Pertussis Antibodies in Postpartum Women and Their Newborns

Abstract

Objectives

To 1) determine the proportion of mothers and infants who had levels of IgG antibody to pertussis antigens predicted to be potentially protective at delivery; 2) evaluate the efficiency of maternal-infant antibody transport; 3) extrapolate infant antibody titers at six weeks; and 4) identify maternal factors associated with potentially protective infant antibodies.

Study Design

Sera from mother-infant pairs from February 2006 through April 2007 were tested for antibody to pertussis antigens by standardized ELISA (enzyme-linked immunosorbent assay). Potentially protective antibody levels were defined as >5 ELISA units (EU) for pertussis toxin (PT) and >10 EU for fimbriae (FIM) and pertactin (PRN). Serologic evidence of previous maternal infection was defined from antibody to four antigens by k-means cluster analysis.

Results

In total, 21% (17/81) of mothers and 26% (21/81) of infants had potentially protective antibody levels at delivery. Mean infant-maternal antibody ratios for PT, FIM and PRN were 1.26, 1.36 and 1.31 respectively. At 6 weeks, 11% (9/81) of infants were predicted to have potentially protective antibody levels. Using cluster analysis, 9% (7/81) of mothers had evidence of previous pertussis infection. Infants born to these mothers were predicted to be more likely to have potentially protective antibodies at 6 weeks (43%) than those born to mothers without (8%) (p = 0.03).

Conclusion

Approximately 75% of infants were born with pertussis antibody levels lower than the modest levels associated with potential protection. Despite effective antibody transfer, nearly 90% of infants were predicted to have little antibody by 6 weeks. Maternal immunization before or during pregnancy might simulate previous pertussis infection and help protect infants through the first months of life.

INTRODUCTION

Pertussis, an endemic and common infectious disease, is of particular importance due to a recent striking increase in the incidence of reported cases and greatest morbidity and mortality in the youngest infants.^1–3 In 2004–2005, a total of 56 deaths from pertussis in children younger than 3 months were reported to the Centers for Disease Control and Prevention (CDC).⁴ Because infants do not complete the primary immunization series against pertussis until their sixth month of life, they are particularly susceptible to pertussis infection and are dependent on maternal antibodies for protection.⁵

Although precise levels of antibody required for protection from acute pertussis infection have been debated,^1,2 modest levels of IgG antibody to fimbriae (FIM), pertactin (PRN) and pertussis toxin (PT) have been associated with disease prevention.^6,7 Although filamentous hemagglutinin (FHA) is a component of all licensed pertussis vaccines and antibody against FHA is associated with natural infection, it has not been proven to play a primary role in prevention of pertussis infection.^6–9

Several articles have postulated that immunizing pregnant women against pertussis may provide protection to their newborns, ^10–13 but the CDC’s Advisory Committee on Immunization Practices (ACIP) does not currently recommend this practice.¹⁴ Previous studies have also shown that infants born at or near term have higher antibody levels to specific pathogens than their mothers as a result of active transport of maternal IgG.^15,16 A better understanding of the natural history of transplacentally acquired pertussis antibodies in infants is critical for predicting whether maternal immunization might provide protection from infection to newborns.

To further elucidate the potential of maternal pertussis antibody to provide protection against pertussis for newborns in the months before their scheduled active immunization, the objectives of our study were to 1) determine the proportion of mothers and infants who had levels of IgG antibody to pertussis antigens predicted to be potentially protective at delivery; 2) evaluate the efficiency of maternal-infant antibody transport; 3) extrapolate infant antibody titers at 6 weeks, and 4) identify maternal factors associated with potentially protective infant antibodies.

METHODS

Protection of human subjects

Approval to conduct this study was granted by the Institutional Review Boards of the University of New Mexico and the University of Utah. Mothers provided informed consent for themselves and their infants.

Study subjects

Women aged 18–45 years of age who delivered healthy term infants ≥ 37 weeks gestation were enrolled from the University of New Mexico Health Sciences Center from February 2006 through April 2007. Mother-infant pairs were excluded for multiple gestation, antenatal detection of a major birth defect in the infant, or serious underlying neurological, cardiac, renal, or pulmonary disease in either mother or infant. Mother-infant pairs were also excluded if the infant required neonatal intensive care.

Data collection

Participants were enrolled by the University of New Mexico General Clinical Research Center (GCRC) pediatric research nurses within 48 h after delivery and before hospital discharge. Demographic data included age, previous pregnancy history, occupation, education, ethnicity, marital status, and number of people in the household. In addition, a history of cough greater than 3 weeks in either mother or any household contact during the pregnancy was obtained. Women enrolled in 2007 (n=17) were asked whether they had received pertussis vaccine (Tdap). Immunization status was not documented for women enrolled in 2006

Laboratory methods

2 ml each of maternal and infant sera were collected by venipuncture, centrifuged and stored at −20 to −40° C, until testing by a standardized enzyme-linked immunosorbent assay (ELISA) in the Pediatric Infectious Disease Research Laboratory at Vanderbilt University School of Medicine in Nashville, Tennessee (KME). Pertussis-specific IgG to PT, FHA, FIM, and PRN were quantified by previously described methods.¹⁷ Briefly, Immulon 2 microtiter plates (VWR International, West Chester, PA, USA) were coated with standard quantities of PT, FHA, FIM, or PRN. Serial dilutions of mother-infant serum pairs were incubated for 2 h at 28° C, and an appropriate dilution of alkaline phosphatase–conjugated goat anti–human IgG was added. The reaction was developed and read at 405 nm. The lower limit of detection of each specific antibody was determined by repeated measurements of serially diluted reference material for each antigen and was 1 EU for PT, FHA, and PRN and 2 EU for FIM.

Rationale for selecting pertussis antibody levels associated with potential protection and statistical analysis

On the basis of reports in the literature, the half-life of transplacental pertussis-specific IgG was defined as 6 weeks.^18,19 Antibody levels determined to be potentially protective were based on vaccine efficacy trial data.^6,7 An analysis by Storsaeter et al.⁶ fit a logistic model that dichotomized ‘low’ pertussis antibody levels for FIM, PRN, PT, and FHA as <5 EU and “high” as ≥5 EU.⁶ A similar analysis by Cherry et al.⁷ determined high/low cut points for these pertussis antibodies from 1.5 to 8 EU and with a cut point of 5.7 EU for PT.⁷ As there was greater variability in cut points for FIM and PRN in previous studies, potentially protective antibody levels for FIM and PRN were defined more conservatively at >10 EU, whereas elevated levels of PT were defined as >5 EU. We hypothesized that mothers with highest antibody levels would give birth to infants with higher antibody levels and that their infants would be more likely to be protected from pertussis infection at birth than infants born to mothers with lower levels. However, because ‘no level of antibody, presence of specific antibodies, or antibody profile has been accepted universally as a quantifiable serologic measure of protection,’¹⁴ we performed a cluster analysis of all four pertussis antibodies to identify women with antibody levels that were statistically greater than the study population as a whole. As it was unlikely that any of the women in our study received adult pertussis immunization, women with the highest antibody titers were likely to have had naturally occurring pertussis infection at some time before the delivery. Infection-acquired immunity may persist for 7–10 years after infection^20,21 and significant antibody titers have been documented 5 years post-immunization.^22,23 A marker for previous maternal infection was generated by a multivariable k-means cluster analysis looking for a high and low cluster of the four antibodies simultaneously. Women with all four antibody titers in the highest cluster were classified as likely to have had previous infection. Mean infant-maternal antibody ratios were calculated to determine the efficiency of maternal-infant antibody transfer. Data were analyzed using Stata Version 10.0 (Stata Corp LP).

RESULTS

Enrollment

A total of 83 mother-infant pairs were enrolled in the study, but two pairs were excluded due to missing infant data, leaving 81 mother infant pairs for analysis. The mean maternal age was 27 years. Self-identified maternal race/ethnicity was: 31% White, 25% Hispanic, 9% Native American, 28% Mixed Race/Ethnicity (most commonly combinations of Hispanic, Native American, or White) and 7% others.

Maternal Pertussis History

None of the mothers reported that they or a household contact had had a 3-week cough during the pregnancy and none of the 17 mothers enrolled in 2007 reported receiving Tdap.

Mother and newborn infant antibody titers

Of the 81 mother-infant pairs with complete data, 21% (17/81) of mothers and 26% (21/81) of infants had PT >5 EU and FIM and PRN >10 EU at delivery. Mean infant-maternal antibody ratios for PT, FIM and PRN were 1.26, 1.36, and 1.31, respectively. There were no differences in maternal age, pregnancy history, occupation, education, ethnicity, marital status, or number of household members in infants with higher antibody titers compared with those with lower titers.

Predicted antibody titers at 6 weeks

Using published half-life data for decline of maternal antibody,^18,19 we predicted that 11% (9/81) of infants would have levels of PT >5 EU and FIM and PRN >10 EU at 6 weeks.

Infant antibody titers analyzed by maternal infection status

None of the women enrolled had antibody evidence of acute pertussis infection, as previously defined by PT levels ≥ 94 EU.²⁴ Maternal PT antibody levels ranged from 1 to 40 EU. However, 9% (7/81) had evidence of previous infection (likely within 7 to 10 years of the delivery date)^20,21 based on the k-means cluster analysis (Figure 1). These seven women did not differ from the other mothers with respect to age, ethnicity, employment, or family size.

Figure 1.

Maternal IgG to FIM, PRN, PT and FHA

The mean levels of FIM, PRN and PT of infants born to the seven mothers with evidence of previous infection were all significantly higher than those of infants born to mothers without serologic evidence of infection (145 and 32.9 EU, p < 0.001,;43.4 and 16.7 EU, p = 0.01; 21.4 and 9.4 EU, p = 0.01, respectively). We predicted that at 6 weeks, 3 (43%) of the 7 infants born to mothers with titers compatible with previous pertussis infection would have PT antibody >5 EU and FIM and PRN antibody >10 EU compared with 8% (6/74) of the infants born to mothers without such evidence (p=0.03).

DISCUSSION

The most important finding from our study is that only one-quarter of infants were born with potentially protective pertussis antibody titers >5 EU for PT and >10 EU for FIM and PRN. In addition, we predicted that by 6 weeks only about 10% would have such levels. Thus, nearly 90% of infants were likely to be susceptible to pertussis infection by 6 weeks of age, despite effective maternal antibody transfer. The timing of predicted antibody loss is before the initiation of the active infant immunization series. Notably, infants born to mothers with serologic evidence of previous pertussis infection were predicted to be more likely to have potentially protective antibody levels at 6 weeks. These data indicate that by simulating previous infection, maternal immunization with acellular pertussis vaccine before or during pregnancy might provide increased transplacental protection to newborns.

The majority of infants in our study had low antibody levels to all pertussis antigens at birth. Effective maternal-infant antibody transfer was confirmed given that mean infant-maternal antibody ratios were 1.26 to 1.36 for the pertussis antigens. In spite of effective antibody transfer, only one-quarter of infants were born with the modest antibody levels that we predicted could confer protection, and by 6 weeks, only 11% were predicted to have antibody levels associated with potential protection. Similarly, Healy et al. measured IgG titers in mothers and infants to three pertussis-specific antibodies and found extremely low levels of maternal and infants pertussis antibody at delivery with a rapid decline in the infants’ antibody levels by 2 months.¹⁹

The current pediatric vaccine schedule recommends initiation of pertussis vaccination at 6 to 8 weeks of age; full infant protection is likely not achieved until after the third vaccination at 6 months. Although 75% of infant mortality from pertussis infection has been reported in children <2 months of age, ²⁵ the current recommendation does not protect infants during this vulnerable period. Halasa et al.²⁶ recently evaluated the effect of an additional dose of Diphtheria, Tetanus, and acellular Pertussis (DTaP) vaccine at birth and found that the additional dose was associated with a significantly lower subsequent antibody response to pertussis and diphtheria vaccine compared with controls. This study shows that immunization strategies focusing on infants may not provide the needed early protection from pertussis infection.

As there is no universally accepted serologic measure of protection from pertussis infection, we hypothesized that mothers with the highest antibody levels would give birth to infants with higher antibody levels, and used cluster analysis to identify women with highest levels. The infants in our study with higher antibody levels were born to women who were identified by the cluster analysis as having the highest antibody levels. These women likely derived their antibody levels through pertussis infection, since none had a documented history of pertussis immunization. None of the mothers identified as having had evidence of previous pertussis infection met the serologic definition of acute pertussis (PT≥94) and none had or were exposed to pertussis during pregnancy, making acute infection as a cause for elevated antibody levels unlikely. Modestly elevated pertussis antibody titers may persist for 7 to 10 years after infection or immunization, ^20–23 and the higher antibody levels shown are most consistent with infection in the years before delivery.

Immunization as a strategy to mimic natural infection offers a potentially safer alternative for increasing maternal antibody. Tetanus, Diptheria and acellular Pertussis immunization trials in adolescents and adults showed that 93% of vacinees achieved PT levels ≥ 5 EU for up to 5 years post-immunization.^22,23 ACIP currently recommends that adolescents receive a single dose of Tdap, with preference for 11 to 12 year-olds.⁴ If protection from immunization lasts ~ 5 years, then most women of child-bearing age, if immunized in early adolescence, would be predicted to have low levels of pertussis antibody.

The principle of utilizing transplacental immunity to protect infants from vaccine-preventable diseases is not novel. Vaccination of pregnant women with tetanus toxoids has markedly decreased the global incidence of neonatal tetanus.^10,11,27,28 A small trial of influenza immunization during pregnancy in women in Bangladesh demonstrated a 63% reduction in laboratory-confirmed influenza and a 29% reduction in the incidence of febrile respiratory illness in infants born to mothers who received influenza vaccination in the third trimester of pregnancy.²⁹

The safety of any vaccine administered during pregnancy is paramount, and there are no published data of Tdap use in pregnant women to document safety in the mother or the fetus. ACIP has cited two potential concerns with respect to Tdap use in pregnancy.¹⁴ First, there is a lack of evidence that transplacental maternal antibody induced by Tdap administered during pregnancy will protect infants against pertussis. Prospective large-scale studies will be needed to address this question. Second is the concern that Tdap administered during pregnancy could blunt an infant's immune response to the primary pediatric pertussis immunization series, resulting in increased disease susceptibility later. We did not examine infant response to the primary pertussis series in this study, but previous studies have shown that infants who receive maternally derived pertussis antibody have either no or only a modestly decreased antibody response to acellular vaccine,^12,30

Our study had several limitations. Our sample size was small and may not be representative of other geographical areas, particularly with respect to ethnicity and race. As the 6-week infant 12 antibody titers were extrapolated rather than measured, it is possible that actual levels would differ. However, the half-life of pertussis antibody has been well-defined in previous studies,^18,19 and we believe that our extrapolation likely provides a reasonable indication of the proportion of 6-week old infants who would be susceptible to pertussis. Previous studies have also shown cross-reactivity between some pertussis antigens with other infectious agents.^31,32 The inclusion of all four antibodies in the cluster analysis attempts to limit the effect of cross-reactivity by defining previous maternal infection based on four antigens associated with acute pertussis infection.

CONCLUSIONS

75% of infants were born with pertussis antibody levels lower than the modest levels associated with potential protection. Despite effective antibody transfer, nearly 90% of infants were predicted to have little antibody by 6 weeks. Maternal immunization before or during pregnancy might simulate previous pertussis infection and help protect infants through the first months of life.