The Protective Role of Maternally Derived Antibodies Against Bordetella Pertussis in Young Infants

Abstract

In 20 infants with polymerase chain reaction–confirmed pertussis <6 months of age, median cord blood anti–pertussis toxin, anti–filamentous hemagglutinin and anti-pertactin IgG antibody values were lower than in 80 matched controls: 10.5 versus 13.5 anti–pertussis toxin IU/mL, 14.5/18.0 (anti–filamentous hemagglutinin) and 6.0/9.0 (anti-pertactin), respectively. Although differences of median (and mean) antibody values between groups were not significant, they are in line with the concept of infant protection by maternal antibodies and thus support the strategy of pertussis booster immunization in pregnant women.

Pertussis, most frequently caused by infection with Bordetella pertussis, is a serious disease with significant morbidity especially in young infants.^1,2 Protection against pertussis can best be achieved by active immunization beginning at age (6-) 8 weeks with 1 or 2 further doses in infancy, followed by regular booster doses in childhood, adolescence and adulthood. Partial protection against complicated pertussis begins after the first dose of immunization³ but does not reach efficacy rates of 80–90% before administration of the 3rd dose and remains less than optimal even with regular booster doses.^4–6

We and others have shown that maternal antibodies against specific B. pertussis antigens do cross the placenta and are measurable in newborns in quantities at least as high as in their mothers and levels gradually decline over the first few months of life.^7,8 However, little is known whether, and if so to what extent, transplacental maternal antibodies do protect infants when exposed to B. pertussis in the first few months of life before they develop immunity by vaccination.⁸ The lack of knowledge is explained by the fact that hardly ever pre-exposure serum specimens are available, which would allow to measure specific antibody levels at the time of exposure.

We took advantage of our collection of serum specimens from cord bloods of newborns born in various obstetric hospitals in the Greater Basel area and a polymerase chain reaction (PCR) laboratory at the University Children’s Hospital in Basel, which provides laboratory confirmation of B. pertussis infection in infants hospitalized with suspected pertussis. Here we present data of specific B. pertussis antibody levels in cord blood specimens of infants <6 months of age with PCR proven pertussis compared with those of age-matched control infants.

PATIENTS AND METHODS

Study Participants

Infants <6 months of age, hospitalized with PCR-confirmed B. pertussis infection at the University Children’s Hospital between 1995 and 2010, were identified from the PCR laboratory’s database. Among these children, those born either in the obstetrical departments of the University Hospital of Basel, Bethesda-Hospital Basel, or the regional hospitals of Bruderholz, Liestal and Rheinfelden (all in Northwestern Switzerland) were eligible for study participation because cord blood specimens have been obtained and conserved at –80°C since 1982 as part of a regional toxoplasmosis screening program. Cord blood specimens were selected for antibody analyses if at least 200 µL were available. For each of the case infants, 4 controls were selected by date of birth. Specifically, cord blood specimens from 2 children born consecutively before and 2 children born consecutively after each case child in any of the collaborating obstetric hospital were chosen as controls. Demographic data such as personal history for pertussis and pertussis immunization status in mothers of these infants are not being obtained and therefore were not available for this study.

Study Procedures

Aliquots of 200 µL serum specimens derived from cord blood of cases and controls were blinded and then sent on ice by overnight courier from Basel to the German pertussis national reference laboratory in Krefeld for antibody analyses.

Enzyme-linked Immunosorbent Assay for Pertussis Antibodies

Antibodies of isotype IgG were measured by in-house enzyme-linked immunosorbent assay with pertussis toxin (PT), filamentous hemagglutinin (FHA) or pertactin (PRN).⁹ In brief, these enzyme-linked immunosorbent assay used purified PT, FHA or PRN, kindly provided by GSK Biologicals SA, Rixensart, Belgium, and employed alkaline phosphatase–conjugated goat antihuman IgG antibodies as tracers (Kierkegard & Perry Laboratories, Gaithersburg, MD). Eight dilutions for every sample were used, and the results were calculated by a 4-parameter logistic. The enzyme-linked immunosorbent assay were standardized to the 1st World Health Organization reference preparation for human pertussis antibodies.¹⁰ The lower level of detection for these assays was 1.7 IU/mL for IgG-anti-PT, 1.8 IU/mL for IgG-anti-FHA and 1.9 IU/mL for IgG-anti-PRN. We set the limit of detection arbitrarily to 2 IU/mL for all tests. For statistical calculation, infant serum specimens with antibody levels below the limit of detection were given a value of 1 IU/mL.

Statistical Analysis

Antibody concentrations were tabulated. For differences in antibody concentration between cases and controls, the Mann–Whitney rank sum was used. Statistics were done with SIGMA Stat and SIGMA-Plot software (Jandel Scientific, Erkrath, Germany).

Ethics

Collection of cord blood specimens had been approved by the ethical committee of the cantons of Basel in 1982 as was the current protocol for this specific study (Ethical committee of Basel and Basel county file number 43/12). Specifically, to obtain informed consent from parents of study or control infants for B. pertussis antibody measurements was not deemed necessary by the ethics committee.

RESULTS

We identified 20 case infants with PCR-confirmed B. pertussis infection and 80 matched control infants. Median anti-PT, anti-FHA and anti-PRN IgG antibody values were consistently lower in cord blood serum specimens of cases than in their controls with values of 10.5 (95% confidence interval: 6–31) and 13.5 (<2–83) anti-PT IU/mL, 14.5 (4–141) and 18.0 (4–117) anti-FHA IU/mL and 6.0 (<2–33) and 9.0 (2–63) anti-PRN IU/mL, respectively. Respective geometric mean values were 14 and 23 IU/mL for anti-PT, 29 and 32 IU/mL for anti-FHA and 10 and 18 IU/mL for anti-PRN in cases and controls, respectively (Fig. 1).

FIGURE 1: Median (horizontal line in grey boxes) IgG antibody values with interquartile range (grey boxes), 10–90% interval (upper and lower lines) and outliers (black dots) above and below for anti-PT, anti-FHA and anti-pertactin in 20 case infants and their 80 matched controls.

When antibody analyses were restricted to those 9 infants who were <3 months of age when they developed pertussis, differences were more pronounced compared with the whole study group with lower antibody levels at birth for case infants compared with their matched controls (Fig., Supplemental Digital Content 1, https://links.lww.com/INF/B505). Geometric mean value (cases/controls) were 11/30 IU/mL, 18/35 IU/mL and 11/20 IU/mL for anti-PT, anti-FHA and anti-PRN, respectively. Median values were 7 (95% confidence interval: 6–30) and 13 (<2–87) IU/mL for anti-PT, 14 (4–56) and 24 (4–118) IU/mL for anti-FHA and 6 (3–42) and 9 (2–88) IU/mL for anti-PRN. Despite consistently lower values for cases, the differences were not statistically significant with P values of 0.418, 0.139 and 0.454 for PT, FHA and PRN, respectively.

DISCUSSION

Our analyses of anti-PT, anti-FHA and anti-pertactin IgG serum antibody levels in cord blood specimens from infants who acquired pertussis in their first few months of life compared with those in control infants did not reveal statistically significant differences. However, the observation of lower median and mean levels for all 3 antigens in cases compared with controls leads us to speculate that certain levels of protection may in fact exist, but our study had not the power to demonstrate this due to a small number of cases and the following 3 methodological considerations: (1) Assuming that serological correlates of protection do exist, antibody levels in some control infants may have been below that threshold and yet because of lack of exposure to B. pertussis they may not have acquired pertussis disease. The ideal control specimens therefore should have been cord blood specimens from unimmunized children with demonstrated significant exposure to B. pertussis in the first few months of life (eg, household contact) and remained healthy as a proof of protection. However, such a set of specimens is not available. (2) Immunization records of cases and controls were not available. Therefore, we cannot rule out the possibility that case and control infants had different pertussis immunization histories that would have influenced their levels of protection against pertussis in their first 6 months of life. However, this would only have had influence on controls who might have been exposed and were protected by immunization rather than remaining maternal antibodies. We consider it unlikely that this would have had a significant impact on the mean and median values in the control group. (3) The decay of maternal IgG antibodies in the serum of infants over the first few months of their lives is significant; therefore, protection by maternal B. pertussis antibodies may only last for a few weeks, even if levels are high at birth. Unfortunately, there were not enough cases to stratify our patients by narrow age groups, but the data of the subset of case infants <3 months of age support this concept (Fig., Supplemental Digital Content 1, https://links.lww.com/INF/B505).

Despite these limitations, we were able to demonstrate that infants who acquired pertussis in the first 3–6 months of life did have specific anti-pertussis antibody levels at birth that were numerically lower than those measured in their matched controls and those from a separate cohort of newborns from our population.⁷

The ultimate goal of pertussis immunization is to decrease morbidity and mortality due to pertussis in young infants. Although cell-mediated immunity and nonspecific mechanisms have been intensively studied in murine models of B. pertussis infections¹¹ and cell-mediated responses can be elicited in young infants,¹² the relative role of antibodies, cell-mediated immunity and innate immune mechanisms for protection or susceptibility to pertussis in humans remains unclear.¹³ Clinical protection against Pertussis, however, can be achieved by immunization. In this context, it is noteworthy to reflect the results of a study in which neonates received either monovalent acellular pertussis (aP) vaccine (containing PT, FHA and PRN) or hepatitis B vaccine (controls) at birth, followed by aP containing standard combination vaccines at 2, 4 and 6 months of age. At 3 months of age, immunization with aP at birth had induced significantly higher antibody responses to pertussis vaccine antigens compared with controls¹⁴ and after completion of the primary series at 7 months of age and after the booster dose at 11–18 months of age antibody levels in the aP group were equivalent to those in controls.¹⁵ Moreover, consistently higher lymphoproliferative immune responses were measured at all study time points in the neonatal aP immunization group.

With regards to potential protection of young infants from pertussis by maternal antibodies, the concept of immunizing pregnant women against pertussis has been developed and recently recommended in the United States.¹⁶ This concept is based on the assumption that transplacental antibody transfer to infants will lead to direct protection in addition to indirect protection by the so called cocoon strategy, where future close contacts of young infants are immunized before or shortly after birth of the newborns to reduce their the risk of exposure to B. pertussis. Therefore, further studies are necessary to better define whether transplacentally acquired anti-pertussis antibodies do protect infants or not.

ACKNOWLEDGMENT

We would like to thank Jacqueline Glaus, microbiological laboratories of the University Children’s Hospital Basel, for excellent management of the cord blood collection.