Plain Language Summary

Infants need protection against whooping cough particularly in the first few months after birth. Vaccination of women during pregnancy is recommended in some countries to protect the infant against the disease through the transfer of the mother’s protective antibodies. We reviewed the published evidence on the immunological and clinical benefits of maternal vaccination against whooping cough. Vaccination of pregnant women elicited robust maternal immune responses against whooping cough and resulted in high antibody levels transferred to the infant. Infants born to women who were vaccinated during pregnancy were better protected against whooping cough in the first few months after birth than those whose mothers were not vaccinated. The best time to vaccinate pregnant women was in the late second or early third trimesters. Although the infant’s immune responses to their primary immunizations were reduced in those born to women vaccinated during pregnancy, this did not affect vaccine effectiveness. This review confirms that maternal vaccination against whooping cough during pregnancy protects infants against the disease.

Introduction

Pertussis, whooping cough, results from infection of the upper respiratory tract by Bordetella pertussis. Infection resulting in at least mild disease was universal among children before the introduction of pertussis vaccination. Disease outbreak cycles occurred every 3–5 years, which primed most individuals through repeated exposure, and provided ongoing immunity after the first episode. Young infants were partially spared through indirect protection from maternal antibodies acquired in utero. Although routine pertussis vaccination of infants and young children reduced the incidence by more than 99% [1], the disease still persists today. In addition, there has been a shift in the age-specific disease profile with, in particular, an increased proportion of cases among infants aged < 1 year, adolescents and the elderly [2,3,4,5,6,7]. This age-specific shift has been attributed to waning of vaccine-induced immunity, and, in young infants, reduced pathogen circulation limiting natural boosting during pregnancy and subsequent passive protection. Infants (in particular those too young to have received the primary immunization series) have the highest rates of pertussis-related morbidity and mortality [6, 8,9,10,11,12,13].

Older siblings and adult close contacts are common reservoirs for pertussis transmission to young infants [14,15,16]. Strategies introduced to increase herd immunity have included booster doses for adolescents and close-contact adults. However, widening coverage with pertussis vaccines appears to have had little impact in preventing transmission to young infants or on the resurgence of periodic epidemic peaks in some countries in recent years [11, 13, 17]. Concerns that close contacts continue to remain common reservoirs for disease transmission to young infants have led to vaccination against pertussis during pregnancy [18,19,20,21,22,23]. Vaccination during pregnancy protects the mother directly, and, more importantly, provides passive protection (via transplacental antibody transfer in utero) for their young infants.

A number of clinical studies assessing the immunogenicity, efficacy and safety of maternal pertussis vaccination during pregnancy have recently been published. As such, it is timely to collate the available evidence and summarize the benefit–risk profile of pertussis vaccination during pregnancy. This will help informed decision-making and identify important gaps in the evidence that may require further investigation. This systematic review focuses on the immunological response to maternal pertussis vaccination during pregnancy as well as on vaccine efficacy in preventing infant pertussis. Maternal immunological responses, placental transfer of antibodies and infant responses to the primary immunization series are summarized. The safety of pertussis vaccination during pregnancy is the focus of an accompanying systematic review (D’Heilly et al., in preparation).

Methods

We conducted a systematic review of the English language literature, following the PRISMA framework [24], to identify studies reporting the immunogenicity and efficacy of pertussis vaccination (tetanus toxoid, reduced-dose diphtheria toxoid, and reduced-dose acellular pertussis [Tdap] vaccine and Tdap-inactivated polio vaccine [IPV]) during pregnancy. The review protocol was registered on PROSPERO (PROSPERO 2016:CRD42016038317) [25].

Our search initially encompassed the period from January 1995 to June 2016, and was later updated to capture relevant studies published between July 2016 and October 2018. Databases searched included PubMed/MEDLINE, EMBASE, Scopus (Elsevier), Cochrane Database of Systematic Reviews, ProQuest, and Science Direct. Reference lists of identified publications were also scanned. ProQuest thesis, Clinicaltrials.gov, Conferences and Congresses, and Trial Trove were searched for unpublished evidence. The search algorithms used were [1]: [(pertussis OR whooping cough) AND (vaccine OR Tdap OR immunization) AND (pregnancy OR pregnant OR pre-partum OR gestation OR maternal)], and [2]: [1] AND (immunogenicity OR immunity OR immune response OR antibody OR antibodies OR effectiveness OR efficacy).

Interventional and observational studies in humans were considered using the Patient Population, Intervention, Comparison, Outcomes, and Setting criteria [26]. The population was pregnant women and their offspring; the interventions were pertussis immunization during pregnancy and the infant series of vaccination; comparison groups were either no vaccination or standard-of-care vaccination, as well as pregnancy or no pregnancy; the outcomes were absolute and relative immunogenicity measures against pertussis vaccine antigens in maternal, umbilical cord, and/or infant blood, as well as infant pertussis antibody responses following each dose of the primary and toddler booster vaccination course, qualitative (clinical symptoms, severity) and quantitative [number, frequencies and relative (e.g., OR, RR)] disease outcome measures, including vaccine efficacy and/or vaccine effectiveness estimates. We excluded reviews, case reports, opinions and letters to editors, meta-analyses, modeling studies, and studies on vaccination program improvement, vaccine uptake increase or health economics, and vaccine acceptability and perception studies.

Selection of publications for inclusion was done in two steps, with two independent reviewers involved at each step. Titles and abstracts of retrieved publications were first screened for relevance. Publications documenting primary research (excluding modeling studies), and specifically dealing with outcomes of pre-partum pertussis vaccination, were retained for full-text review to determined compliance with inclusion and exclusion criteria. Relevant data from included studies were extracted using pre-structured MS Excel forms. Where possible, vaccine brand names are presented to help ascertain the number of pertussis components assessed. A meta-analysis of the immunologic or efficacy/effectiveness findings was not feasible because of heterogeneity in the study designs, vaccines used, measured outcomes, measurement/analysis methods, and background pertussis incidence rates.

This article is based on previously conducted studies and does not contain any studies with human participants or animals performed by any of the authors.

Results

A total of 293 articles were identified, of which 43 articles (summarized in Tables 1 and 2, respectively) were included in this review (Fig. 1).

Table 1 Characteristics and results of studies included in the immunogenicity systematic review
Full size table
Table 2 Characteristics and results of studies included in the effectiveness systematic review
Full size table
Fig. 1
figure 1

PRISMA diagram of results of search strategy

Full size image

Immune Indicators

Maternal immune responses to pertussis antigens in both the mother and child were considered in this review. However, extrapolation of persisting antibody geometric mean concentrations (GMCs) to protection against pertussis is not possible as there is no agreed immunological correlate of protection.

Maternal Immune Responses

Vaccination with Tdap during pregnancy induces a robust maternal immune response (Table 1) [27,28,29,30,31,32,33,34,35,36,37,38,39,40], irrespective of maternal body mass index [30]. Pertussis antibody levels measured ≥ 1 month after Tdap immunization, or at delivery, were substantially higher against all vaccine antigens [pertussis toxin (PT), filamentous hemagglutinin (FHA), pertactin, and fimbriae (FIM)], in the Tdap-vaccinated group versus Tdap unvaccinated control groups across studies (Table 1). The higher antibody levels in Tdap-vaccinated women persisted through to delivery, and for up to 12–15 months post-delivery in some follow-up studies [28, 32].

Antibody responses to Tdap vaccination during pregnancy appear similar to those in non-pregnant women and women immunized post-partum [36, 38, 41]. However, although antibody GMTs increased for all antigens post-vaccination in both pregnant and non-pregnant women in one study [42], those against PT and FHA were significantly higher in non-pregnant women. The difference in the latter study may have simply reflected that a higher portion of non-pregnant women had received prior Tdap vaccination (65% vs. 53%).

Infant Serological Outcomes—Transfer of Antibodies

Newborn infants of mothers who received Tdap during pregnancy consistently had higher anti-pertussis antibody levels (antibodies to PT, FHA, pertactin and FIM) versus newborns of non-vaccinated mothers (p ≤ 0.002 across antibodies and studies; Table 1) [27, 29, 32, 34, 35, 38, 40, 43,44,45,46,47]. Estimates of cord to maternal blood antibody ratios from studies assessing transplacental antibody transfer ranged from 1.03 to 3.47 across the four pertussis vaccine antigens [27, 29, 32, 33, 35, 38, 39, 41, 46, 48]. A linear positive correlation was found between maternal and newborn’s antibody levels [29, 34, 38, 44, 49]. This correlation may be important for defining the optimal vaccination window during pregnancy since timing may impact the level of antibodies in the mother pre-partum and consequently the level of antibodies transferred to the infant at birth.

The relationship between the timing of Tdap vaccination during pregnancy has been explored, both with respect to gestational week and interval before delivery, and anti-pertussis antibody GMCs in the newborn. A study involving 105 pregnant women found no differences in umbilical cord serum levels and gestational week of Tdap vaccination (13–19 vs. 20–23 vs. 24–27 vs. 28–31 vs. 32–36 weeks) [29]. However, in a larger cohort (335 pregnant women), Eberhardt et al. [50] showed that vaccinating women in the second versus the third trimester resulted in higher umbilical cord antibody levels (p < 0.001) and infant seropositivity rates (PT antibody concentration > 30 EU/mL at birth; p < 0.001). When second trimester vaccination was compared with vaccination at 26–36 weeks gestation, superiority was again shown for second trimester vaccination [50].

Transplacental transfer of anti-pertussis antibodies to the newborns appears more efficient with maternal Tdap immunization earlier in the third trimester than later [27, 34, 43, 45, 47]. The transfer ratio was also found to be positively impacted by earlier Tdap vaccination [27, 34, 45]; for example, transfer ratio estimates fell from 1.12–1.45 across antigens in the group vaccinated between 27–30 weeks to 0.97–0.98 in the group vaccinated after 36 weeks of gestation [27]. The optimal time for vaccination was suggested as 30 gestational weeks, based on umbilical cord antibody GMCs and estimated GMCs in infants aged 2 months [45]. However, pertussis antibody concentrations in umbilical cord blood at delivery did not differ significantly between groups with maternal vaccination at 27–30 gestational weeks or at 31–35 gestational weeks in another study [49].

Although umbilical cord pertussis antibody levels were positively correlated with the interval between vaccination and delivery (a period of about 2–13 weeks of vaccine exposure was considered) [47], this correlation does not seem to extend to pertussis immunization very early in pregnancy. Healy et al. found that, although vaccination before pregnancy or early in pregnancy both resulted in increased antibodies at birth, there was no difference (p ≥ 0.45) between GMCs of newborns in the two groups [41]. Similarly, cord blood PT and FHA antibody GMCs were similar in infants whose mothers were immunized < 2 weeks prior to delivery or not immunized, suggesting that vaccination < 2 weeks prior to delivery was unlikely to effectively protect the newborn [50]. Optimal timing of vaccination, based on umbilical cord antibody GMCs, was determined to be 30–120 days before delivery (range considered < 15– > 150 days) [50], although 15 days was considered sufficient to significantly elevate umbilical cord antibody GMCs in a preterm population [51]. Other studies have suggest that vaccination ≥ 8 weeks before delivery may maximize antibody levels to pertussis antigens in umbilical cord blood [27, 34, 48], with the optimal timing being 57–84 days before delivery (range considered 1–84 days) [27]. Of note, the timing of vaccination may not only affect the quantity of antibodies transferred but also their functionality, since the relative avidity of antibodies in cord blood increased linearly with time between Tdap vaccination and delivery [43]. In contrast, two studies found no correlation between time from Tdap vaccination to delivery (range 3–21 weeks and about 2–14 weeks, respectively) and pertussis antibody concentrations in umbilical cord blood at delivery [39, 49].

Transplacentally acquired antibodies in early infancy decay over time up until receipt of the first dose of the primary pertussis vaccination series [29, 32, 33, 35, 38,39,40, 46, 52]. Two studies quantified this decay at 58–76% across the pertussis-specific antibodies [29, 38]. Nonetheless, the anti-pertussis antibody levels remained higher through the period before the first pertussis vaccine (generally 8 weeks), with 3.2–24.4-fold higher antibody GMCs persisting in infants of Tdap-vaccinated versus non-vaccinated mothers [32, 33, 35, 38, 40, 45, 46]. Using previously estimated half-life values for maternally acquired antibodies in newborns to model the persistence of anti-pertussis antibodies, it was determined that transferred maternal antibodies would remain detectable at age 2 months in 51–89% of infants born to mothers who received Tdap during pregnancy [39, 41, 52, 53]. These findings suggest that maternal pertussis immunization could bridge the susceptibility gap until the infant receives the first dose of pertussis vaccine.

Infant Immune Responses—Interference with Primary Vaccination

There is a theoretical risk that transplacentally acquired maternal antibodies may blunt the immune responses and thus the protection elicited by the primary diphtheria tetanus acellular pertussis vaccine (DTaP) immunization series. The available studies on the immunological effect found blunting of antibodies to at least one pertussis antigen following all or part of the primary vaccination series among infants whose mothers received pertussis immunization during pregnancy compared with infants of mothers who did not (Table 1) [33, 35, 38, 40, 46, 54]. Whether the blunting effect also persists with the infant booster vaccination is unclear. Some studies suggest that pertussis-specific antibody GMCs either remained lower [32, 55] or did not differ notably [33, 38, 56] 1 month after the booster dose in infants of mothers who received Tdap during pregnancy compared with those of mothers who did not. However, all of the latter studies had small sample sizes, which may have precluded detection of significant differences. In one study, FHA and pertactin antibody levels did not differ 1 month after the booster dose between infants of Tdap-vaccinated versus unvaccinated mothers, but PT antibody levels were significantly lower in the infants of vaccinated mothers [57]. Nonetheless, increases in all relevant pertussis antibody GMCs following the primary and/or booster immunization series are observed in the infants of mother immunized during pregnancy [33, 35, 46, 54, 56, 57].

A prospective cohort study conducted following implementation of the program of immunization of pregnant women in the UK with Tdap-IPV examined the issue of immunological blunting from a broader perspective, exploring the effect on the whole range of vaccines administered to infants in the first few months after birth [53]. Immunological responses to pertussis, diphtheria, tetanus, Haemophilus influenzae type B, pneumococcal, and meningococcal vaccines in infants of vaccinated mothers were compared with historical data from infants vaccinated before program implementation. Although infants born to vaccinated mothers achieved a 2.64-fold increase (p < 0.001) in PT antibodies 1 month after completion of the primary immunization program relative to GMCs before the first primary immunization dose, GMCs for antibodies to FHA and FIM were lower than pre-dose GMCs. In addition, response to all three pertussis antigens was blunted by 33–49% in comparison with the historical data (p < 0.001 for all). In the absence of a pertussis vaccine booster dose in the UK national immunization program, the study could not identify whether this blunting persisted with the booster in second year of life.

Antibody Transfer in Breast Milk

Limited data were identified evaluating the potential for antibody transfer through breast milk and it is unclear whether Tdap vaccination during pregnancy meaningfully affects pertussis antibody concentrations in colostrum or breast milk [58, 59]. While, the value of transfer of breast milk immunoglobulin (Ig)A for clinical protection remains unclear, the data are indicative that secretory antibodies generated by the vaccine are also likely to have been transferred in utero to the fetus and contribute to protection. Abu Raya et al. (2014) found that GMCs of IgA to PT were similar in the colostrum and breast milk of Tdap-vaccinated and unvaccinated women at weeks 2, 4 and 8 post-partum [58]. However, IgA to FHA was significantly higher in colostrum and breast milk from vaccinated women at 2 weeks. IgA to both pertussis antigens significantly declined over the 8-week study period. In the study by De Schutter et al. (2015), women vaccinated during pregnancy (p = 0.012) or at, or shortly after, birth (p = 0.001) had higher levels of secretory IgA to PT in breast milk at a median of about 58 days post-partum versus women not vaccinated for at least 5 years before delivery [59].

Effectiveness Measures

In infants aged < 2 or < 3 months, the effectiveness of pertussis immunization during pregnancy ranges from 53 to 93% in preventing pertussis (Table 2) [60,61,62,63,64,65,66,67]. In addition, Tdap during pregnancy reduces pertussis-related hospitalization in infants, with vaccine effectiveness estimated to be 58–94% [65, 66, 68]. Among hospitalized infants with pertussis, hospital stays were shorter in those whose mothers were vaccinated versus those of unvaccinated mothers (median 3 vs. 6 days; p = 0.02) and, notably, no infants of vaccinated mothers had seizures, required intubation, or died [68]. However, a much smaller study reported that the duration of hospitalization did not differ between infants whose mothers received pertussis vaccination and those of unvaccinated mothers (median 4 vs. 3.5 days; p = 0.58) [63]. Vaccine effectiveness against pertussis-related death was estimated to be 95% in one UK-based study [61].

Vaccine effectiveness was affected by the time between vaccination and delivery, and the gestational week at vaccination. Indeed, effectiveness appeared higher if maternal vaccination occurred during gestational weeks 27–36 than vaccination outside this period during pregnancy [66, 67]. No reduction in pertussis rates were observed in infants of mothers who received Tdap at < 27 gestational weeks in one study with infant follow-up to 18 months [69]. The protection afforded by vaccination prior to pregnancy or post-partum also appears at best low [62, 66]. In the UK program of maternal immunization, effectiveness was similar in infants aged < 2–3 months whose mothers were vaccinated 7–27 days before delivery and in those whose mothers were vaccinated earlier in pregnancy (≥ 28 days before delivery), but lower in those whose mothers were vaccinated between 6 days before and 13 days after delivery (91% vs. 91% vs. 38–43%, respectively) [60, 61].

Three studies investigated the clinical impact of potential immunological blunting of the infant’s response to their primary immunization series [61, 62, 70]. Two studies reported that the relative effectiveness of maternal immunization after each of the three doses of the primary series waned but remained positive, indicating the absence of a negative effect of maternal immunization on the protection afforded by the primary immunization, regardless of any immunological blunting (Table 2 [61, 62]). In the third study, which followed 408 infants of women vaccinated during pregnancy in New Zealand (94.2–98.5% received each of the three infant DTaP doses on time), no infant developed pertussis over a period of up to 12 months despite nine infants having contact with a confirmed case, and there being a high rate of pertussis in the community [70].

Discussion

The concept of “cocooning” was previously considered an appropriate intervention for preventing disease among infants too young to be vaccinated. Over time, it was realized that the broad vaccination coverage required to achieve acceptable protection of the newborn was unrealistic, at least beyond the parents and siblings. Given the resurgence of pertussis, a number of countries have introduced maternal pertussis vaccinations during pregnancy as a strategy to confer protection in young infants.

Our review confirms that maternal pertussis vaccination during pregnancy elicits robust and durable responses to all vaccine antigens similar to those observed in non-pregnant women. Moreover, maternal immunization results in efficient transplacental transfer of anti-pertussis antibodies to the fetus, demonstrated by higher antibody GMCs in cord blood than maternal blood. Pertussis-specific antibodies persisted in the infant and reduced the risk of disease during the period before childhood pertussis vaccination. Although a minimum interval of 2 weeks between maternal vaccine administration and delivery appears to be required, longer intervals, including vaccination during the second trimester or early in the third trimester, lead to higher antibody concentrations in newborn infants and allow for maturation of the immune response, thus improving the quality (i.e., avidity) of antibodies transferred to the fetus [43].

The available evidence suggests that maternal Tdap vaccine effectiveness is maintained during the infant’s primary pertussis vaccination series. It is encouraging that no negative impact on the effectiveness of the primary DTaP series was reported. However, maternal pertussis immunization causes a relative blunting of the infants’ immune response to pertussis antigens after primary immunization with acellular pertussis-component vaccines, although the specific antigen(s) affected varied across studies. Nevertheless, long term consequences of this blunting effect remain to be determined and may take several years to establish. Another question raised may be the potential consequences of blunting on other diseases and in other settings. In addition to reporting the immunogenicity of pertussis vaccine, a UK study identified blunting of responses to other antigens, particularly diphtheria toxoid or diphtheria protein-conjugated vaccines, although most infants (97.7% and 84.4%, respectively) achieved protective antibody concentrations [53]. Short- and long-term clinical consequences need to be ascertained, particularly as the strategy of immunization during pregnancy expands globally and may be introduced in settings where diphtheria remains a public health concern. It is also important to understand whether blunting occurs in infants receiving whole-cell pertussis primary series and whether there are associated short- and long-term consequences.

Additional research will also be needed to clarify whether the findings of this review are applicable to other pertussis vaccines currently in development. Genetically detoxified pertussis toxin-based vaccines are being developed to enhance immunogenicity against PT, both quantitatively and qualitatively. The impact of these vaccines on maternal immunization will need to be studied as part of their research and development. Finally, it should be noted that the overwhelming majority of pregnant women receiving pertussis vaccination were originally primed with whole-cell pertussis vaccines. Whole-cell pertussis vaccine priming induces a robust immune response after subsequent booster doses than does acellular pertussis vaccine priming. It therefore seems reasonable to question, as cohorts of women of child-bearing age shift from whole-cell pertussis vaccine to acellular pertussis vaccine priming, whether their responses to pertussis vaccination in pregnancy will remain robust enough to sufficiently protect the newborn.

Our systematic review has a number of limitations. The heterogeneity in immunological endpoints, the diagnostic assays, seroconversion thresholds, and criteria applied between the studies limit their comparability. In addition, the protective threshold of pertussis-specific antibodies remains to be established, which limits the interpretation of the antibody GMCs presented. A substantial number of studies reviewed were of small sample size, potentially affecting the external validity of the evidence, as well as the power of these studies to detect associations. The restriction to articles published in the English language may have caused some international studies to be overlooked. The strength of this review lies in our adherence to established methods for conducting systematic reviews, including extensive literature searching methods across several databases, and a wide inclusive publication date range. It is reassuring that, despite these limitations, the generally concordant findings between studies supports our conclusions.

There have been four recent reviews assessing the effectiveness and safety of pertussis vaccination in pregnancy [71,72,73,74]. Gkentzi et al. [71], McMillan et al. [72] and Furuta et al. [74] all performed reviews of the literature published up to May 2016, and Campbell et al. through to April 2017 [73]. Our review includes several more recently published studies not captured in the earlier reviews, reflecting the high general interest in maternal pertussis vaccination during pregnancy. The findings in the current review and our accompanying safety review (D’Heilly et al., preparation) are in concordance with the earlier reviews.

Conclusion

In conclusion, this systematic review provides evidence that maternal pertussis immunization with Tdap during pregnancy provides sufficient maternally-derived pertussis antibodies in infants, and protects infants aged < 2 or < 3 months against pertussis. Vaccine effectiveness was not adversely affected by the relative blunting of the infants’ immune response to some, but not all, pertussis antigens after primary immunization with acellular pertussis-component vaccines. However, there remains a need to better characterize the impact of maternal pertussis vaccination, so as to determine how it will evolve as the wider maternal population is exposed to pertussis vaccines during pregnancy, and how it will affect the epidemiology of the disease. Nonetheless, a number of countries now recommend maternal pertussis immunization as part of their overall pertussis control program.