vaccination in infants younger than 12 months. Poor immunogenicity was associated with the persistence of antibodies acquired transplacentally from mothers whose measles immunity was induced by natural infection.7-13 The recommendation for vaccination at 12 to 15 months of age was made to ensure that all, or almost all, infants had lost passive antibody when immunized and that optimal herd immunity was achieved.14 In contrast with the first 3 decades after measles vaccine was introduced, most infants in the United States are now born to mothers who have vaccine-induced immunity to measles.15-19 These infants can be expected to lose maternal antibodies by 9 to 12 months of age.19,20 Our recent study showed only 29% of 9month-old infants and 5% of 12-monthold infants had persistent passive antibodies.20 As a result, more infants younger than 12 months now lack both passive and active measles immunity, leaving them unprotected and in the highest-risk group for life-threatening complications. Maintaining high seroprevalence rates to achieve herd immunity protects young infants but local epidemics confirm past studies showing that a small population of susceptible individuals can sustain a measles outbreak.11,20 Little is known about the maturation of virus-specific immune responses in healthy infants following infection or immunization.21-24 Newborn infants have deficiencies in primary antigen presentation by dendritic cells, limited T-cell proliferation, impaired B-cell function, and reduced production of cytokines by helper T cells of the type 1 subset (TH1), including interleukin 2 (IL-2) and interferon (IFN- ).24-30 Whether these deficiencies, which could diminish the immunogenicity of measles vaccine, are still present in 6- to 9-month-old infants has not been determined. Since many infants now have early loss of passive antibodies, it is feasible to distinguish the relative impact of deficiencies attributed to the maturation of the immune re527

JAMA, August 12, 1998Vol 280, No. 6

Immune Response to Measles VaccineGans et al

1998 American Medical Association. All rights reserved. Downloaded from jama.ama-assn.org by guest on February 22, 2012

sponse from passive antibody inhibition of measles vaccine immunogenicity. The purpose of our study was to assess whether there are intrinsic immunologic barriers to immunization of infants younger than 12 months with measles vaccine, or whether neutralization of the vaccine virus by passive antibodies constitutes the only significant obstacle to early vaccination. In the past, potential interference by passive antibodies prevented the analysis of developmental changes in the host response to measles.13,31 METHODS Study Population Subjects included healthy infants, without documented intercurrent illnesses, who were 6 (n = 27), 9 (n = 26), or 12 (n = 34) months of age and were seen for their well-child visit at the Palo Alto Medical Foundation, Palo Alto, Calif. A total of 87 infants were enrolled in the study; 2 infants never participated after consent was given because their mothers decided against phlebotomy (6-montholds: n = 1, 12-month-olds: n = 1), 13 infants provided only the initial blood sample (6-month-olds: n = 1, 9-montholds: n = 2, 12-month-olds: n = 10), and 72 infants provided both the initial and follow-up samples (6-month-olds: n = 25, 9-month-olds: n = 24, 12-month-olds: n = 23), but not all T-cell and B-cell assays were performed for each sample. The study was approved by the Stanford University Committee for the Protection of Human Subjects and the institutional review board of the Palo Alto Medical Foundation; written consent was obtained from parents or guardians. Children born before 36 weeks gestation, whose birth weight was less than 2500 g, or who had chronic underlying illnesses were excluded. Mothers were grouped by birth date, as those born before 1957 (n = 17), between 1957 and 1963 (n = 34), after 1963 (n = 13), or unknown (n = 3). Blood samples (2-5 mL) were collected from infants before vaccination and 12 weeks later. Samples were also obtained after the Measles, Mumps, and Rubella Virus Vaccine (M-M-R, Merck & Co Inc, West Point, Pa) (given as routine well-child care at age 12 months) from infants who did not respond to initial measles vaccination. Healthy adults who had received at least 1 measles vaccination were also evaluated. No cases of measles were identified in our geographic area during the study period. Measles Vaccine Six- and 9-month-old infants were immunized with Measles Virus Vaccine Live (Attenuvax; Merck & Co Inc; lot 1002A;
528 JAMA, August 12, 1998Vol 280, No. 6

expiration, August 15, 1996; strength, 1000 median tissue culture infective doses [TCID50] of the US reference measles virus); these infants subsequently received M-M-R at 12 months. Twelvemonth-old infants were immunized with M-M-R Virus Vaccine Live (Merck & Co Inc; lot 1315A; expiration, October 12, 1996; containing Measles Virus Vaccine Live; strength, 1000 TCID50 of the US reference measles virus). Measles Antibody Assays Serum samples were stored at 80C. Paired specimens from before and after vaccination were run in parallel when possible and tested for measles neutralizing antibody using a modified plaque reduction neutralization (PRN) assay.32 This assay was used instead of commercially available enzyme-linked immunosorbent assay (ELISA) kits because of the superiority in detection of low titers of measles antibodies.33 Briefly, serial 4fold dilutions, starting at 1:4, of heatinactivated serum (56C for 30 minutes) were mixed with an equal volume of a low-passage strain of Edmonston measles virus containing 25 to 35 plaqueforming units. Each dilution was incubated in duplicate in plastic 24-well plates with Vero cell monolayers for 1 hour 45 minutes at 36C and 5% carbon dioxide. A reference serum calibrated against the international reference was included in duplicate in each run. The PRN titer was defined as the serum dilution that reduced the number of plaques by 50%. Titers less than 1:4 were considered negative and assigned a value of 1 for statistical analysis. A PRN titer of 1:120 shown previously to neutralize infection34 was used as the definition for seroprotection. Seroconversion was defined as a 4-fold rise in antibody titer after levels prior to vaccination were corrected for decay over 3 half-lives (1 half-life = 1 month); lack of seroconversion was considered primary vaccine failure. T-cell Proliferation Assay Fresh peripheral blood mononuclear cells (PBMCs) were separated from whole blood by Ficoll-Hypaque gradient and added to 96-well microtiter plates at concentrations of 3.0 105 per well in RPMI 1640 (Gibco, Gaithersburg, Md), 10% normal human sera (Sigma, St Louis, Mo). Measles antigen, prepared from infected Vero cell lysates, or an uninfected cell control was added at dilutions of 1:16 and 1:32 to triplicate wells. T-cell proliferation was measured by adding tritiated thymidine (2.5 Ci per well) after 5 days for 6 to 18 hours. The stimulation index (SI) was calculated as the mean counts per minute (cpm) in

measles antigenstimulated wells divided by the mean cpm in control wells. A positive SI to measles was 3.0 or greater, based on the mean and SD of responses in infants before vaccination. Phytohemagglutinin (Difco, Detroit, Mich) was used as a positive control. Assays for Cytokine Production Supernatants from PBMCs stimulated with measles antigen prepared from infected Vero cell lysates or uninfected cell controls were collected from duplicate wells for 8 consecutive days, stored at 70C, and tested using ELISA assays, with sensitivities of detection defined by reference standards in each assay. The supernatants collected on days 1 through 8 after initial incubation of T cells with measles antigen were tested in parallel to determine the peak cytokine response; this peak concentration was used as the value for statistical analysis. Interleukin 4 and IL-10 production were measured using assays from Genzyme Inc (Cambridge, Mass). Specimens were also tested for IL-4 using Cytoscreen ultrasensitive assay (Biosource Inc, Carmillo, Calif). The ELISA method from Endogen Inc (Cambridge, Mass) was used to measure IL-2 and IFN- . Statistical Analysis The reciprocals of the measles PRN titers were transformed, and geometric mean titers (GMTs) were calculated. Differences in antibody titers among groups were evaluated by the Mann-Whitney U test. Stimulation indexes and cytokine responses in individual patients before and after vaccination were compared using the paired Student t test; the unpaired t test was used to compare study populations, but only on paired data points. Analysis of variance was performed to evaluate differences among the means of all 3 groups. The 2 and Fisher exact tests were used to compare the number of vaccinees in each cohort who had antibody or proliferation responses. The Spearman rank coefficient was used to evaluate correlations between SI and GMTs or cytokine responses. Statistical significance was defined at P .05 for all analyses performed. RESULTS Humoral Immune Responses Measles neutralizing antibody titers were determined in 65 infants before vaccination. Twelve (52%) of 23 6month-old infants had detectable passive antibodies compared with 7 (35%) of 20 9-month-old infants and 0 (0%) of 22 12-month-old infants (6 months vs 12 months, P = .003; 9 months vs 12 months, P = .01) (Table). There were no
Immune Response to Measles VaccineGans et al

1998 American Medical Association. All rights reserved. Downloaded from jama.ama-assn.org by guest on February 22, 2012

statistical differences in measles GMTs before vaccination when comparing 6and 9-month-old infants born to the oldest mothers, who were born before 1957, and the youngest mothers, who were born after 1963. Measles-neutralizing antibody titers after vaccination, determined in the same 65 infants and expressed as GMTs, were 35 (95% confidence interval [CI], 13-95), 201 (95% CI, 69-585), and 972 (95% CI, 669-1415) in 6-, 9-, and 12month-old infants, respectively (6 months vs 9 months, P = .003; 6 months vs 12 months, P<.001; 9 months vs 12 months, P = .01) (Figure 1). The seroconversion rate (4-fold rise in antibody titer) in 6-month-old infants was only 65% (15/23) compared with 90% (18/20) of 9-month-old infants and all of the 22 infants who were 12 months old (6 months vs 12 months, P = .01). Six-month-old infants were also less likely than 9- and 12month-old infants to develop seroprotective neutralizing antibody titers of 120 or more. Only 10 (43%) of 23 6-month-old infants had seroprotective titers after vaccination compared with 17 (85%) of 20 9-month-old infants and 21 (95%) of 22 12-month-old infants (6 months vs 9 months, P = .01; 6 months vs 12 months, P = .001). In the presence of passive antibodies, determined by PRN, there was no statistical difference in the seroconversion rates and GMTs among all infants, regardless of age. Sixty-three percent of 6-month-old infants who had passive antibody before vaccination seroconverted, with a postvaccination GMT of only 45. This rate of response was not statistically different from the 40% of 9month-old infants who seroconverted in the presence of passive antibodies and who had a GMT of 85 after vaccination. In contrast, age-related differences were observed among infants who had no detectable passive antibodies by PRN prior to vaccination. Among these infants, the measles GMTs after vaccination were lower in 6-month-old infants compared with 9-month-old infants (27 vs 578; P = .01) and 12-month-old infants (27 vs 972; P = .001) (Figure 2). Analysis of variance showed P = .001. The seroconversion rate in these 6-month-old infants was 67%, and only 36% of these infants reached seroprotective GMTs after vaccination, compared with 100% of the 9and 12-month-old infants lacking detectable passive antibody prior to vaccination. The GMTs, seroconversion, and seroprotective rates of the 9-month-old infants who lacked detectable passive antibodies showed no statistical difference compared with those of 12-month-old infants, all of whom lacked detectable passive antibodies prior to vaccination.
JAMA, August 12, 1998Vol 280, No. 6

Humoral and Cell-Mediated Immune Responses to Measles Vaccine*

Age at Vaccination, mo 6 Neutralizing antibody Passive antibody, No. (%) Seroconversion, No. (%) Seroprotection, No. (%) GMT (95% CI), prevaccine GMT (95% CI), postvaccine T-cell proliferation SI 3, No. (%) Mean (SE) SI prevaccine Mean (SE) SI postvaccine Cytokine response, mean (SE), pg/mL IL-2 prevaccine [n] Peak IL-2 postvaccine Interval to peak IL-2, d IFN- prevaccine [n] IFN- postvaccine Interval to peak IFN- , d IL-10 prevaccine [n] IL-10 postvaccine Interval to peak IL-10, d 12/23 (52) 15/23 (65) 10/23 (43) 10 (4-26) 35 (13-95) 15/24 (63) 2.1 (0.38) 3.8 (0.52) 46.2 (18.00) [17] 148.6 (25.77) 3.1 41.4 (10.93) [16] 194.4 (75.85) 4.9 81.9 (15.54) [18] 73.9 (26.44) 6.9 9 7/20 (35) 18/20 (90) 17/20 (85) 3 (2-6) 201 (69-585) 15/22 (68) 1.3 (0.27) 7.5 (2.80) 65.3 (16.59) [12] 161.1 (52.4) 3.6 49.8 (19.69) [14] 366 (125.37) 4.8 84.46 (17.87) [11] 77.36 (25.8) 6.4 12 0/22 (0) 22/22 (100) 21/22 (95) 1 (1-1) 972 (669-1415) 12/21 (57) 1.4 (0.22) 5 (1.10) 32.5 (9.83) [12] 99.8 (24.32) 3.8 42.5 (13.69) [12] 110.4 (27.83) 4.9 59.9 (11.98) [9] 50.4 (10.49) 6.2 Adults NA NA NA NA NA 13/18 (72) NA 8 (1.59) NA 306.3 (60.84) 3.5 NA 300.6 (147.66) 5.4 NA 152.4 (32.03) 6.7

*NA indicates not applicable; GMT, geometric mean titer; SI, stimulation index; IL-2, interleukin 2, IFN- , interferon ; and IL-10, interleukin 10. Seroconversion is dened as a 4-fold increase in GMT after vaccination compared with before vaccination. Seroprotection is dened as a GMT of 120 or more.

Six infants who had primary measles vaccine failure at 6 months (n = 5) or 9 months (n = 1) were evaluated for neutralizing antibodies up to 2 years after revaccination with M-M-R. Their GMTs increased significantly, from 4.7 to 403 (P = .02). No correlations were found between measles GMT and cytokine responses to measles in any of the infant groups. T-cell Proliferation T-cell proliferation to measles antigen was measured in 67 infants and 17 vaccinated adults. Infants from all age groups showed a significant increase in T-cell proliferation to measles antigen after vaccination (6 months, P = .02; 9 months, P = .04; 12 months, P = .006) (Figure 3). A positive response (SI 3) was detected in 15 (63%) of 24 6-month-old infants, 15 (68%) of 22 9-month-old infants, 12 (57%) of 21 12-month-old infants, and 72% of adults (Table). Response rates were not statistically different among the infant groups and adults. The mean (SE) SIs after vaccination of 6-, 9-, and 12-month-old infants were 3.8 (0.52), 7.5 (2.80), and 5.0 (1.10), respectively, which were not statistically significant. These SIs did not differ statistically from vaccinated adults, who had a mean (SE) SI of 8.0 (1.60) (Table). The presence or absence of passive antibody did not influence T-cell proliferation responses to measles antigen in the 6- or 9-month-old infants. Measles SI did not correlate with neutralizing antibody titers in individual patients after vaccination.

Five infants who had serologically defined primary vaccine failure had SIs of 3 or greater after measles immunization. Among infants who had passive antibodies prior to vaccination, 8% (1/12) of 6month-old infants and 14% (1/7) of 9month-old infants had neither humoral nor T-cell proliferation to measles vaccine. All infants who lacked passive antibodies had either humoral or cellmediated immunity or both after immunization. Cytokine Production Interleukin 2 production in response to measles antigen was measured in 41 infants and 8 adults. Before vaccination, the mean (SE) IL-2 concentrations were 46.2 (18.0), 65.3 (16.59), and 32.5 (9.83) pg/mL in 6-, 9-, and 12-month-old infants, respectively (Table). A significant rise in IL-2 production was detected after vaccination of only 6- and 12-month-old infants, to 148.6 (25.77) and 99.8 (24.32) pg/mL, respectively (6 months, P = .003; 12 months, P = .03). No age-related differences were detected among infant groups but the mean IL-2 concentration in vaccinated adults was 306.3 (60.84) pg/ mL, which was higher than levels among all infants (P = .002). The production of IFN- by T cells stimulated with measles antigen was measured in 42 infants and 8 adults. Mean (SE) IFN- concentrations before vaccination were 41.4 (10.93), 49.8 (19.69), and 42.5 (13.69) compared with concentrations of 194 (75.85), 366.0 (125.37), and 110.4 (27.83) pg/mL after
529

Immune Response to Measles VaccineGans et al

1998 American Medical Association. All rights reserved. Downloaded from jama.ama-assn.org by guest on February 22, 2012

1000

Before Vaccination After Vaccination

Passive Antibody 1000 No Passive Antibody 10

Before Vaccination After Vaccination

800

800

600

600

Stimulation Index

GMT

GMT

400

400

200

200

12

12

12

Age at Time of Vaccination, mo

Figure 1.Neutralizing antibody responses of infants before and after measles immunization. Shown are the geometric mean titers (GMTs) as measured by plaque reduction neutralization assay before and after measles vaccination in infants who were 6, 9, or 12 months of age at the time of measles immunization. The Table provides the 95% condence intervals. A neutralizing antibody titer of 120 or greater, dened as protective, is indicated by the dashed line.

Age at Time of Vaccination, mo

Figure 2.Neutralizing antibody responses of infants to measles immunization in the presence and absence of passive antibodies. Shown are the geometric mean titers (GMTs) as measured by plaque reduction neutralization assay after measles vaccination of infants in the presence and absence of passive antibodies. Infants were 6, 9, or 12 months of age at the time of measles immunization. The Table provides the 95% condence intervals. A neutralizing antibody titer of 120 or greater, dened as protective, is indicated by the dashed line.

Age at Time of Vaccination, mo

Figure 3.T-cell proliferative responses of infants before and after measles immunization. Shown is the stimulation index to measles antigen before and after measles vaccination in infants who were 6, 9, or 12 months of age at the time of immunization. Error bars indicate SEs. A positive stimulation index is dened as 3 or greater.

vaccination of 6-, 9-, and 12-month-old infants, respectively (Table). However, when responses of individual infants were evaluated by paired t test, significant increases were detected after vaccination of 9- and 12-month-old infants but not 6-month-old infants (6 months, P = .06; 9 months, P = .03; 12 months, P = .04). The mean (SE) IFN- concentration was 300.6 (147.66) pg/mL in vaccinated adults, which was not significantly different from the infants. Interleukin 4 release was not detected after stimulation of T cells from infants or adults with measles antigen. Interleukin 10 production by PBMC stimulated with measles antigen was evaluated in 38 infants and 8 vaccinated adults. The mean (SE) IL-10 concentrations before and after vaccination were not statistically different in all age groups; the responses were 81.9 (15.54) vs 73.9 (26.44) pg/mL in 6-month-old infants, 84.5 (17.87) vs 77.4 (25.80) pg/mL in 9-month-old infants, and 59.9 (11.98) vs 50.4 (10.49) pg/mL in 12month-old infants. Vaccinated adults had higher IL-10 concentrations than infants with a mean (SE) of 152.4 (32.03) pg/mL (P = .02). No age-related differences in the kinetics of cytokine production, defined as the interval to detection of the peak concentration, were detected. There was no correlation between SI and cytokine responses in individual patients. COMMENT Interference due to passively acquired antibodies among infants younger
530 JAMA, August 12, 1998Vol 280, No. 6

than 12 months has been observed since the live attenuated measles vaccine was introduced in the 1960s.7-13 Mothers born prior to 1957 who had measles immunity from natural disease transferred more neutralizing antibodies to their infants than mothers with vaccine-induced immunity to measles.16,19,20,35 Most women of childbearing age in the United States today have been immunized against measles.19,20,35 Consequently, infants lose measles neutralizing antibody sooner after birth, 15-19 and could benefit from measles vaccination before age 12 months to provide active immunity against infection. Although rare, measles outbreaks in the United States are associated with high rates of morbidity and mortality among infants who have not yet received routine measles vaccination.2,6,36,37 Vaccination of infants younger than 12 months is recommended when exposure to measles is likely.38 However, more information about the immunogenicity of measles vaccine in younger infants is needed to reassess the optimal age for routine vaccination against measles, since the proportion of infants born to mothers with vaccineinduced immunity compared with natural immunity can be expected to increase. Since 48% of 6-month-old infants in our population had undetectable levels of passive antibodies by the most sensitive PRN assay,33 it was possible to evaluate the capacity of the developing immune system to respond to measles vaccine without the confounding vari-

able of interference by neutralizing antibodies acquired from the mother. Vaccination of 6-month-old infants who had no detectable passive antibodies elicited seroconversion in only 67% and seroprotective titers in only 36% of these infants. In contrast, 100% of 9-month-old infants lacking passive antibodies seroconverted and achieved titers considered seroprotective; their responses were not statistically different from those of 12-month-old infants. Our study demonstrates that the persistence of passive antibodies remains an obstacle to measles immunization, affecting responses in about one third of 9-month-old infants and half of 6-monthold infants. Yet, the deficiency of the humoral immune response to measles vaccine among 6-month-old infants without detectable passive antibodies in this study, compared with that among 9- or 12-month-old infants, indicates that some component of the immune response to measles antigens undergoes maturation late in the first year of life. Younger infants may have a functional defect in the TH-cell response to measles antigen since IFN- production did not increase significantly after vaccination of 6-month-old infants. Infants infected with herpes simplex virus and cytomegalovirus also have low or absent IFN- .24,39 T-cell recognition of measles antigens, as measured by in vitro proliferation, was detected among vaccinated infants and adults but the response rates were less than 75% among all cohorts, as has been described in previous studies of cell-mediated immunity to measles.40-42
Immune Response to Measles VaccineGans et al

1998 American Medical Association. All rights reserved. Downloaded from jama.ama-assn.org by guest on February 22, 2012

The limited proliferation of measles-specific memory T cells in in vitro assays may be related to the down-regulation of IL-12 production shown previously to be triggered by measles.43 Interleukin 12, which is secreted by monocytes, is important for optimal TH1 differentiation and plays an integral role in the acquisition of the cell-mediated immune response.44-46 If infant T cells have a diminished capacity to produce IFN- , decreased IL-12 production after measles infection or immunization would be expected to further impair the induction of measles-specific cellular immunity. In vivo, natural measles infection and live attenuated measles vaccine induce transient, nonspecific immunosuppression, characterized by predominance of type 2 TH cell responses and associated with an increase in the spontaneous release of IL-4 and IL-10.42,47-49 Infants may have an age-related susceptibility to the generalized immunosuppression elicited by measles virus in vivo, including attenuated vaccine strains.50 In contrast with the increased spontaneous release of these cytokines, our evaluation of the measles-specific production of IL-4 and IL-10 did not demonstrate a shift toward a predominant TH2 cell response. Naive T cells of neonates shift preferentially toward a TH2 cell response when stimulated by foreign antigens, but whether this pattern persists later in the first year of life has not been determined.39 The decreased synthesis of measles neutralizing antibodies in younger infants may represent impaired T-cell and B-cell interactions. Specifically, a maturational deficiency in CD40 ligand expression by activated T cells could inhibit the development of T-celldependent Bcell immunity to measles.27,29,44,46,51,52 It is also possible that the limited immunogenicity of measles vaccine in 6-month-old infants reflects a defect in antigen presentation by dendritic cells. The switch from naive CD45RA T cells to memory T cells expressing CD45RO is crucial for acquisition of an effective antigen-specific immune response and requires antigen processing by dendritic cells, which have been shown to be functionally immature in neonates.25,26 Finally, diminished humoral immunity to measles in 6-monthold infants could represent a primary B-cell deficiency, although impaired immunity to viral pathogens has usually been associated with altered T-cell dependent responses. Infants have decreased T-cellindependent B-cell activation by polysaccharides and heavily glycosylated proteins made by bacteria and viruses.53 Regardless of the underlying immunologic mechanism, our experience was that the maturational deficiency in the
JAMA, August 12, 1998Vol 280, No. 6

humoral response to measles vaccine was transient in individual infants. Revaccination at age 12 to 15 months resulted in a marked increase in measles neutralizing antibody titers among infants with serologically defined primary vaccine failure. These observations suggest that early immunization does not induce tolerance to subsequent doses of measles vaccine, which was an earlier concern in considering the immunization of infants younger than 12 months.12,54 Our observations are consistent with recent reports that these infants seroconvert after a second dose of measles vaccine.55-58 Most 6-month-old infants with poor humoral immune responses to measles had detectable T-cell proliferation to measles antigen after vaccination. Whether the presence of measles-specific T cells predicts protection, despite low or undetectable titers of neutralizing antibodies, is not known. The importance of cellular immunity against measles is suggested by clinical experience demonstrating that patients with cellular immunodeficiencies are susceptible to severe or fatal measles, whereas children with congenital agammaglobulinemia had no complications from measles and developed immunity to reinfection.14 Nevertheless, seroconversion after measles vaccination, with induction of measles neutralizing antibody titers greater than 120, correlates with protection against wildtype measles infection. 34 Our study demonstrates that host response of most infants immunized at age 6 months is insufficient to achieve this criterion of protective immunity. Given the heterogeneity of the population of women of childbearing age in the United States, which includes women with natural as well as vaccine-induced immunity to measles, interference by passive antibodies will continue to affect the responses of many 6- and 9month-old infants to measles vaccine. Our study indicates that postnatal maturation of the immune system is also likely to restrict the immunogenicity of measles vaccine in 6-month-old infants. This possibility warrants further investigation when considering lowering the recommended age of measles immunization for infants whose mothers have vaccineinduced immunity.
This work was supported by grant AI37127 from the National Institute of Allergy and Infectious Diseases, Bethesda, Md. We are indebted to the families, pediatricians, nursing staff, and laboratory staff of the Palo Alto Medical Foundation, Palo Alto, Calif, for their assistance with this study. References 1. Clements CJ, Cutts FT. The epidemiology of measles: thirty years of vaccination. Curr Top Microbiol Immunol. 1995;191:13-33.

2. Cherry JD, Feigin RD, Lobes LA Jr, et al. Urban measles in the vaccine era: a clinical, epidemiologic, and serologic study. J Pediatr. 1972;81:217-230. 3. Hardy GE Jr, Kassanoff I, Orbach HG, Case GE, Witte JJ. The failure of a school immunization campaign to terminate an urban epidemic of measles. Am J Epidemiol. 1970;91:286-293. 4. Black FL. Measles. In: Evans AS, ed. Viral Infections of Humans: Epidemiology and Control. New York, NY: Plenum Medical Book Co; 1976: 297-316. 5. Babbott FL Jr, Gordon JE. Modern measles. Am J Med Sci. 1954;228:334-361. 6. Aaby P, Clements J, Orinda V. Mortality from measles: measuring the impact. In: World Health Organization, ed. Expanded Programme on Immunization. Geneva, Switzerland:World Health Organization; 1991. 7. Gindler JS, Atkinson WL, Markowitz LE, Hutchins SS. Epidemiology of measles in the United States in 1989 and 1990. Pediatr Infect Dis J. 1992; 11:841-846. 8. Krugman RD, Rosenberg R, McIntosh K, et al. Further attenuated live measles vaccines: the need for revised recommendations. J Pediatr. 1977;91: 766-767. 9. Schluederberg A, Lamm SH, Landrigan PJ, Black FL. Measles immunity in children vaccinated before one year of age. Am J Epidemiol. 1973;97: 402-409. 10. Shelton JD, Jacobson JE, Orenstein WA, Schulz KF, Donnell HD Jr. Measles vaccine efficacy: influence of age at vaccination vs. duration of time since vaccination. Pediatrics. 1978;62:961-964. 11. Shasby DM, Shope TC, Downs H, Herrmann KL, Polkowski J. Epidemic measles in a highly vaccinated population. N Engl J Med. 1977;296:585-589. 12. Wilkins J, Wehrle PF. Additional evidence against measles vaccine administration to infants less than 12 months of age: altered immune response following active/passive immunization. J Pediatr. 1979;94:865-869. 13. Yeager AS, Davis JH, Ross LA, Harvey B. Measles immunization: successes and failures. JAMA. 1977;237:347-351. 14. Cherry JD. Measles. In: Feigin R, Cherry J, eds. Textbook of Pediatric Infectious Diseases. 3rd ed. Philadelphia, Pa: WB Saunders Co; 1992:1591-1609. 15. Black FL. Measles active and passive immunity in a worldwide perspective. Prog Med Virol. 1989; 36:1-33. 16. Lennon JL, Black FL. Maternally derived measles immunity in era of vaccine-protected mothers. J Pediatr. 1986;108:671-676. 17. Markowitz LE, Preblud SR, Fine PE, Orenstein WA. Duration of live measles vaccine-induced immunity. Pediatr Infect Dis J. 1990;9:101-110. 18. Pabst HF, Spady DW, Marusyk RG, et al. Reduced measles immunity in infants in a well-vaccinated population. Pediatr Infect Dis J. 1992;11:525529. 19. Markowitz LE, Albrecht P, Rhodes P, et al, for the Kaiser Permanente Measles Vaccine Trial Team. Changing levels of measles antibody titers in women and children in the United States: impact on response to vaccination. Pediatrics. 1996;97:53-58. 20. Maldonado YA, Lawrence EC, DeHovitz R, Hartzell H, Albrecht P. Early loss of passive measles antibody in infants of mothers with vaccine-induced immunity. Pediatrics. 1995;96:447-450. 21. Sullender WM, Miller JL, Yasukawa LL, et al. Humoral and cell-mediated immunity in neonates with herpes simplex virus infection [published erratum appears in J Infect Dis. 1987;155:838]. J Infect Dis. 1987;155:28-37. 22. Pass RF, Dworsky ME, Whitley RJ, August AM, Stagno S, Alford CA Jr. Specific lymphocyte blastogenic responses in children with cytomegalovirus and herpes simplex virus infections acquired early in infancy. Infect Immun. 1981;34: 166-170. 23. Hayward AR, Herberger MJ, Groothuis J, Levin MR. Specific immunity after congenital or neonatal infection with cytomegalovirus or herpes simplex virus. J Immunol. 1984;133:2469-2473.

Immune Response to Measles VaccineGans et al

531

1998 American Medical Association. All rights reserved. Downloaded from jama.ama-assn.org by guest on February 22, 2012

24. Burchett SK, Corey L, Mohan KM, Westall J, Ashley R, Wilson CB. Diminished interferongamma and lymphocyte proliferation in neonatal and postpartum primary herpes simplex virus infection. J Infect Dis. 1992;165:813-818. 25. Hunt DW, Huppertz HI, Jiang HJ, Petty RE. Studies of human cord blood dendritic cells: evidence for functional immaturity. Blood. 1994;84: 4333-4343. 26. Caux C, Massacrier C, Vanbervliet B, et al. Activation of human dendritic cells through CD40 cross-linking. J Exp Med. 1994;180:1263-1272. 27. Brugnoni D, Airo P, Graf D, et al. Ineffective expression of CD40 ligand on cord blood T cells may contribute to poor immunoglobulin production in the newborn. Eur J Immunol. 1994;24:1919-1924. 28. Nonoyama S, Penix LA, Edwards CP, et al. Diminished expression of CD40 ligand by activated neonatal T cells. J Clin Invest. 1995;95:66-75. 29. van Essen D, Kikutani H, Gray D. CD40 ligandtransduced co-stimulation of T cells in the development of helper function. Nature. 1995;378:620-623. 30. Wedgwood JF, Palmer R, Weinberger B. Development of the ability to make IgG and IgA in newborns and infants. Mt Sinai J Med. 1994;61:409-415. 31. Albrecht P, Ennis FA, Saltzman EJ, Krugman S. Persistence of maternal antibody in infants beyond 12 months: mechanism of measles vaccine failure. J Pediatr. 1977;91:715-718. 32. Albrecht P, Herrmann K, Burns GR. Role of virus strain in conventional and enhanced measles plaque neutralization test. J Virol Methods. 1981;3: 251-260. 33. Ratnam S, Gadag V, West R, et al. Comparison of commercial enzyme immunoassay kits with plaque reduction neutralization test for detection of measles virus antibody. J Clin Microbiol. 1995;33: 811-815. 34. Chen RT, Markowitz LE, Albrecht P, et al. Measles antibody: reevaluation of protective titers. J Infect Dis. 1990;162:1036-1042. 35. Johnson CE, Nalin DR, Chui LW, Whitwell J, Marusyk RG, Kumar ML. Measles vaccine immunogenicity in 6- versus 15-month-old infants born to mothers in the measles vaccine era. Pediatrics. 1994; 93:939-944.

36. Measles outbreaksouthwestern Utah, 1996. MMWR Morb Mortal Wkly Rep. 1997;46:766-769. 37. Atkinson WL, Hadler SC, Redd SB, Orenstein WA. Measles surveillanceUnited States, 1991. MMWR CDC Surveill Summ. 1992;41:1-12. 38. American Academy of Pediatrics. Measles. In: Peter G, ed. 1997 Red Book: Report of the Committee on Infectious Diseases. 24th ed. Elk Grove Village, Ill: American Academy of Pediatrics; 1997: 344-357. 39. Martinez X, Brandt C, Saddallah F, et al. DNA immunization circumvents deficient induction of T helper type 1 and cytotoxic T lymphocyte responses in neonates and during early life. Proc Natl Acad Sci U S A. 1997;94:8726-8731. 40. Ward BJ, Boulianne N, Ratnam S, Guiot MC, Couillard M, De Serres G. Cellular immunity in measles vaccine failure: demonstration of measles antigen-specific lymphoproliferative responses despite limited serum antibody production after revaccination. J Infect Dis. 1995;172:1591-1595. 41. Markowitz L, Katz S. Measles vaccine. In: Plotkin SA, Mortimer EA Jr, eds. Vaccines. 2nd ed. Philadelphia, Pa: WB Saunders Co; 1994:229-276. 42. Griffin DE, Ward BJ, Esolen LM. Pathogenesis of measles virus infection: an hypothesis for altered immune responses. J Infect Dis. 1994;170(suppl 1): S24-S31. 43. Karp CL, Wysocka M, Wahl LM, et al. Mechanism of suppression of cell-mediated immunity by measles virus. Science. 1996;273:228-231. 44. Stuber E, Strober W, Neurath M. Blocking the CD40L-CD40 interaction in vivo specifically prevents the priming of T helper 1 cells through the inhibition of interleukin 12 secretion. J Exp Med. 1996;183:693-698. 45. Schmitt E, Hoehn P, Huels C, et al. T helper type 1 development of naive CD4+ T cells requires the coordinate action of interleukin-12 and interferongamma and is inhibited by transforming growth factor-beta. Eur J Immunol. 1994;24:793-798. 46. Kennedy MK, Picha KS, Fanslow WC, et al. CD40/CD40 ligand interactions are required for T cell-dependent production of interleukin-12 by mouse macrophages. Eur J Immunol. 1996;26: 370-378.

47. Ward BJ, Griffin DE. Changes in cytokine production after measles virus vaccination: predominant production of IL-4 suggests induction of a Th2 response. Clin Immunol Immunopathol. 1993;67: 171-177. 48. Griffin DE. Immune responses during measles virus infection. Curr Top Microbiol Immunol. 1995; 191:117-134. 49. Griffin DE, Ward BJ. Differential CD4 T cell activation in measles. J Infect Dis. 1993;168:275-281. 50. Hussey GD, Goddard EA, Hughes J, et al. The effect of Edmonston-Zagreb and Schwarz measles vaccines on immune response in infants. J Infect Dis. 1996;173:1320-1326. 51. Oxenius A, Campbell KA, Maliszewski CR, et al. CD40-CD40 ligand interactions are critical in T-B cooperation but not for other anti-viral CD4+ T cell functions. J Exp Med. 1996;183:2209-2218. 52. Klaus SJ, Berberich I, Shu G, Clark EA. CD40 and its ligand in the regulation of humoral immunity. Semin Immunol. 1994;6:279-286. 53. Siegrist CA. Potential advantages and risks of nucleic acid vaccines for infant immunization. Vaccine. 1997;15:798-800. 54. Linnemann CC Jr, Dine MS, Roselle GA, Askey PA. Measles immunity after revaccination: results in children vaccinated before 10 months of age. Pediatrics. 1982;69:332-335. 55. Cutts FT, Nyandu B, Markowitz LE, et al. Immunogenicity of high-titre AIK-C or EdmonstonZagreb vaccines in 3.5-month-old infants, and of medium- or high-titre Edmonston-Zagreb vaccine in 6-month-old infants, in Kinshasa, Zaire. Vaccine. 1994;12:1311-1316. 56. McGraw TT. Reimmunization following early immunization with measles vaccine: a prospective study. Pediatrics. 1986;77:45-48. 57. Murphy MD, Brunell PA, Lievens AW, Shehab ZM. Effect of early immunization on antibody response to reimmunization with measles vaccine as demonstrated by enzyme-linked immunosorbent assay (ELISA). Pediatrics. 1984;74:90-93. 58. Stetler HC, Orenstein WA, Bernier RH, et al. Impact of revaccinating children who initially received measles vaccine before 10 months of age. Pediatrics. 1986;77:471-476.

532 JAMA, August 12, 1998Vol 280, No. 6

Immune Response to Measles VaccineGans et al

1998 American Medical Association. All rights reserved. Downloaded from jama.ama-assn.org by guest on February 22, 2012