Pneumococcal Immune Response in Infants Whose Mothers Received Tetanus, Diphtheria and Acellular Pertussis Vaccination During Pregnancy

Abstract

Background: 

Maternal immunization with a tetanus, diphtheria and acellular pertussis (Tdap) vaccine may blunt infant pneumococcal immune responses after a primary series of vaccines.

Methods: 

As part of a prospective controlled cohort trial of Tdap (Boostrix; GSK Biologicals, Rixensart, Belgium) vaccination in pregnancy, infants born to vaccinated mothers and controls were immunized at 8 and 16 weeks and 12 months of age with 13-valent pneumococcal conjugate vaccine (Prevenar13; Pfizer, Wyeth, United States). Sera were tested for pneumococcal antibody concentrations against vaccine serotypes following primary and booster immunization.

Results: 

Geometric mean concentration of antibodies to serotypes 1, 3, 4, 5, 6A, 7F, 9V, 14 and 19A was significantly lower after 2 doses of Prevenar13 vaccine in the offspring of the mothers vaccinated in pregnancy. This blunting effect disappeared after a booster dose at 12 months of age, except for serotypes 1 and 4. Despite this blunting, the percentage of children achieving the threshold of protection of 0.35 µg/mL was comparable in the vaccine and the control group both after primary and booster vaccination with only a significant lower rate of seroprotection in the vaccine group for serotype 3 after primary vaccination. After booster vaccination, seroprotection rates increased further for serotypes 3, 5, 6B, 9V and 23F.

Conclusions: 

The present results indicate a blunting effect after primary vaccination for some serotypes resolving after booster vaccination. Seroprotection rates were comparable both after primary and booster vaccination, except for serotype 3 with a significant lower seroprotection rate in the vaccine group after primary vaccination.

Pertussis, caused by the Gram-negative bacteria Bordetella pertussis, is a global endemic respiratory disease and an important cause of morbidity and mortality among infants.¹ Worldwide, pertussis is estimated to cause 63,000 deaths in less than 5-year-old children (2013), although there is considerable uncertainty around these estimates because of paucity of reliable surveillance data, particularly from developing countries.¹ In industrialized countries, outbreaks of pertussis have been reported during the last decade, and the overall incidence is increasing, with most severe cases and fatalities among young, not fully vaccinated infants.²

An important way to offer protection against pertussis disease from birth is, with the currently available vaccines and vaccination schedules, to include immunization during pregnancy. High concentrations of maternal antibodies, elicited by maternal vaccination, are transported transplacentally to the fetus, offering protection until the start of the primary infant vaccination schedule and thereby closing the susceptibility gap for infection.^3–7 Several industrialized countries have already put in place a recommendation for this strategy although some safety and immunologic aspects were unknown at implementation. While blunting of infant immune responses to pertussis antigens included in the infant vaccination has been described after a primary series of infant vaccination in the presence of maternal antibodies,^4,7 the impact of maternal Tdap vaccination on other non-Tdap antigens, included in the regular infant vaccination schedule, is a potential concern.

A UK study⁷ showed a blunting effect on some serotypes of the primary pneumococcal infant humoral immune response, in infants born to mothers immunized with a Tdap vaccine (Repevax; Sanofi Pasteur, Lyon, France) during pregnancy. The proposed explanation is that the diphtheria vaccination during pregnancy interferes with the immune response of the infants to the CRM-197 carrier protein included in the infant pneumococcal vaccine. Any additional data from controlled clinical trials on the beneficial and possible unknown side effects of maternal Tdap vaccination are therefore welcomed to assist decision-making with regard to maternal immunization as a global strategy to protect young infants.

We report the effect of maternal Tdap vaccination (Boostrix; GSK Biologicals, Rixensart, Belgium) on infant responses to pneumococcal conjugate vaccine 1 month after 2-dose pneumococcal priming (at 8 and 16 weeks of age) and 2.5 months after the pneumococcal booster dose (at 12 months of age). Importantly, the last booster dose is not foreseen in the global Expanded Program on Immunization schedule. This output adds to the body of knowledge on potential blunting of infant immune responses to childhood vaccination following maternal Tdap vaccination. The study will ultimately contribute to inform decision-making bodies on implementing maternal immunization, both in industrialized and low- and middle-income countries (LMIC).

MATERIALS AND METHODS

A prospective controlled cohort study was conducted in Belgium in 2011–2015, in accordance with the Declaration of Helsinki, International Conference on Harmonization - Good Clinical Practice (ICH-GCP), and procedures established by Belgian law (clinicaltrials.gov identifier: NCT01698346). The present analysis was a post hoc study on leftover blood samples from a trial on maternal Tdap vaccination^4,8 and was approved by the ethics committee of the University of Antwerp, Belgium. Written informed consent was obtained from both parents of the participating infants, and details on the study procedures can be consulted in previous publications.^4,8

Participating women were included in either a vaccine group: women vaccinated with Tdap (Boostrix) between 18 and 34 weeks of gestation (as per protocol), or a control group of pregnant women not vaccinated with a pertussis containing vaccine for at least 10 years. The offspring was included in 2 groups according to the vaccination status of their mother. Within the regular health care system, infant pneumococcal vaccines were administered at well-baby clinics, by a pediatrician or general practitioner at 8 and 16 weeks and 12 months of age. All infants were vaccinated according to the standard Belgian vaccination schedule with hexavalent vaccine, measles, mumps and rubella vaccine and rotavirus vaccine.

Study Vaccines

Licensed Tdap vaccine (Boostrix) was used to vaccinate pregnant women. Boostrix contains 5 Lf of tetanus toxoid, 2.5 Lf of diphtheria toxoid, 8 µg inactivated pertussis toxin, 8 µg filamentous hemagglutinin and 2.5 µg pertactin.

Infants were vaccinated with the 13-valent pneumococcal vaccine Prevenar13 (Pfizer, United Kingdom), containing 2.2 µg of serotypes 1, 3, 4, 5, 6A, 7F, 9V, 14, 18C, 19A, 19F and 23F and 4.4 µg of serotype 6B.

Study Procedures

Blood samples were collected from the infants at month 5 (28–35 days after the second pneumococcal vaccine dose) and at month 15 (2.5 months after the pneumococcal booster dose). Blood samples were centrifuged at 2000 rpm within 24 hours after blood collection and stored at –20°C. Because this study is a spin-off study from a maternal Tdap vaccination trial measuring infant pertussis immune responses, bleeding time points were originally not chosen to measure pneumococcal immune responses resulting in late bleeding after the pneumococcal booster dose.

Laboratory

Sera were tested for antibodies to the 13 vaccine-type capsular polysaccharides (1, 3, 4, 5, 6A, 6B, 7F, 9V, 14, 18C, 19A, 19F and 23F) at the University College London, Institute of Child Health. Immunoglobulin G (IgG) antibody levels were measured by the World Health Organization reference enzyme-linked immunosorbent assay after adsorption with cell wall and 22F polysaccharide.⁹ The lower limit of quantification has been set at 0.15 μg/mL. The protective threshold for all 13 pneumococcal serotypes is set at 0.35 μg/mL.¹⁰

Laboratory procedures used to test the samples were exactly the same as in the study by Ladhani et al⁷ which was also analyzed in the World Health Organization Reference Laboratory. Results of both studies are therefore comparable.

Statistics

Statistical tests included parametric tests: (paired) t tests and χ² tests, and nonparametric alternatives: (paired) Wilcoxon tests and Fisher exact tests whenever the underlying normality and sparseness assumptions of parametric tests were violated. Linear regression models were used to identify characteristics that could potentially impact infant antibody concentrations. The analysis was performed using SPSS statistical software 23.0 (IMB corporation). Two-sided P value of <0.05 was considered as statistical significant. Pneumococcal antibody concentrations below the lower limit of quantification were replaced by a value of 0.075 μg/mL, which is half of the limit of detection, to perform the analysis.

Blunting of the infant immune response is defined as a statistically significant lower geometric mean concentration (GMC) of serotype-specific antibodies in infants from 1 group compared with the other group.

RESULTS

General Characteristics of the Study Population

All infants were vaccinated with Prevenar13 at the respective immunization visits. Fifty-two mother–infant pairs were included in the vaccine group and 25 pairs in the control group. Women in the vaccine group were vaccinated with a Tdap vaccine (Boostrix) between 22 and 32 weeks of gestation. Women in our study were not vaccinated with a pneumococcal vaccine during childhood or later in life. However, they could have pre-existing antibodies because of natural exposure. Children were born between April 2, 2012, and April 16, 2014. Blood samples were taken between August 27, 2012, and July 27, 2015. The mean interval between Tdap vaccination and delivery was 77.7 days. The mean gestational age at vaccination was 28.8 weeks. Table 1 summarizes the characteristics of all mother–infant pairs, vaccination data and intervals between vaccination and blood sampling in the infants. All women in the control group had additional education after secondary school resulting in a significant difference in education (P = 0.008) between both groups. Significantly lower mean age at hexavalent vaccine dose 1 (P = 0.026) and dose 2 (P = 0.015) was calculated in the vaccine compared with the control group (Table 1).

TABLE 1.: Demographic Characteristics of Mother–Infant Pairs and Vaccination and Blood Sampling Data From Infants in Both Study Groups

Seroprotection Results

Table 2 provides an overview of the percentage of infants with an antibody concentration above the correlate of protection (0.35 µg/mL) 1 month after the second and 2.5 months after the third Prevenar13 dose. After priming, a significantly lower seroprotection rate was seen in the vaccine group compared with the control group for serotype 3 (P = 0.045). Low levels of seroprotection are described in both study groups for serotypes 6B and 23F after priming, but there is a significant rise after the third vaccine dose for both serotypes. After the booster dose, comparable seroprotection rates are found in both study groups for all serotypes (Table 2).

TABLE 2.: Percentage of Children With Pneumococcal Antibody Concentration Above the Correlate of Protection (0.35 μg/mL) with 95% CI in Both Study Groups 1 Month After the Second Pneumococcal Vaccine Dose and 2.5 Months After the Third Pneumococcal Vaccine Dose

Overall, following primary vaccination, the vaccine was immunogenic in both groups with similar proportions achieving protective concentrations. For most infants in the vaccine group, immune responses after the booster vaccination were higher than immune responses after primary immunization despite the fact that postprimary immune responses were measured 1 month, but booster responses 2.5 months after vaccination.

Serotype-specific IgG Concentrations

Table 3 provides an overview of the GMCs per serotype in both groups at all time points. One month after the priming, significantly lower GMCs are seen in the vaccine group for serotypes 1, 3, 4, 5, 6A, 7F, 9V, 14 and 19A. For serotypes 6B, 18C, 19F and 23F, comparable but lower antibody concentrations were found in the vaccine group. After the administration of the third pneumococcal vaccine dose, significantly lower GMCs were only seen in the vaccine group for serotypes 1 and 4, with a slight increase in antibody concentration after the booster for serotype 1. In general, the increase in antibody concentration between both time points is higher in the vaccine group, except for serotypes 4, 6B, 18C and 23F. Figure 1 shows the reverse distribution curve of the data for serotype 1, as an example. Reverse distribution curves for other serotypes can be found in Supplemental Digital Content 1, http://links.lww.com/INF/C704 (Table 3 and Figure 1).

TABLE 3.: Geometric Mean Concentrations With 95% CI for Antibodies to Serotype 1, 3, 4, 6A, 6B, 7F, 9V, 14, 18C, 19A, 19F and 23F Postprimary and Postbooster Vaccination in Both Groups of Infants

FIGURE 1.: Reverse cumulative distribution of serotype 1, as an example.

Results From the Regression Analysis

We only report significant influences of variables on the serotype-specific pneumococcal antibody titers after primary and booster vaccinations. Some variables like gender, lactation, gestational age at delivery and weight randomly influence serotype-specific antibody concentrations at one time point. There is no consistency in these effects, and they are hard to explain (Table 4). Other factors do seem to influence more consistently serotype-specific antibody responses. For example, a higher age at blood sampling correlates with a higher antibody concentration in the vaccine group. In contrary, a higher age at blood sampling correlates with a lower antibody concentration in the control group. A higher age at pneumococcal booster vaccination correlates with higher antibody concentration in both study groups, and a higher interval between pneumococcal vaccine dose 3 and blood sampling correlates with a lower antibody concentration in both study groups (Table 4).

TABLE 4.: Influencing Factors on Distinct Serotypes in the Multiple Regression Analysis

DISCUSSION

This study of pneumococcal vaccine responses in infants born to mothers who received Tdap vaccine (Boostrix) during pregnancy shows that while infant responses to Pneumococcal Vaccine (PCV) are blunted, proportions of infants achieving protective concentrations of serotype-specific IgG are similar irrespective of maternal vaccination status. Our study is the first study that has addressed the impact of Tdap in pregnancy on the responses to a pneumococcal booster dose. All serotype-specific concentrations reached a high percentage of seroprotection after the booster dose, with lowest rate for serotypes 4 and 23F (87.78%), but no significant differences in seroprotection rates between both study groups. This finding is reassuring that the blunting effect is temporary and time limited.

These results support those previously published for British mothers and their infants by Ladhani et al.⁷ Our findings are important as it confirms that the impact of maternal Tdap vaccination on pneumococcal humoral responses is not an isolated result in the United Kingdom and needs to be considered whenever recommending the maternal pertussis vaccination strategy with the available combination vaccines. In comparison with the UK data, we confirm low levels of seroprotection for serotypes 3 (significantly lower in the vaccine group), 6B and 23F but otherwise good levels of seroprotection (>65%) in both study groups for most other serotypes after primary vaccination. In contrast to the UK data, we did not confirm lower protection for serotypes 5 and 9V. Reaching the seroprotection level of antibodies ensures closing the susceptibility gap of infection for invasive pneumococcal disease (IPD).

Belgian and British populations are expected to be quite similar in characteristics, and both countries have a vaccination program with high coverages. Circulation of vaccine serotypes is therefore limited, and the protective relevance of blunting needs to be interpreted carefully. This is certainly different in countries where other schedules and coverage of pneumococcal vaccination are reported, for example, LMIC, where no booster doses are foreseen or coverage of pneumococcal vaccination might be lower resulting in ongoing circulation of vaccine included serotypes.

When comparing serotype-specific IgG concentrations, we found a significant blunting effect for serotypes 1, 3, 4, 5, 6A, 7F, 9V, 14 and 19A after 2-dose primary immunization. In the United Kingdom, blunting of the immune response was described for the same serotypes apart from serotypes 14 and 19A. While our vaccinees and controls were recruited in the same study period, the United Kingdom used historical control data collected in a period where maternal Tdap immunization was not yet in place. The use of different vaccines for maternal Tdap vaccination (Repevax vs. Boostrix) and possible boosting by nasopharyngeal carriage of pneumococci in the historical controls may explain why responses to 14 and 19A were comparable between both study groups in the UK study, but blunted in the vaccine group of our study after 2-dose primary immunization.

In our study, the second blood sample was taken relatively late after the third pneumococcal vaccination (2.5 months), hence the concentrations measured at that point might already be waning. However, as the study was randomized, we assume that waning of concentrations is similar in both groups of infants, thus allowing comparison of the results between both study groups. Postbooster GMC results were comparable for both groups, and blunting only persisted for serotypes 1 and 4. Booster responses to serotypes 6A, 6B, 7F, 14, 19A and 19F were excellent. The high response to these specific serotypes has already been reported previously.^11,12

Interestingly, booster responses in the control group were lower for 7 serotypes 2.5 months after boosting compared with primary responses. This may be due to rapid decay following boosting and the delayed time point of blood sampling in the infants which may have missed the peak of the response. This phenomenon was only seen for 2 serotypes (4 and 18C) in the infants born to vaccinated mothers.

Comparison of GMC to postvaccination data after vaccination with distinct vaccination schedules is useful to interpret the present results. Spijkerman et al¹² compared several PCV13 vaccination schedules. Children receiving a 3 + 1 schedule reached higher levels of antibodies after the primary vaccination when compared with our study, yet all schedules had similar high seroprotection levels after completion. Nevertheless, we report similar antibody concentrations after a short primary schedule as do Rodgers et al¹³ in their comparative overview on different priming schedules with PCV13. Higher response rates are reported for serotype 4 but similar lower rates for serotype 23F.

Gestational age at maternal vaccination did not influence the antibody concentrations. Our cohort was not powered to detect differences in GMC when vaccinating at different gestational ages, as has been recently suggested.^14,15 Nevertheless, some variables like age at blood sampling, age at pneumococcal booster vaccination and interval between pneumococcal vaccine dose 3 and blood sampling at month 15 do seem to have influence on several serotype antibody responses. However, we have to be careful drawing conclusions on influencing factors taking into account the relatively small sample size of the study.

In Belgium, the circulating strains causing IPD in children <2 years of age, the period where children are most vulnerable to IPD, were 3, 7F, 19A and non PCV13 types in 2014.¹⁶ In the same year, 186 IPDs were diagnosed in children below 16 years of age (N = 283 in 2013 and N = 334 in 2012). The overall incidence of IPD <2 years of age was 53.2/10,000 (compared with 156/10,000 before the start of the vaccination recommendation). Coverage for the third dose of pneumococcal vaccine among infants in Flanders is >96.5%. The present results show interference, resolving after the booster dose. Because there is low circulation of the vaccine serotypes, the blunting effect will probably not have clinical significance unless long-term blunting in the population might affect the circulation of vaccine strains again.

In view of global recommendations for maternal immunization to protect infants from disease, the effect of maternal Tdap vaccination on infant immune responses regarding both seroprotection rate and clinical efficacy is important. In particular when considering implementation of recommendations for maternal vaccination in LMIC with varied regional epidemiology and infant immunization schedules (3-dose priming without booster dose) as used in the Expanded Program on Immunization because no deduction from the Belgian study can be drawn and no data from LMIC are available to confirm these results.

The study has a few limitations. The present project builds on an existing serum bank of leftovers and is therefore a convenience sample, without having a power calculation on beforehand to answer this research question. In addition, the main research aim was initially not to identify blunting of pneumococcal immune responses, and the moments for blood sampling are therefore not adapted to the pneumococcal vaccination schedule. Maternal samples were not tested for pre-existing anti-pneumococcal antibody titers; hence, maternal antibody interference could not be analyzed, but as the mothers were randomized, this should not impact on the analysis. The interval between vaccination and blood sampling after the primary series (1 month) differs from the interval between the booster dose and blood sampling (2.5 months). Also no functional data were reported in this paper, because the sample volumes of the infants after multiple testing were limited, and we were therefore not able to undertake additional analysis of functional anti-pneumococcal antibody. However, the correlation between serotype-specific IgG measured by enzyme-linked immunosorbent assay compared with Opsonophagocytic Assay (OPA) is good following PCV in infancy especially after the booster dose.¹⁷ Despite this limitation, the samples were all taken at the same time point, allowing us to compare the results after booster vaccination between both groups, taking into consideration the possible waning of antibodies.

CONCLUSION

The blunting effect of maternal Tdap vaccination on pneumococcal immune responses in young infants is confirmed in the present study. The clinical effect on protection from pneumococcal disease will likely be low in the Belgian setting, because protective levels of antibody are achieved for almost all serotypes and circulation of vaccine serotypes is almost nonexistent.

ACKNOWLEDGMENTS

Bill & Melinda Gates foundation (BMGF) is acknowledged gratefully for the support of the present study. Aline Bontenakel is gratefully acknowledged for taking all blood samples from the participating children.