Within this phase IV single-center and open-label study, invitations were sent to middle-aged adults (between 50 and 65 years of age) in Amersfoort, a city in the middle of the Netherlands, during August/September 2014. The Dutch community-based administration was used to send the invitations. Potential participants were excluded based on the following criteria: antibiotic use or fever (>38°C) within the last 14 days, serious diseases demanding immune suppressive medical treatment within the last 3 months, a known or suspected immune deficiency, a blood coagulation disorder, a neurologic disorder, administration of blood products in the past 6 months, serious surgery within the last 3 months, the use of hormone supplementation, pregnancy, a suspected allergy toward the vaccine components, history of serious adverse events after previous vaccinations, a previous meningococcal vaccination, a previous meningococcal episode, a tetanus vaccination within the last 5 years, and any vaccination in the month before enrollment. Written informed consent was obtained from all participants prior to enrollment and all procedures were in accordance with the Declaration of Helsinki. The medical ethical committee: Medical Research Ethics Committees United (MEC-U) approved the study and the study was registered at the Dutch trial register (NTR4636).

A pre-vaccination blood sample was taken from all participants. Subsequently, all participants received the tetravalent meningococcal vaccine conjugated to TT (MenACWY-TT, Nimenrix, GlaxoSmithKline) by intramuscular administration. Blood samples were taken 7 days, 28 days, and 1 year post-vaccination. Serum samples were collected using serum clotting tubes (BD Biosciences) and were stored at −20°C until further use. All participants were sampled during evening hours. In addition, all participants filled in a short health questionnaire during the first appointment.

IgG and IgG subclass concentrations specific for the polysaccharides of the meningococcal groups C, W, and Y (MenCWY), and TT were determined by the fluorescent-bead-based-multiplex immuno assay as previously described (25–28). An internationally accepted TT-specific IgG concentration of 0.01 IU/ml was used as protection level (29). A similar method was used to determine the MenCWY-specific IgM concentrations, with the modification of using a donkey anti-human IgM RPE conjugate (Fc5μ specific, Jackson ImmunoResearch).

MenA-specific measurements were left out of the analysis due to interference of varying levels of cross-reacting antibodies to MenA, present in these older age cohorts, in the immunoassay. Moreover, the analysis of MenCWY-specific responses is perceived sufficient to answer our primary research question in this study.

The MenCWY-specific protective antibody titers were determined with the serum bactericidal antibody assay using baby rabbit complement (rSBA) (Pelfreez, LOT#13035EL), as described previously (30, 31). The MenC, MenW, and MenY strains used in the rSBA were C11, MP01240070, and S-1975, respectively, kindly donated by Prof. Dr. Ray Borrow from the Vaccine Evaluation Unit at Manchester (PHE). The bactericidal titer was expressed as the reciprocal of the highest serum dilution yielding ≥50% killing after 60 min of incubation. The internationally accepted correlate of protection used was an rSBA titer of ≥8 for all groups, whereas a titer of ≥128 was used as a more conservative measure for long-term protection (31, 32). Participants with an rSBA titer below detection level (rSBA seronegative) were assigned an rSBA titer of two for statistical purposes. A representative group of 100 persons was selected for the functional rSBA analysis, based on varying IgG levels from low to high concentrations, for all three meningococcal groups. For MenW and MenY, the same selection of participants was used. This selection had an overlap of 71 participants with the selection for MenC.

Gullsorb reagent human IgG (Meridian Biosciences™) was used for IgG depletion and goat-Anti-human IgM (μ-chain specific)-Agarose antibody beads (Sigma Aldrich) for IgM depletion in the serum samples.

Normal distribution of the data was checked prior to each analysis and only cases with samples available at all time points were included in the analysis. Geometric mean concentrations (GMCs) with 95% confidence intervals [95% CI] were calculated for the MenCWY-PS- and TT-specific IgG responses, as well as the MenCWY-PS-specific IgG subclass and IgM responses. MenCWY-PS-specific IgG and IgM concentrations were log transformed, after which the repeated measurements ANOVA was used to analyze the post-vaccination response. rSBA geometric mean titers (GMTs) with the corresponding 95% CI were calculated. Proportions and 95% CI of participants with an rSBA ≥8 and ≥128 were calculated with the Wilson/Brown test. The proportion of participants with an rSBA ≥8 and ≥128 were compared between participants with and without a pre-vaccination serum bactericidal assay (SBA) titer with the Chi-squared test, whereas differences in GMTs were analyzed with the Mann–Whitney U test. Differences in IgG ratios between the meningococcal groups were analyzed with the Mann–Whitney U test preceded by the Kruskal–Wallis test and corrected for multiple comparisons with the Bonferroni correction.

Correlations between the rSBA titers and the IgG and IgM responses 28 days and 1 year post-vaccination were determined using the Spearman’s rho correlation test. Graphpad Prism V7 and SPSS V22.0 were used for the statistical analysis. A p-value <0.05 was considered statistically significant.

To study the duration of protection, we used a Bayesian approach with a multilevel longitudinal model to statistically predict the dynamics of antibody concentrations (33). We performed Markov-chain Monte Carlo simulations to find the appropriate joint distribution of parameters which best could adjust the model to the measured rSBA titer datasets for MenC, MenW, and MenY. Assuming a simple exponential decay after reaching the time to peak (t₁) would underestimate the persistence of protection (33–35). Therefore, we modeled the rSBA titer decay assuming that it follows a bi-exponential decay curve of the form: y(t>t1)=y1(e−α1(t−t1)+fe−α2(t−t1))/(1+f) where y₁ is the peak antibody concentration level, α₁ and α₂ are the respective decay rates of each exponential component, and f regulates the contribution of the exponential components.

The initial raise of antibody levels was assumed to be exponential until the time to peak t₁. Per individual, the three measured antibody levels were used to fit the multilevel model with their respective time point. The time to peak (t₁) in the model was set to follow a lognormal distribution with a mean of 12 days (SD 8.5 days). The long-term antibody persistence was described using the decay curves predicted by the models. To assess the duration of protection, we focused on the predicted percentages of participants that possessed an rSBA titer >8 and >128, for any given time up to 10 years post-vaccination.

A total of 204 middle-aged adults (mean age: 57.7 years; range 50–65 years; 52% males) participated in the study (Figure 1). All participants received the MenACWY-TT vaccine. Overall, 194 participants (95.1%) completed the study, with blood drawings before vaccination and at 7, 28 days, and 1 year post-vaccination. Additional baseline participant characteristics are presented in Table S1 in Supplementary Material.

At first, longitudinal MenCWY and TT-specific IgG responses were investigated (Figures 2A–C; Figure S1 in Supplementary Material). Pre-vaccination, low IgG levels were observed for the different meningococcal groups. Seven days post-vaccination, IgG responses were enhanced for all three groups (Figures 2A–C), although a higher increase in the IgG response (ratio day 7/pre) was observed for MenC as compared to MenW and MenY (Table 1). This was also reflected in the number of persons with an IgG fold increase above four at day 7, which was 58.8, 20.1, and 22.5% for MenC, MenW, and MenY, respectively (Table 1). Moreover, compared to the other groups, a significant higher increase in MenC-specific IgG was observed 28 days post-vaccination (ratio day 28/pre) (Table 1). Taken together, these results are suggestive of a booster response to MenC as compared to a naïve response in the majority of the participants to MenW and MenY. Nonetheless, based on the IgG response 7 days post-vaccination, a few participants also showed booster responses for MenW and MenY as well as naïve responses to MenC. In addition, MenC-specific IgG concentrations showed a larger decay from 28 days to 1 year (ratio day 28/1 year) post-vaccination than seen for MenW and MenY IgG levels, resulting in comparable IgG GMCs 1 year post-vaccination for all three groups (Table 1). Notably, these IgG concentrations were still significantly enhanced compared with pre-vaccination levels (Figures 2A–C). The TT conjugate clearly induced a booster response in the participants, since 71.6% of the persons showed highly significant increases in IgG concentrations 7 days post-vaccination (Table 1). In the majority of the middle-aged persons, pre-vaccination TT-specific IgG concentrations were above the protection level of 0.01 IU/ml and these were significantly enhanced up to 1 year post-vaccination (Table 1; Figure S1 in Supplementary Material). Within this cohort, we did not observe any effects of age, gender, and CMV seropositivity on the IgG responses (Table S2 in Supplementary Material).

Remarkably, a robust IgM response was observed for all meningococcal groups and these IgM levels were still enhanced 1 year post-vaccination (Figures 2D–F). A significant negative correlation was found between the IgM responses and age for MenC and MenW, whereas a negative trend was observed for MenY (Table 2). However, the low predictive value of the model (R² model) indicates that age and pre-existing IgM concentrations were not the major factors predicting the IgM concentrations post-vaccination.

A representative group of 100 persons was selected for the functional rSBA analysis, based on the IgG concentrations. As expected, the majority of the participants possessed pre-vaccination rSBA titers below the internationally accepted protection level of 8. MenC, MenW, and MenY-specific pre-vaccination rSBA titers ≥8 were observed in 18, 23, and 27% of the middle-aged participants (Table 3). Moreover, 28 days post-vaccination, the vast majority of the participants possessed an rSBA titer ≥8, which was 94, 99, and 97% for MenC, MenW, and MenY, respectively. In addition, 92, 97, and 95% for MenC, MenW, and MenY, respectively, possessed an rSBA titer ≥128, the cutoff titer used for long-term protection (Table 3). One-year post-vaccination, protective antibody titers were still found in 76–94% of the participants (Table 3). Surprisingly, at this time point, slightly higher rSBA titers were observed for MenW and MenY compared with MenC. Moreover, a small, but significant negative correlation of the rSBA titer was found with age; 28 days post-vaccination for MenC (rho: −0.239, p: 0.017) and 1 year post-vaccination for MenW (rho: −0.300, p: 0.002) (Table S3 in Supplementary Material). No gender differences were observed (data not shown).

Using a bi-exponential decay model, the proportion of participants with rSBA titers above the protective cutoff level of 8 at 10 years post-vaccination was estimated to be around 40, 40, and 60% for MenC, MenW, and MenY, respectively (Figures S2A–C in Supplementary Material). Moreover, 20, 20, and 40% of the participants were expected to possess rSBA titers above the protective cutoff level of 128 at 10 years post-vaccination for MenC, MenW, MenY, respectively (Figures S2D–F in Supplementary Material).

In one-fourth of the participants, a detectable pre-vaccination rSBA titer (rSBA ≥ 4) was observed (Figures 3A–C). Subsequently, we compared the vaccine response between participants being seropositive (rSBA ≥ 4) or being seronegative at the pre-vaccination time point (Figure 3; Table S4 in Supplementary Material). Similar rSBA titers were observed in both groups 28 days post-vaccination for all meningococcal groups (Figures 3A–C; Table S4 in Supplementary Material), indicating a robust increase in rSBA titer in the seronegative participants. At day 28, 89.9, 96.1, and 93.2% of the seronegative participants showed an rSBA titer ≥128 for MenC, MenW, and MenY, respectively, compared with 100% of the seropositive participants (Figures 3D–F; Table S4 in Supplementary Material). One-year post-vaccination, rSBA titers showed a significantly higher decay in the seronegative participants, reaching significance for MenW and MenY, and resulting in rSBA titers below the protection limit (rSBA ≥ 8) in some participants (Figure 3; Table S4 in Supplementary Material).

Since the rSBA titers and IgG responses for the different groups did not show a similar pattern, the correlations between the IgG and rSBA responses for the three meningococcal groups were determined (Figures 4A–C). A moderate correlation was found for MenC (28 days: rho 0.561, p < 0.001, 1 year: rho 0.548, p < 0.001) (Figure 4A), whereas the correlations for MenW (28 days: rho 0.356, p < 0.001, 1 year: rho 0.307, p < 0.002) and MenY (28 days: rho 0.201, p = 0.045, 1 year: rho 0.214, p = 0.033) were rather low (Figures 4B,C).

This encouraged us to investigate whether MenC-, MenW-, and MenY-specific IgM responses could explain this low correlation. Remarkably, a good correlation was found between the rSBA titers and the IgM responses for both MenW (28 days: rho 0.687, p < 0.001, 1 year: rho 0.761, p < 0.001) and MenY (28 days: rho 0.690, p < 0.001, 1 year: rho 0.729, p < 0.001) (Figures 4E,F), whereas a moderate correlation was found for MenC (28 days: rho 0.594, p < 0.001, 1 year: rho 0.529, p < 0.001) (Figure 4D). In order to confirm the important contribution of the IgM antibodies to the rSBA response, IgG was depleted in a subset of the samples varying in SBA titer (Figures S3A,C in Supplementary Material). MenW and MenY rSBA titers remained almost equal after IgG removal. By contrast, these rSBA titers were largely reduced after IgM depletion (Figures S3B,D in Supplementary Material), although the rSBA titer was mostly retained in participants with very high IgG concentrations. The data suggest that the protective rSBA titers are mediated via induction of IgM type antibodies, mainly for MenW and MenY.

Finally, we investigated whether IgG subclass responses differed between the naïve (MenW and MenY) and booster (MenC) responses at middle-age, as IgG subclass responses have previously been shown to affect the antibody functionality. IgG subclass responses were highly variable for all three groups. A slightly skewed IgG2 response was observed for MenC and MenW, whereas the response to MenY showed IgG1 skewing, as shown by the IgG1/IgG2 ratio (Table 4). Besides a primarily MenC-specific IgG2 skewed response, 1 year post-vaccination IgG1 and IgG2 contributed equally to the IgG response, suggesting a more rapid decay of IgG2 than IgG1 over time (Table 4), which was not seen for the IgG2 skewed MenW response. Also, the IgG1 skewed response to MenY remained stable.