Subjects, Materials, and Methods

Study design The participants of the present study had been enrolled in a randomized, blinded, controlled study of healthy adolescents conducted at 11 clinical centers in the United States. Three years later, as part of an extension of that trial, follow-up serum samples were obtained from a convenience sample of subjects, 76 of whom had received MCV-4 and 77 of whom had received MPSV-4 (figure 1). Control serum samples were obtained from 88 newly recruited healthy subjects, 14–22 years old, with no history of meningococcal vaccination or disease. These subjects were enrolled at the same centers that had participated in the initial study. For the serum samples from the initial study, a preliminary report of the bactericidal titers measured using a rabbit complement assay was included as part of the ACIP recommendations for prevention and control of meningococcal disease [1], and a more detailed analysis has been published elsewhere [5]. We evaluated serum samples with sufficient volume for testing, which represented approximately two-thirds of the samples collected at the 3-year postvaccination visit. Use of the serum samples for our investigations was approved by the institutional review board of Children’s Hospital and Research Center, Oakland, California

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1

Flow diagram showing the sources of serum samples used in the study. MCV-4, quadrivalent meningococcal conjugate vaccine; MPSV-4, quadrivalent meningococcal polysaccharide vaccine

Serological assessment All assays were performed by personnel who were blinded to the vaccine groups. Serum quantities were limited, and relatively large volumes were required for the passive protection assay; therefore, we limited serological analyses to capsular serogroups C, W-135, and Y. Serogroup A disease has not been observed in North America for >20 years [2, 6] and is rare in Europe [7]. Thus, capsular serogroups C, W-135, and Y account for nearly all vaccine-preventable meningococcal disease in these regions. The strains and methods used in the human complement serum bactericidal antibody assay have been described elsewhere [8, 9]. Bactericidal activity was measured using log-phase, washed bacteria. The bacteria had been grown for 2 h at 37°C in Mueller-Hinton broth supplemented with 0.25% glucose (wt/vol), were rediluted 1:7 in fresh Mueller-Hinton broth supplemented with 0.25% glucose (wt/vol) and 0.02 mmol/L cytidine-5′-monophospho-N-acetylneuraminic acid (Sigma-Aldrich), and were incubated for an additional 2 h to an OD at 620 nm of ∼0.6. Test serum samples were heated to 56°C for 30 min, to inactivate internal complement. The extrinsic complement source was serum samples from healthy adults with no detectable intrinsic bactericidal activity [10] and anticapsular antibody concentrations <0.08 μg/mL, as measured by a radioantigen binding assay (RABA) [8, 9]. The bactericidal titer was defined as the dilution of serum resulting in a 50% decrease in the number of colony-forming units per milliliter versus that at time 0 in negative control samples. Bactericidal titers in the rabbit complement assay were measured at Sanofi Pasteur (Swiftwater, PA), and assays were performed as described elsewhere [5]. Concentrations of total (IgM, IgG, and IgA) serum anticapsular antibody to serogroups C, W-135, and Y were measured using RABA, as described elsewhere [8]. Serogroup C anticapsular antibody avidity constants, K _a, were measured using RABA [8, 9]. Assays of avidity were limited to serum samples with serogroup C anticapsular antibody concentrations between 0.3 and 1.9 μg/mL, because it was not possible to perform the assays on samples with <0.3 μg/mL of antibody, and, in previous studies, nearly all serum samples with >1.9 μg/mL of antibody, irrespective of avidity, conferred passive protection [9, 11]

Passive protection As described elsewhere [9], 5–7-day-old out-bred Wistar rat pups (Charles River Laboratories) were randomly redistributed to nursing mothers. At time 0, 100 μL of a 1:4 dilution of serum was administered intraperitoneally (ip) to 3–6 rats/serum sample. Two hours later, rats were challenged ip with 100 μL of washed, log-phase N. meningitidis serogroup C strain 4243. In different experiments, the challenge dose of bacteria ranged from 1230 to 5300 cfu/rat. Eighteen hours later, blood samples were obtained from anesthetized rats by puncture of the heart with a needle and syringe containing ∼25 U of heparin without preservative (American Pharmaceutical Partners). Aliquots equivalent to 1, 10, and 100 μL of blood were plated onto chocolate agar. The number of colony-forming units per milliliter of blood was determined after overnight incubation of the plates at 37°C in 5% CO₂. Protection was defined as a 95% decrease in the geometric mean number of colony-forming units per milliliter in the 3–6 rats given a specific test serum, compared with that in a group of rats treated with a negative control serum that gave no protection (average, ∼400,000 cfu/mL in blood obtained 18 h after challenge)

Statistical analysis For geometric means, serum samples with values below the lower limit of detection were assigned a value of half of the lower limit (for bactericidal titers, 1:2; for anticapsular antibody concentrations, 0.04 μg/mL). The respective geometric means of the bactericidal titers or anticapsular antibody concentrations and associated 2-sided 95% confidence intervals (CIs) were computed from the log₁₀ values. The proportion of vaccinated participants with bactericidal titers ⩾1:4 (which is considered to be a protective titer against serogroup C infection [12]) was computed. Differences found in the proportion of participants in the groups with bactericidal titers ⩾1:4 were analyzed by χ² or Fisher’s exact test. All statistical comparisons between groups were 2-tailed

Results

Serum samples were obtained between February and April 2004. Results for the primary analysis of serum samples, geometric mean bactericidal titers measured in the rabbit complement assay, and the percentage of serum samples with titers ⩾1:128 were similar in the subset of samples we tested and the full set of samples collected for the 3-year follow-up study (table 1). Of the serum samples we tested, 52 were from MCV-4 recipients (mean±SD age at vaccination, 14.0±1.8 years), 48 were from MPSV-4 recipients (mean±SD age at vaccination, 14.3±1.9 years), and 60 were from controls. The mean±SD age at the 3-year follow-up was 17.2±1.8 years for MCV-4 recipients and 17.5±1.9 years for MPSV-4 recipients. The mean±SD age of controls was 16.1±2.1 years

Anticapsular antibody responses Figure 2 shows the reverse cumulative distribution of anticapsular antibody concentrations in serum samples obtained 3 years after vaccination. For all 3 capsular serogroups, geometric mean anticapsular antibody concentrations were higher in vaccine recipients, compared with those in controls (P⩽.0001). No significant differences in the serum anticapsular antibody concentrations for any capsular serogroup were observed between MCV-4 and MPSV-4 recipients

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 2

Reverse cumulative distribution of serum anticapsular antibody concentrations 3 years after vaccination with either quadrivalent meningococcal conjugate vaccine (MCV-4) or quadrivalent meningococcal polysaccharide vaccine (MPSV-4). The respective antibody concentrations in the vaccine groups were not significantly different (P=.70, 1.0, and .09, for antibodies against serogroups C, W-135, and Y, respectively). Both vaccine groups had higher antibody concentrations than did the control group (P⩽.0001)

Serum bactericidal antibody responses As is shown in figure 3, a higher proportion of MCV-4 and MPSV-4 recipients had bactericidal titers ⩾1:4, compared with controls (MCV-4 recipients, P⩽.003 for all 3 capsular serogroups; MPSV-4 recipients, P⩽.01 for serogroups C and Y and P=.06 for serogroup W-135). The proportion of vaccinated participants with titers ⩾1:4 against serogroup W-135 was higher in the MCV-4 group, compared with that in the MPSV-4 group (44% vs. 21%; P=.01), whereas the respective proportions with titers ⩾1:4 against serogroups C and Y were similar in the 2 vaccine groups (figure 3). The higher titers against W-135 in MCV-4 recipients may reflect the relatively lower W-135 bactericidal responses of the group vaccinated with MPSV-4, compared with the respective responses to serogroups C and Y

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 3

Serum bactericidal activity 3 years after vaccination with either quadrivalent meningococcal conjugate vaccine (MCV-4) or quadrivalent meningococcal polysaccharide vaccine (MPSV-4). A Reciprocal geometric mean titers (GMTs) against serogroups C, W-135, and Y. B Percentage of participants with bactericidal titers ⩾1:4. In both panels, error bars show 95% confidence intervals, and an asterisk indicates that the difference between the respective vaccine groups is significant (P⩽.02)

Passive protective activity Bactericidal activity is measured using bacteria grown in artificial media. Although the results correlate with protective immunity [12], bacteria grown on mucosal surfaces or in blood may express different genes than do those grown in culture media [13–15], and this can potentially affect their susceptibility to bactericidal activity. Therefore, we assessed the ability of serum antibody to confer passive protection in vivo against serogroup C meningococcal bacteremia in an infant rat model. Serum samples from 51 MCV-4 recipients, 47 MPSV-4 recipients, and a subset of 25 randomly selected controls were tested for their ability to confer protection against serogroup C meningococcal bacteremia in infant rats. Protective activity was present in 76% of serum samples from MCV-4 recipients, 49% of serum samples from MPSV-4 recipients, and 32% of serum samples from controls (P<.0005, χ² test) (table 2). Compared with negative control rats treated with serum that gave no protection, rats treated with serum from MCV-4 recipients had a mean decrease in bacterial concentrations that was 0.8 log₁₀ cfu/mL greater than that in rats treated with serum from MPSV-4 recipients (2.5 log₁₀ cfu/mL decrease vs. 1.7 log₁₀ cfu/mL decrease, respectively; P=.01). The mean decrease in bacterial concentrations in rats treated with serum from controls was 0.9 log₁₀ cfu/mL (P<.0001, vs. serum from recipients of either vaccine)

As is shown in figure 4A the vast majority of serum samples from vaccinated participants with bactericidal titers ⩾1:4 conferred protection: 100% of the serum samples from MCV-4 recipients and 78% of the serum samples from MPSV-4 recipients (for MCV-4 recipients vs. MPSV-4 recipients, P<.05). Of interest, protection also was conferred by 63% of the serum samples with bactericidal titers <1:4 from MCV-4 recipients and 31% of the serum samples with bactericidal titers <1:4 from MPSV-4 recipients (P=.02). The respective results were similar when the passive protection data were stratified by serum samples with bactericidal titers <1:128 or ⩾1:128 in the rabbit complement assay (for MCV-4 recipients vs. MPSV-4 recipients, P=.02) (figure 4B )

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 4

Passive protection of infant rats against serogroup C meningococcal bacteremia. A Data stratified by bactericidal activity measured in a human complement assay. B Data stratified by bactericidal activity measured in a rabbit complement assay. C Data stratified by serogroup C anticapsular antibody concentrations measured in a radioantigen binding assay. In all panels, error bars show 95% confidence intervals, and an asterisk indicates that the difference between the respective vaccine groups is significant (P⩽.05)

As is shown in Figure 4C serum samples with <0.3 μg/mL of serogroup C anticapsular antibody measured by RABA were unlikely to show protective activity, irrespective of the vaccine group, whereas the majority of serum samples with ⩾1 μg/mL of serogroup C anticapsular antibody conferred protection against bacteremia (96% of the serum samples from MCV-4 recipients vs. 76% of the serum samples from MPSV-4 recipients; P<.05). In serum samples containing 0.3–0.99 μg/mL antibody, the difference in protective activity between the vaccine groups was more pronounced, with 77% of the serum samples from MCV-4 recipients versus 25% of the serum samples from MPSV-4 recipients conferring protection (P<.01)

As is described in Subjects, Materials, and Methods, we measured antibody avidity in serum samples with serogroup C anticapsular antibody concentrations between 0.3 and 1.9 μg/mL. The mean avidity of serogroup C anticapsular antibodies in serum samples from MCV-4 recipients (75.5 nmol/L) was significantly higher than that in serum samples from MPSV-4 recipients (54.2 nmol/L; P<.01) or controls (50.5 nmol/L; P<.01). No significant between-group difference in avidity was observed in serum samples from MPSV-4 recipients and controls (P>.5)

Discussion

We observed higher W-135 bactericidal antibody titers in serum samples obtained 3 years after MCV-4 vaccination of healthy adolescents, compared with those in adolescents vaccinated with MPSV-4. Vaccination with either MPSV-4 or MCV-4 was associated with an increase in the ability of serum samples to provide passive protection against capsular serogroup C bacteremia in infant rats. The MCV-4 group also had higher serogroup C anticapsular antibody avidity. These results and previous data on bactericidal titers measured using a rabbit complement assay in the same population [1] are consistent with a longer duration of antibody persistence and protection after vaccination with MCV-4

Meningococcal polysaccharide conjugate vaccines were developed to provide increased immunogenicity in infants and higher antibody avidity (compared with that conferred by plain polysaccharide vaccines) and priming for booster antibody responses at all ages [11, 16–19]. Monovalent conjugate vaccines against capsular serogroup C, which were licensed in the United Kingdom in 1999 [20, 21] and were subsequently introduced in other European countries and in Canada, have provided these benefits. Use of these vaccines also has decreased nasopharyngeal colonization and transmission of serogroup C meningococci in the general population [22, 23]. In a recent study in the United Kingdom, the effectiveness of monovalent C conjugate vaccines given at 2, 3, and 4 months of age, although high during the first year, decreased dramatically between 1 and 4 years after vaccination, whereas the effectiveness of these vaccines remained high for at least 4 years after vaccination of adolescents [24]. These results paralleled the duration of bactericidal titers, which lasted in infants for 1 year but were present in adolescents for 4 years [25]. Borrow et al. showed that infants vaccinated with serogroup C conjugate vaccines at 2, 3, and 4 months of age had evidence of immunologic memory, which persisted for at least 4 years [26]. Thus, immunologic memory alone, without protective anticapsular antibody concentrations, may not confer protection against serogroup C disease in infants. Therefore, in the present study, we focused on the persistence of serum antibodies as being the most important marker of protective immunity [12]

Serum bactericidal antibody activity 3 years after vaccination of adolescents with MCV-4 or MPSV-4, as measured in a baby rabbit complement assay, was described in a recent study [1, 5]. Baby rabbit complement augments bactericidal titers, compared with the effect of human complement [10, 27, 28]; consequently, titers obtained using baby rabbit complement are higher [10, 29]. We used human complement to assay a subset of the same serum samples that were previously tested with rabbit complement. Our assay differed from the baby rabbit complement assay in a number of other important respects, including the use of log-phase broth-grown organisms, compared with bacteria grown on solid media in the baby rabbit complement assay, and different buffers and test strains. Thus, the respective data are not directly comparable. In the study using the baby rabbit complement assay, 71%, 83%, and 96% of adolescents vaccinated with MCV-4 had titers ⩾1:128 against capsular serogroups C, W-135, and Y, respectively, compared with 57%, 68%, and 83%, respectively, of adolescents vaccinated with MPSV-4 [1]. In our study, the proportions of participants with protective bactericidal titers (⩾1:4 in the human complement assay) were lower in both the MCV-4 group and the MPSV-4 group. Despite these differences, in both studies, serum bactericidal titers 3 years after vaccination of 11–18-year-olds with MCV-4 or MPSV-4 were consistently higher than those in controls, and bactericidal titers after MCV-4 vaccination were either similar to or higher than those after MPSV-4 vaccination

We found no significant differences in the total serum anticapsular antibody concentrations against capsular groups C, W-135, and Y as measured by RABA between the 2 vaccine groups. The RABA as performed in the present study provides a measure of the total serum anticapsular antibody concentration and does not distinguish between low- and high-avidity antibody, which can affect antibody functional activity [8]. As was described previously in 2-year-olds vaccinated with MCV-4 [11, 16], there was evidence in adolescents vaccinated with MCV-4 that the quality of the serum anticapsular antibody was superior to that of the antibody elicited by MPSV-4. Significantly higher bactericidal titers against W-135 were observed in MCV-4 recipients, compared with those in MPSV-4 recipients. MCV-4 also elicited higher serogroup C anticapsular antibody avidity and greater serogroup C passive protective activity in the infant rat bacteremia model, compared with MPSV-4. The difference in passive protection between the vaccine groups was particularly striking for serum samples with bactericidal titers <1:4; compared with serum samples from the MPSV-4 group, twice as many serum samples from the MCV-4 group conferred protection. A similar result was found between the 2 vaccine groups when we analyzed the protective activity of serum samples with bactericidal titers <1:128, as measured in a rabbit complement assay. The most likely mechanism for passive protective activity with low or absent complement-mediated bactericidal activity is Fc- and C3b-mediated opsonization [30, 31]

Although a serum bactericidal titer ⩾1:4 is specific for protection in the human complement assay, this threshold may be an insensitive indicator of protection [12]. In our study and in a previous study of naturally acquired serogroup C immunity [9], bactericidal titers ⩾1:4 predicted protection against bacteremia in the infant rat model. However, one-third to two-thirds of serum samples with bactericidal titers <1:4 also conferred passive protection. Similarly, Goldschneider et al. found that bactericidal titers ⩾1:4 were highly protective in military recruits, but 62% of recruits with bactericidal titers <1:4 and who were colonized with the epidemic serogroup C strain did not develop meningococcal disease [12]. These data suggest that bactericidal titers ⩾1:4, when measured in a human complement assay, predict protection but may underestimate the extent of protective meningococcal immunity in unvaccinated and vaccinated populations. In contrast, the higher titers measured in a rabbit complement assay may be more sensitive for estimating protective immunity than are titers measured in a human complement assay [10], but they may also be less specific. For example, in the infant rat model, for the MCV-4 group in the present study, the sensitivity for passive protection of a titer ⩾1:4 measured using human complement was 49% and the specificity was 100%. The corresponding sensitivity and specificity of a titer ⩾1:128 measured using rabbit complement were 79% and 42% (table 3)

The present study had certain limitations. The unvaccinated controls were not part of the original randomized study. However, they were enrolled at the same sites that participated in the follow-up of the vaccine recipients. Also, the sex and age distributions of the unvaccinated controls were similar to those of the vaccine recipients at the time of follow-up. Thus, the serum anticapsular antibody concentrations in controls should reflect the concentrations in vaccinated participants had they not received vaccine. Another limitation is that we tested passive protection in serum samples from only a randomly selected sample of 25 controls, which resulted in wide CIs about the point estimate of the proportion of serum samples with protective activity. However, the sample size of control serum samples was sufficient to detect significant differences in protective activity between the control group and the vaccinated groups (table 2). Finally, the available volume of serum from all 3 groups was too small to allow us to test antibody responses to all 4 capsular serogroups covered by the vaccines. Additional studies will be needed to identify any differences in responses between capsular serogroup A and the 3 capsular serogroups we tested

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Analysis of the serum samples and demographics of the study population

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 2

Passive protection against bacteremia in infant rats challenged with Neisseria meningitidis serogroup C

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 3

Sensitivity and specificity of bactericidal activity, measured in a human or a rabbit complement assay, for predicting passive protection

With the recent ACIP recommendations for use of MCV-4 at 11–12 years of age, the duration of protection after vaccination is a critical clinical question. The possible requirement of a booster dose—and its timing, whether given before entry to high school or college—needs to be ascertained, because the risk of acquiring invasive meningococcal disease increases after 12 years of age, and death and other devastating outcomes are more common in this age group than in children or adults [1, 32]. Our data indicate that the duration of protective immunity conferred after MCV-4 vaccination of adolescents was at least equal to or greater than that conferred after MPSV-4 vaccination, which is estimated to be 3–5 years [1]. Follow-up studies of the adolescents who participated in this vaccine trial are ongoing, and the results will provide additional information on the duration of immunity