Direct Comparison of Immunogenicity Induced by 10- or 13-Valent Pneumococcal Conjugate Vaccine around the 11-Month Booster in Dutch Infants.

BACKGROUND & AIMS: Since 2009/10, a 10- and a 13-valent pneumococcal conjugate vaccine (PCV) are available, but only the 10-valent vaccine is now being used for the children in the Netherlands. As the vaccines differ in number of serotypes, antigen concentration, and carrier proteins this study was designed to directly compare quantity and quality of the antibody responses induced by PCV10 and PCV13 before and after the 11-month booster.
METHODS: Dutch infants (n = 132) were immunized with either PCV10 or PCV13 and DTaP-IPV-Hib-HepB at the age of 2, 3, 4 and 11 months. Blood samples were collected pre-booster and post-booster at one week and one month post-booster for quantitative and qualitative immunogenicity against 13 pneumococcal serotypes, as well as quantitative immunogenicity against diphtheria, tetanus, pertussis and Haemophilus influenzae type b. We compared immunogenicity induced by PCV13 and PCV10 for their ten shared serotypes.
RESULTS: One month post-booster, pneumococcal serotype-specific IgG geometric mean concentrations (GMCs) for the PCV13 group were higher compared with the PCV10 group for six serotypes, although avidity was lower. Serotype 19F showed the most distinct difference in IgG and, in contrast to other serotypes, its avidity was higher in the PCV13 group. One week post-booster, opsonophagocytosis for serotype 19F did not differ significantly between the PCV10- and the PCV13 group.
CONCLUSION: Both PCV10 and PCV13 were immunogenic and induced a booster response. Compared to the PCV10 group, the PCV13 group showed higher levels for serotype 19F GMCs and avidity, pre- as well as post-booster, although opsonophagocytosis did not differ significantly between groups. In our study, avidity is not correlated to opsonophagocytotic activity (OPA) and correlations between IgG and OPA differ per serotype. Therefore, besides assays to determine IgG GMCs, assays to detect opsonophagocytotic activity, i.e., the actual killing of the pneumococcus, are important for PCV evaluation. How differences between the two vaccines relate to long-term protection requires further investigation. TRIAL REGISTRATION: www.trialregister.nl NTR3069.

RCT Entities:   Population Interventions Outcomes

Introduction

Streptococcus pneumoniae (SP) is an important cause of morbidity and mortality worldwide, with the highest disease incidence among children under 2 years of age. Carriage of SP is often asymptomatic but can result in non-invasive mucosal infections or invasive pneumococcal disease (IPD). More than 90 SP serotypes of have been identified, of which about 20 serotypes tend to cause IPD [1-3]. The 2006 introduction of a 7-valent pneumococcal conjugate vaccine (PCV7) for infants in the Netherlands dramatically decrease SP carriage and IPD for its seven serotypes [4-10], in line with other countries that implemented PCV7 [11, 12]. However, carriage and IPD related to serotypes not covered by PCV7 have increased since its introduction of PCV7 [4–7, 9, 10]. In 2009, two new pneumococcal conjugate vaccines were licensed that provide protection against 10 (PCV10) or 13 (PCV13) serotypes. They share serotypes 1, 4, 5, 6B, 7F, 9V, 14, 18C, 19F and 23F, and PCV13 in addition includes 3, 6A and 19A. In the Netherlands, all children born since March 2011 have received PCV10 instead of PCV7. Carriage of the three additional serotypes of PCV10 (1, 5 and 7F) was already low before its introduction, and scarcely changed in the first 1.5 year afterward [9]. However, overall IPD incidence decreased in the youngest age group (<2 years), and some decrease has been seen in adults >50 years of age [10]. Even if PCV10 should prove superior to PCV7, the question remains whether PCV13 might offer additional improvement. Besides differing in the number of serotypes, PCV10 and PCV13 differ in the carrier protein and the amount of antigen per serotype, factors that might influence the induction of B-cell responses and consequently antibody responses [13-19]. Of the PCV10 serotypes, 8 are conjugated to protein D of non-typeable H. influenzae, while 18C is conjugated to tetanus toxoid and 19F is conjugated to diphtheria toxoid. For PCV7 and PCV13, the carrier protein is CRM197, a non-toxic variant of diphtheria toxin.

Although PCV10 and PCV13 have both been compared with PCV7 as to induction of immunogenicity [20-26], to our knowledge no direct comparison between PCV10 and PCV13 for all serotypes has been published. In this study we evaluate not only the quantity of the antibodies (the serotype-specific total IgG antibody concentrations), but also the quality or functional activity of the antibodies. This was evaluated through analysis of avidity, i.e., the strength of antibody-antigen binding interactions, and opsonophagocytosis, i.e., the ability to opsonize bacteria to facilitate phagocytosis.

The current study compared, head-to-head, PCV10 and PCV13 administered in the primary series (at 2, 3 and 4 months) and as a booster dose at 11 months of age. Here we report the immunogenicity data, including pneumococcal serotype-specific IgG antibody concentrations and avidity in serum samples, and opsonophagocytosis responses in plasma samples, obtained pre-booster and post-booster at one week and one month. In addition to pneumococcal data, serum IgG concentrations against diphtheria, tetanus, pertussis and H. influenzae type b were compared. Our primary objective was to compare immunogenicity at one month post-booster induced by immunizations with PCV10 or PCV13. Secondary objectives were: 1) to compare immunogenicity pre-booster and one week post-booster induced by immunizations with PCV10 or PCV13, 2) to investigate pre-booster and post-booster the possible influence of PCV10 or PCV13 immunizations on the immune responses induced by diphtheria, tetanus, pertussis and H. influenzae type b.

Material and Methods

Study design

A single-centre, parallel-group intervention study with two groups (PCV10 recipients and PCV13 recipients) was conducted in the Netherlands among infants eligible for the routine National Immunization Program (NIP): vaccinations at 2, 3, 4 and 11 months of age, or the 3+1 (primary plus booster) schedule (S1 Text). Through their parents, infants born in August 2011 were invited for the PCV13 group, and infants born in September-November 2011 were invited for the PCV10 group (Fig 1). Exclusion criteria were a serious and/or immunological disease that could interfere with the results of the study; previous administration of plasma products; presence of bleeding disorders, and prematurity (gestational age <37 weeks). No financial compensation for participants was provided. Infants in the PCV10 group received Synflorix (GlaxoSmithKline Biologicals s.a., Rixensart, Belgium) during regular well-baby clinic visits according to the NIP schedule. Infants in the PCV13 group received Prevenar-13 (Pfizer Limited, Sandwich, Hampshire, UK) during home visits by the study team according to the NIP schedule. Both groups received concomitantly DTaP-IPV-Hib-HepB (Infanrix hexa, GlaxoSmithKline Biologicals s.a., Rixensart, Belgium).

Fig 1

Enrollment and Study Participation.

PCV10 indicates 10-valent pneumococcal conjugate vaccine, Synflorix; PCV13 indicates 13-valent pneumococcal conjugate vaccine, Prevenar-13. * When the enrolment target was reached, further children were considered redundant, and their parents’ informed consent procedures were cancelled. ** Plasma cell and memory B-cell analysis published by van Westen et al.[27].

For B-cell and immunogenicity assays, an 8-ml blood sample was collected by venipuncture and randomly allocated “pre-booster” for half of the children (group A) and “one week post-booster” for the other half (group B). From children of the PCV10 group, who gave the 8-ml post-booster sample, an additional pre-booster sample (300 µl) was collected by heel stick. In all subjects, to collect the one month post-booster blood (300 µl), which was the primary endpoint, the heel stick was used because it was easier for the children. This was an open-label study for parents and study staff, but immunogenicity analysis was performed blinded. An acknowledged national ethics committee approved the study (METC Noord Holland, NTR3069). All study-related activities were performed in accordance with European guidelines for Good Clinical Practice, which includes provisions of the Declaration of Helsinki, with written informed consent from the child’s parents or the guardian.

In the current article, we publish the quantitative IgG data for all 132 subjects, of all three time-points (pre-booster, one week post-booster, and one month post-booster) and all 13 serotypes. In addition, we publish the qualitative data, i.e., avidity and opsonophagocytosis of pneumococcal serotypes and IgG data of diphtheria, tetanus, pertussis and H. influenzae type b. Results on the B-cell assays were previously published, open access, by van Westen et al. [27] who also included IgG data for a subset of the children for 6 of the 13 serotypes, pre-booster and one week post-booster.

Humoral analyses

After blood collection, sera for serological analyses and plasma for opsonophagocytosis assays were separated and stored at -80°C. Pneumococcal serotype-specific IgG antibody concentrations against the 13 pneumococcal serotypes were measured using a fluorescent bead-based multiplex immunoassay (MIA), as described previously [28]. Avidity for all 13 serotypes was determined with a MIA using 0.5 M ammonium isothiocyanate (NH4SCN) compared with phosphate buffered saline (PBS) [29]. Opsonophagocytosis activity (OPA) was determined for all 13 serotypes by multiplex opsonophagocytosis assay performed at UCL Institute of Child Health (London, UK) [30]. Antibody concentrations against DTaP-Hib antigens were determined by MIA, as described previously [31-33].

Statistical analyses

In this type of study, a 2- to 2.5-fold difference is considered to represent a true difference in immunogenicity [20, 34]. A minimum of 47 children per group were needed in order to provide 80% power to detect a 2-fold difference in IgG geometric mean concentrations (GMCs) between the two groups, with an assumed variance of the log-transformed concentration of 0.27 (based on serotype 6B, the serotype with the highest variance). The following formula was used n = 2k2σ2 / δ2. Since 66 children per group had already participated in the study for the B-cell assays [27], a 1.8-fold difference could be detected at one month post-booster the primary endpoint. For the other time-points, collected blood samples were fewer but sufficient to detect a 2- to 2.5-fold difference. Baseline characteristics including sex, age at vaccination and blood sampling, day-care attendance at 12 months of age, birth-weight, gestational age, presence of siblings < 5 year of age, duration of breast-feeding, and passive smoking were compared between the two groups, using a t-test for continuous variables and a chi-square test for categorical variables.

For each group, pneumococcal serotype-specific IgG GMCs, mean avidity indices (MAI), and OPA geometric mean titers (GMTs) with 95% confidence intervals (CI) were calculated. Percentages of vaccine recipients reaching an IgG threshold of 0.35 μg/ml for all time-points and 1.0 μg/ml post-booster were calculated. IgG GMCs were calculated diphtheria, tetanus, pertussis and H. influenzae type b antigens, and their internationally accepted thresholds for protection were used for seroprotection rate analysis. Differences in IgG GMCs, MIA, and OPA GMTs between groups were analyzed with a t-test, and differences in the percentage of subjects reaching a threshold were tested with a chi-square test or, when appropriate, a Fisher’s Exact test. Spearman correlations between IgG and OPA or avidity and OPA were calculated. For analyses between post-booster time-points (one week versus one month), a paired t-test was used.

Since the age at vaccination and blood sampling differed slightly between the groups, we performed secondary analyses in which we adjusted for age in days at vaccination and blood sampling using regression analysis.

All data were analyzed using SPSS 19. A p value of <0.05 was considered statistically significant.

Results

Study population

In total, 132 infants were enrolled in the study, both groups (PCV10 recipients and PCV13 recipients) consisting of 66 infants (Fig 1). Baseline characteristics of the two groups are summarized in Table 1. Subjects of the PCV10 group, immunized during regular well-baby clinic visits, were slightly older at vaccination and blood sampling than children of the PCV13 group, who were vaccinated at home by the study team. This difference reflects the normal delays that occur when families plan clinic visits. Compared to the standard analyses described below, regression analyses for all time-points, with adjustment for age at vaccination and blood sampling (S1–S3 Tables), showed no change in the pattern of differences between the PCV13 group and the PCV10 group, although some differences were no longer statistically significant.

Table 1

Characteristics of the participants.

	PCV13 (n = 66)	PCV10 (n = 66)	p-value
Male sex, %	62.1	50.0	0.161
Month of birth, median (min-max)	Sept (Sept-Sept)	Nov (Oct-Dec)	-
Day-care attendance a , %	81.5b	78.8	0.693
Half-days at day-care a , mean (sd)	4.5 (1.6)	4.7 (1.7)	0.568
Birth-weight in grams, mean (sd)	3578 (458)	3636 (471)	0.477
Gestational age in weeks, mean (sd)	39.6 (1.3)	39.8 (1.3)	0.579
Presence of siblings < 5 years, %	40.9	50.0	0.294
Breast-feeding, %	81.8	81.8	1.000
Duration of breast-feeding in months, mean (sd)	5.6 (3.0)	6.3 (3.7)	0.309
Passive smoking, %	0.0	1.5	0.315
Age at 2-month vaccination, mean (sd)	1.9 (0.12)	2.0 (0.17)	<0.001

a At 12 months of age

One month post-booster immunogenicity for the ten shared pneumococcal serotypes

One month post-booster, IgG GMCs were higher for serotypes 5, 6B, 9V, 14, 19F and 23F in the PCV13 group compared with the PCV10 group (Fig 2 and S1 Table). Avidity indices of IgG antibodies induced by PCV13 post-booster were higher for serotype 6B and 19F, but lower for 5 serotypes compared with the PCV10 group (Fig 2 and S2 Table). Nearly all participants reached the 0.35 μg/ml seroprotection threshold for all serotypes except for 19F in the PCV10 group (86%) (Table 2). The percentage of children with antibody concentrations >1.0 μg/ml was >90% for most serotypes, except in the PCV13 group for serotype 4 (89%) and in the PCV10 group for serotype 19F (73%).

Fig 2

Pneumococcal serotype-specific measurements.

IgG geometric mean concentrations (GMCs) pre-booster, one week post-booster and one month post-booster; Mean avidity indices of IgG antibodies pre-booster, one week post-booster, and one month post-booster; opsonophagocytosis geometric mean titers (GMTs) pre-booster and one week post-booster. *indicates significant differences (<0.05), higher in the PCV10 group than the PCV13 group #indicates significant differences (<0.05), higher in the PCV13 group than the PCV10 group **sample size for opsonophagocytosis assays: PCV10 group N = 22–27, PCV13 group N = 25–31

Table 2

Seroprotection rate of the antibody concentrations against 13 pneumococcal serotypes for the PCV10 group and the PCV13 group with p-values for differences between the groups.

	Pre-booster			One month post-booster
	PCV13 (N = 31)	PCV10 (N = 57)		PCV13 (N = 65)		PCV10 (N = 66)
Serotype	>0.35	>0.35	p-value >0.35	>0.35	>1.0	>0.35	>1.0	p-value* >0.35	p-value* >1.0
1	77	88	0.207	100	100	100	97	-	0.496
4	29	39	0.370	97	89	100	92	0.244	0.561
5	94	90	0.707*	100	99	100	100	-	0.496
6B	68	93	0.002	99	99	100	100	0.496	0.496
7F	100	97	0.538*	99	99	100	100	0.496	-
9V	77	97	0.008*	100	100	100	100	-	-
14	94	90	0.707*	100	100	100	96	-	0.244
18C	84	95	0.124*	100	99	100	100	-	0.496
19F	87	65	0.026	100	100	86	73	0.003	<0.001
23F	55	83	0.005	99	99	100	94	0.496	0.365
3	61	30	0.004	100	95	55	23	<0.001	<0.001
6A	81	19	<0.001	100	100	76	44	<0.001	<0.001
19A	71	42	0.010	100	92	76	32	<0.001	<0.001

* Fisher Exact test

Pre-booster immunogenicity for the ten shared pneumococcal serotypes

Compared with the PCV10 group, pre-booster IgG GMCs were lower in the PCV13 group for 5 of the 10 shared serotypes but higher for serotype 19F (Fig 2 and S1 Table). Avidity indices of the PCV13 group were lower for serotypes 5 and 14 but higher for 19F, compared with the PCV10 group (Fig 2). Functional OPA GMTs in the PCV13 group were lower for serotype 6B and 19F (Fig 2 and S3 Table). In contrast, the IgG GMC and avidity index for serotype 19F were higher for PCV13-vaccinated children compared with PCV10-vaccinated children. The percentage of children with antibody concentrations >0.35 μg/ml was above 75% for most serotypes (Table 2).

One week post-booster immunogenicity for the ten shared pneumococcal serotypes

Surprisingly, all one week post-booster IgG GMCs were significantly higher (p<0.002) than one month post-booster IgG GMCs (p values not shown), except for serotype 19F in the PCV10 group. Differences in IgG GMC and avidity data between the PCV13 and PCV10 groups were similar at one week and one month. However, in the one week post-booster groups, which included only half of the children of the PCV10 and PCV13 groups, the differences were in most cases not significant (Fig 2 and S1 and S2 Tables). OPA GMTs were significantly higher in the PCV13 group for serotypes 9V and 23F (Fig 2 and S3 Table). Spearman correlations were serotype-specific and found only found between IgG GMC and OPA; no correlation between avidity and OPA or IgG was found except for PCV10 serotype 1 (S5 Table and S1 Fig).

PCV13-specific serotypes 3, 6A and 19A

PCV13 induced IgG GMCs for serotype 3, 6A and 19A with pre-booster seroprotection rates ranging from 61% to 81% whereas the rates one month post-booster were 100% for antibody concentrations >0.35 μg/ml and 92–100% for antibody concentrations >1.0 μg/ml (Table 2 and S1 Table). PCV13 avidity indices these serotypes were comparable to other serotypes (Fig 2 and S2 Table), for all three time-points. In the PCV10-group IgG GMCs were slightly increased at one week and one month post-booster, compared with pre-booster, for serotypes 3, 6A and 19A (Fig 2, S1 Table), but avidity was not measured. One week post-booster in the PCV10 group, OPA GMTs had increased for serotypes 6A and 19A but were much lower than for the PCV13 group (S3 Table).

One month post-booster immune response of DTaP-Hib

Diphtheria- and pertussis-specific IgG GMCs and seroprotection rates induced by the concomitantly given DTaP-Hib vaccine did not differ between the groups (Table 3 and S4 Table). In both groups, all vaccine components showed a good booster response. However, Tetanus-specific IgG GMCs were significantly lower in the PCV13 group than the PCV10 group, whereas seroprotection rates were similar. Hib IgG GMCs were only significantly lower in the PCV13 group compared with the PCV10 group in the age-adjusted analyses.

Table 3

Seroprotection rate (%) for diphtheria, tetanus, pertussis, and Hib antigens for the PCV10 group and the PCV13 group with p-values for differences between the groups.

	Pre-booster			One month post-booster
Antigen	PCV13 (N = 31)	PCV10 (N = 56)	p-value	PCV13 (N = 65)	PCV10 (N = 66)	p-value
Diphtheria > = 0.10 IU/ml	68	63	0.625	100	100	1.000
Pertussis PT > = 20 EU/ml	39	39	0.958	99	100	0.496*
Tetanus > = 0.10 IU/ml	90	100	0.042*	100	100	1.000
Hib > = 0.15 mcg/ml	21**	68	<0.001*	97	100***	0.496*

*Fisher Exact test

**N = 29 for PCV13

***N = 65 for PCV10

Discussion

The current study compared immunogenicity induced by PCV13 and PCV10 given at 2, 3, 4 and 11 months of age in a Dutch cohort of 132 infants. In addition to the quantitative antibody levels (IgG), also the quality of these IgG antibodies was examined by measuring avidity and functional opsonophagocytic activity (OPA) for the pneumococcal vaccine serotypes before and after the booster vaccination.

Post-booster data

Post-booster, the IgG GMC levels for six shared serotypes were higher for the PCV13 group than for the PCV10 group. The PCV13 group also had higher functional OPA responses for two serotypes but higher avidity levels only for one serotype. The age at vaccination and blood s<span class="Chemical">ampling differed slightly between the groups, and the vaccination schedule varied more in the PCV10 group. However, the age differences between the groups did not appear to affect immune responses, as the age-adjusted analysis showed similar post-booster data for both products. Likewise, the results of the study of Spijkerman et al. showed that four different primary PCV13 vaccination schedules resulted in similar one month post-booster immunogenicity outcomes [29]. Differences found between our two groups could therefore be due to actual differences between the two PCVs, such as serotype concentrations, conjugates, and additives.

The most pronounced difference in immunological response between the two vaccines was found for serotype 19F, which showed a significantly higher IgG GMC and avidity index in the PCV13 group compared with PCV10 group. In our study, however, the serotype 19F-specific IgG GMC (2.09 μg/ml) induced by PCV10 one month post-booster was much lower than found in previous studies (4.23 to 9.72 μg/ml) [20, 21, 35–39]. Also, the related seroprotection rate (>0.35 μg/ml), which was 86% for serotype 19F for our PCV10 group, is lower than the 97–100% reported in other studies [20, 21, 35–39]. Seroprotection rates are based on internationally accepted ELISA IgG GMCs cut-off levels of >0.35 μg/ml for all serotypes [40]. Recently, Andrews et al. proposed serotype-specific ELISA thresholds for protection based on effectiveness data for PCV13 [41]. For serotype 19F, they determined a threshold of 1.17 μg/ml, which would translate to a serotype 19F-specific seroprotection rate of 70% for our PCV10 group and 100% for the PCV13 group one month post-booster. Moreover, our seroprotection rate was calculated using IgG GMCs determined in a MIA, which in general results in slightly higher levels than the WHO ELISA [28]. Serotype 19F IgG GMCs for our PCV13 group, analyzed in same assay run as for PCV10, indeed were relatively high (24.37 μg/ml), especially compared with data generated using the WHO ELISA. This finding indicates that the already low values for the PCV10 group might have been even lower if the WHO ELISA had been used. In spite of low serotype 19F IgG GMCs and avidity for the PCV10 group, OPA responses and OPA seroprotection rates were comparable to those in the PCV13 group, They were 88.9% and 100% for the PCV10 group and the PCV13 group, respectively, when using the 19F-specific seroprotection cut-off level of 430 [41]. This suggests a low correlation between IgG and OPA, which was indeed only 0.39 in the PCV10 group and 0.49 in the PCV13 group. The relatively high serotype 19F OPA response in the PCV10 group, despite the low IgG results could be explained by a higher IgM antibody response, since IgM is be effective at opsonizing bacteria [42, 43]. Moreover, the timing of the OPA samples in this study should be considered as we measured at one week post-booster, which could reflect the recall of the memory response, whereas if we would have measured at one month post-booster the time-point normally used for IgG antibodies, our measurements could have reflected a mixture of memory and newly generated antibodies.

Protection against disease by PCV10 seems adequate, SP carriage has remained low (1%) since its 2009 introduction in the Netherlands and no PCV10 serotype 19F vaccine failures have been observed [9, 44] despite low serotype 19F IgG GMCs and avidity.

Studies in Finland and Brazil have likewise found PCV10 to be 84–100% effective against IPD [45, 46].

For PCV13 serotype 19F, the current study showed an IgG GMC of 24.37 μg/ml, which is higher than in previous studies of both PCV7 (2.4–5.6 μg/ml)[20, 22, 23, 38, 47] and PCV13studies (6.6–11.06 μg/ml)[22, 23, 47] and could be related to the use of MIA instead of ELISA [28]. However, Grant et al. and Dagan et al. [22, 23], using ELISA, also reported higher 19F-specific responses for PCV13 than seen in previous PCV7 studies. Cross-reaction by serotype 19A could play a role [22, 48–50]. Overall, the reported higher serotype 19F-specific IgG GMCs, avidity, and OPA induced in our PCV13 group could result in better protection against pneumococcal disease than induced by PCV7, for which six vaccine failures in 5 years occurred in the Dutch population [44].

Van Westen et al. published the plasma and memory B-cells results of described trial for 6 serotypes, showing no clear correlation between IgG and plasma or memory B-cells. In our paper, we show a weak correlation between IgG GMCs and OPA and no correlation between avidity and OPA. Together with the fact that OPA measures actual killing of the bacteria and avidity index is a measure of antigen-antibody binding strength implies that for comparison of differences in vaccine-induced immunogenicity, IgG GMCs combined with OPA analysis should be performed.

Pre-booster data

Pre-booster seroprotection rates for pneumococcal vaccine serotypes remained high for both our groups, indicating that protection of children is adequate during the whole period between the primary series and the booster dose. Whether the 5 serotypes with lower IgG GMCs pre-booster in the PCV13 group, compared with PCV10, reflect a lower initial response after the primary series or a steeper decline cannot be concluded from data collected in this trial.

Interestingly, Spijkerman et al. showed that pre-booster antibody levels at 11 months of age were higher when the primary vaccinations started at an older age (three versus two months) or had longer intervals between doses (two versus one month) [29]. Children in our PCV13 group followed the NIP schedule for the primary vaccinations more strictly than the children in the PCV10 group, who in general had somewhat longer intervals between the doses. This might explain the differences between the groups in pre-booster GMCs at 11 months. However, the differences in the intervals are relatively small (at most 0.4 months at the third vaccination), and a more plausible explanation for the results is therefore the difference in vaccine composition between PCV13 and PCV10. Despite somewhat lower IgG GMCs for the PCV13 group, compared with the PCV10 group, PCV13 is still expected to be effective in the period between primary and booster vaccination. Its effectiveness has been shown in countries where it has been implemented, and even with herd protection induced by PCV7, a further decline of IPD caused by PCV7 serotypes is seen after introduction of PCV13 [51-53]. Also in other studies, comparable seroprotection rates for PCV13 at one month after the primary series have been shown [26, 29, 41].

Added value of serotype 3, 6A and 19A

Although serotypes 3, 6A and 19A are not present in PCV10, IgG GMCs against these serotypes could be detected in the PCV10 group, pre- as well as post-booster dose. For serotype 6A and 19A, a small but significant increase in IgG GMCs was found after the booster dose along with functional post-booster OPA (S1 Table). Both findings could be a result of carriage of 6A and 19A serotypes by 9–15% of the Dutch children at 11 months [4, 5, 9] or cross-reactivity between serotype 6B and 6A and between 19F and 19A antibodies [22, 48–50]. However, post-PCV10-booster, IgG GMCs for these serotypes were 10-30-fold lower than the IgG GMCs reached in PCV13-vaccinated children, and only 9 of 27 PCV10-vaccinated children showed serotype 19A-specific OPA titers comparable to the OPA titers measured in PCV13-vaccinated children, all others being lower. One week post-booster, 89% of the children of the PCV10 group and 100% of the PCV13 group reached the serotype 19A-specific OPA seroprotection threshold of 48 calculated by Andrews et al. as protective against serotype 19A [41]. In the study published by van Westen et al., the PCV10 group had no post-booster increase in plasma and memory B-cells for serotype 6A and 19A, possibly because the cross-reactive vaccine serotypes did not generate an optimal memory response for the non-vaccine serotypes [27]. Whether PCV10 induces protection against serotype 19A disease remains to be seen.

Influence on other NIP components

Serum IgG GMCs against the concomitantly given tetanus and H. influenzae type b (Hib) vaccine components were significantly lower in the PCV13 group compared with the PCV10 group. Enhancement of the antibody responses to Hib, diphtheria and tetanus vaccines by PCV10 vaccination have been reported [54] and could be related to the conjugate proteins used for this vaccine. The clinical relevance of these differences is unknown.

Strength and limitations of this study

The strength of our study is the head-to-head comparison within one trial of immunogenicity induced by PCV10 and PCV13, with all samples analyzed in the same lab using the same methods. An additional strength is that post-booster responses were evaluated at two short-term time-points (one week and one month after the booster), and multiple immunogenicity parameters, including functional activity, were evaluated (IgG GMC, avidity and opsonophagocytosis), generating a more complete immunological profile. We could therefore also determine if IgG or avidity is a better predictor of vaccine effectiveness. A study limitation is the small difference in age at vaccination and blood sampling between the two groups, due to different vaccination practices (well-baby clinic visits for the PCV10 group and home visits for the PCV13 group); however, the age-adjusted analyses showed no difference in immunogenicity.

Conclusion

Both PCV10 and PCV13 are immunogenic and able to induce a good booster response. In the PCV13 group compared to the PCV10 group, six shared serotypes showed higher IgG GMCs, and two higher avidity at one month post-booster, whereas pre-booster, five serotypes showed lower IgG GMCs. Pneumococcal serotype 19F-specific immune responses induced by PCV13 were higher than PCV10-induced responses, pre- as well as post-booster. Beyond the quantitative serum IgG GMC measurements, assessment of the qualitative activity of generated antibodies by avidity and OPA analysis showed the added value, particularly for serotype 19F. In general, IgG GMCs correlated better with opsonophagocytosis than avidity with opsonophagocytosis. Further investigation is required to determine the clinical relevance of the differences between PCV10- and PCV13-induced responses and their impact on SP carriage and long-term protection.

Approved protocol.

(PDF)

Click here for additional data file.

Trend statement checklist.

(PDF)

Click here for additional data file.

Geometric mean concentrations (GMC) with 95% CI of the antibody concentrations against 13 pneumococcal serotypes for the PCV10 group and the PCV13 group.

(PDF)

Click here for additional data file.

Mean avidity index with 95% CI for 13 pneumococcal serotypes for the PCV10 group and the PCV13 group.

(PDF)

Click here for additional data file.

Geometric mean titers (GMT) with 95% CI of opsonophagocytosis for 13 pneumococcal serotypes for the PCV10 group and the PCV13 group.

(PDF)

Click here for additional data file.

Geometric mean concentrations (GMC) with 95% CI of the antibody concentrations against diphtheria, tetanus, pertussis and Hib antigens for the PCV10 group and the PCV13 group.

(PDF)

Click here for additional data file.

Spearman correlations one week post-booster between IgG, OPA titer and avidity for the PCV10 group and the PCV13 group.

(PDF)

Click here for additional data file.

(PDF)

Click here for additional data file.