BACKGROUND
Placental
transfer of maternal IgG antibodies to the fetus is an important mechanism that
provides protection to the infant while his/her humoral response is
inefficient. IgG is the only antibody class that significantly crosses the
human placenta. Factors determining placental antibody transfer from mothers to
their infants may include maternal total and specific IgG levels, maternal
infectious diseases, placental integrity, IgG subclass, half-life of the
passive antibodies, nature of the antigen and timing of vaccination (or
infection) [1-3]. Maternally acquired passive immunity generally starts to wane
soon after birth, reaches a nadir at an estimated 6 months of life, and is
replaced by antibodies made by the infant in response to active immunization or
natural infection. Antibody response to T-cell dependent (protein) antigen is
more intense compared to T-cell independent (polysaccharide) antigen during the
neonatal period- a shift from Th1 (i.e., oriented towards cell-mediated
immunity) towards Th2 (i.e., oriented towards humoral immunity) occurs during
third trimester of pregnancy [4,5] and continues during infancy.
Neonatal IgG
levels usually correlate with maternal levels. It has been suggested that once
maternal total IgG levels reach a threshold (>15 g/L), neonatal Fc receptor (FcRn, or Brambell receptor) on placenta can
become saturated [1,6] and IgG then competes for a finite number of FcRn
receptors. Unbound IgG molecules are subsequently destroyed through the
lysosomal degradation process within the cells. This finding had also been
supported by studies demonstrating reduced IgG transfer ratios were associated
with higher maternal total IgG levels [7-9]. Levels of IgG in human
immunodeficiency virus (HIV) infected mothers depend on the immunological and
clinical status of their disease condition. It is well established in
literature that maternal chronic infections (malaria, HIV), due to
hypergammaglobulinemia from the infection may reduce placental transfer of
specific IgG to vaccine induced immunity [1,8-10]. Thus, HIV- infected women, who have
hypergammaglobulinemia due to polyclonal activation of their B cells [10,11], may have non-specific antibodies
block the receptors (FcRn) on syncytiotrophoblast cells and reduce placental
transfer of specific antibodies. The reduced transfer of antibody is
thought to be influenced by concomitant infection and inflammation of the
placenta, a reduction in FcRn-antibody binding avidity, or via induction
of hyper-gammaglobulinemia (IgG > 15 g/L) [12]. This phenomenon can also be applied to pregnant women in low- and
middle-income countries (LMIC), where there is a high burden of infectious
diseases, superimposed on malnutrition (an immune-compromising condition).
Therefore, concern should remain whether there is inadequate placental transfer
of specific IgG in HIV-infected women and pregnant women in LMIC. This concern
remains valid regardless of whether the infant becomes infected with HIV or
not. The author has already demonstrated an overall decreased antibody levels
to routine vaccine preventable infections in HIV-exposed infants compared to
their unexposed counterparts [13-15].
IgG is the
only immunoglobulin class that can cross human placenta in significant amounts.
There is preferential transfer of IgG subclasses across
placenta; level of IgG1, but not IgG2 in cord blood has been demonstrated to
exceed maternal levels of the antibodies [2,16]. Thus, placental transport of
IgG2, is shown to be considerably less efficient than that of IgG1, IgG3 or
IgG4 [17,18].
The primary objective of this manuscript is to
review, and reanalyze data obtained from several studies of the author, from a
different perspective to see if there is preferentially decreased transfer of
IgG to polysaccharide antigens compared to the protein antigens, in addition to
assess if maternal HIV infection has any impact on the neonate. The secondary
objective is to discuss justification for immunizing mothers during pregnancy
against vaccine preventable infections, particularly against polysaccharide
antigens.
METHODS
Data from
relevant studies of the author were reviewed and reanalyzed. All study
protocols were approved by the Institutional Review Board at Meharry Medical
College (MMC). All studies were performed at Meharry Medical College and
adjacent General Hospital in Nashville, Tennessee (TN). All women were enrolled
during their first, second, or third trimesters of pregnancy and informed
written consents were obtained. All relevant information was obtained from the
subjects’ medical records. Pregnant women
were followed prospectively up to their deliveries when cord blood samples were
obtained.
Study Population
Fifteen
HIV-infected and 34 HIV-uninfected pregnant women were enrolled into the study.
The mean (range) ages of pregnant women at the time blood specimens were
obtained were 27 (18-40) and 25 (15-41) years for HIV-infected and HIV-
uninfected women, respectively. Of the fifteen HIV-infected women, sixty-two
percent were on antiretroviral therapy. Their mean (range) CD4 cell count was
484 (210- 1053) cells/mm3 and their mean (range) viral load was 47,320 (400-
278,167) copies/ml.
Pregnant women
There is no record of previous exposure to pneumococcal or Hemophilus influenzae type b vaccines in
the clinical histories of the women enrolled in this study.
Infants
All preterm
infants born at less than 38 weeks were excluded from the study. Not all cord
blood samples were mother-infant pairs.
Sample size and power
calculation
The sample size
calculation was based on a convenient sample of HIV-infected women who were
available to enroll over a period of fourteen months (June’ 2000- August’
2001). The control group included HIV-uninfected women at a ratio of
approximately 2:1 with HIV-infected women. Due to reasons such as insufficient
quantity, accidental spillage, and missed opportunity for sample collection,
the final number of samples for analyses varied slightly for each category of
polysaccharide and protein antigens.
Determination of antibody
levels
IgG levels
against twenty-three serotypes (1-5, 6B, 7F, 8, 9N, 9V, 10A, 11A, 12F, 14, 15B,
17F, 18C, 19F, 19A, 20, 22F, 23F, 33F) as a total of pneumococcal capsular
Polysaccharide (PCP) of Streptococcus pneumoniae (SPN), polyribosyl-phosphate
(PRP) of Hemophilus influenzae type b (Hib) and tetanus toxoid (TT) were
performed in the authors’ laboratories by enzyme linked immunosorbent assay
(ELISA) utilizing the kits BINDAZYMETM (Anti- Hemophilus B IgG “Kit MK 016” for
H. influenzae type b and Anti-PCP IgG “Kit MK 012” for S. pneumoniae
[ww.bindingsite.co.uk]). Correlates of immunity were defined as antibody levels
>0.35 µg/ml for PCP of Streptococcus pneumoniae [19,20], >1.0
µg/ml for PRP of Hemophilus influenzae type b [21], and >0.01 IU/ml
for tetanus toxoid [22]. IgG against measles (MS) was performed by ELISA at
specialty laboratories, Santa Monica, California. Correlates of protective
immunity for measles was defined as a level >1.09 optical density ratio
(ODR) [23].
Immunologic and
virologic studies
T cell analyses were performed by flow cytometry
in the authors’ laboratories at MMC for HIV-seronegative group and at
Vanderbilt University Medical Center (VUMC) core laboratories; Nashville, TN for
HIV infected mothers and HIV-exposed infants. HIV viral loads in HIV- infected
pregnant women and HIV deoxyribonucleic acid (DNA) testing in their infants
were performed by polymerase chain reaction (PCR) at VUMC.
Statistical analysis
The
statistical analyses were performed using the software program Intercooled
Stata version 8. Continuous variables were rounded to the nearest tenth and
presented as mean + SD; a p ≤ 0.05 was considered significant.
1. Prevalence
of protective immunity for a total of 23 serotypes of SPN (>0.35 µg/ml),
Hib (>1.0 µg/ml), TT (>0.01 IU/ml) and MS (>1.09
ODR) between groups of pregnant women and their infants was performed by two-
tailed Fisher’s exact test.
2. Paired mother-infant differences in mean IgG levels
between groups were determined by two-tailed t test.
3. Effect of
confounding variables (socioeconomic status, age of mother, gestational age of
babies, and number of pregnancies) on immune response was done using ANOVA with
Bonferroni correction, yielding a significance level of 0.01.
RESULTS
Number of available samples
Total of fifteen
HIV-infected and thirty-four HIV-uninfected mothers. Sample size distribution
for each category of antigen in mothers and their infants are shown in Table
1.
SPN Immunity in HIV-infected and HIV-uninfected mothers: The mean anti-PCPIgG level was 2-fold lower in the
eight HIV-infected women (77.63 ug/ml [SD + 66.04]) compared to their
twenty-four HIV-uninfected (152.39.ug/ml [SD + 190.27]) counterparts.
All HIV-infected and HIV-uninfected mothers had protective level (anti-PCP >0.35 ug/ml) of IgG
antibodies.
SPN immunity in HIV-exposed and HIV-unexposed infants: Mean anti-PCPIgG was 2.5-fold lower in the
HIV-exposed (50.83 ug/ml [SD + 52.58]) compared to their HIV-unexposed
(124.42 ug/ml [SD + 167.69]) counterparts. Protective level (anti-PCP >0.35ug/ml) of IgG was
prevalent in all HIV-exposed and-HIV-unexposed cord bloods.
SPN immunity in matched mother-infant pairs: While mean anti-PCPIgG was only slightly lower in HIV-exposed infants (61.02 ug/ml
[SD + 65.59]) compared to their HIV-infected mothers (77.63 ug/ml [SD +
66.05]), mean anti-PCP was
2-fold lower in the eighteen HIV-unexposed infants (98.98 ug/ml [SD +
134.82]) compared to their HIV-uninfected mothers (180.81 ug/ml [SD +
246.59).
Hib immunity in HIV-infected versus HIV-uninfected mothers: Mean anti-PRP IgG was 4-fold lower in the eight
HIV-infected women (1.15 ug/ml [SD + 1.41]) compared to their twenty-six
HIV-uninfected counterparts (4.77 ug/ml [SD + 6.77]). Protective
antibody levels (anti-PRP >1
ug/ml) were detected in only two of eight (25%) of HIV-infected mothers as
compared with fourteen of twenty-six (54%) of the HIV-uninfected mothers.
Hib immunity in HIV-exposed versus HIV-unexposed infants: Mean anti-PRP IgG was slightly lower (2.07
ug/ml [SD + 4.23]) in the thirteen HIV-exposed compared to twenty-six
HIV-unexposed (2.73 ug/ml [SD + 3.03]) cord bloods. Protective antibody
levels (anti-PRP> 1.0
ug /ml) were significantly (p=0.05) lower in HIV-exposed (three of thirteen
[23%]) compared to fifteen of twenty-six (58%) of the HIV-unexposed cord
bloods.
Hib immunity in matched mother-infant pairs: While mean anti-PRPIgG was comparable between the eight HIV-exposed (1.39 ug/ml
[SD + 2.18]) and their HIV-infected mothers ((1.16 ug/ml [SD +
1.41]), mean anti-PRP was
2-fold lower (p=0.09) in twenty HIV-unexposed infants (2.93 ug/ml [SD +
3.0]) compared to their HIV-uninfected mothers (5.98 ug/ml [SD + 7.27]).
Protein Antigens
(Tetanus Toxoid [TT] and Measles [MS])
Tetanus
toxoid immunity in HIV-infected versus HIV-uninfected mothers: The mean anti-TT IgG was 1.5- fold
lower in the nine HIV-infected (1.37 IU/ml [SD + 1.37]) women compared
to their twenty-four HIV-uninfected (2.09 IU/ml [SD + 2.39])
counterparts. Protective antibody level (anti-TT > 0.01 IU/ml) was prevalent
in seven of nine (78%) HIV-infected compared to thirteen of twenty-four (54%)
of the HIV-uninfected mothers.
Tetanus
toxoid immunity in HIV-exposed versus HIV-unexposed infants: Mean anti-TT IgG was
2-fold lower in the nine HIV-exposed infants (1.59 IU/ml [SD + 1.70])
compared to their twenty-four HIV-unexposed (2.85 [SD + 4.21])
counterparts. Protective antibody level (anti-TT > 0.01IU/ml) was prevalent
in seven of nine (78%) HIV-exposed compared to thirteen of twenty-four (54%) of
the HIV-unexposed cord bloods.
TT immunity
in matched mother-infant pairs: Mean anti-TT IgG was slightly higher in both nine HIV- exposed (1.59
IU/ml [SD + 1.70]) and twenty-four HIV- unexposed cord bloods (2.85
IU/ml [SD + 4.21]) compared to their HIV-infected (1.37 IU/ml [SD +
1.37]) and HIV-uninfected mothers (2.09 IU/ml [SD + 2.39]),
respectively.
Measles
immunity in HIV-infected versus HIV-uninfected mothers: Mean anti- MS IgG was
significantly (p=0.03) lower in the eleven HIV-infected (1.39 ODR [SD ± 0.77]) women compared to their thirty-three HIV-uninfected (2.09 ODR [SD ± 0.97]) counterparts.
Protective antibody levels (anti-MS >1.09 ODR) were prevalent in six
of eleven (54%) HIV-infected mothers compared to twenty-eight of thirty-three
(85%) of HIV-uninfected mothers.
Measles
immunity in HIV-exposed versus HIV-unexposed infants: Mean anti-MS IgG was
considerably lower (p=0.06) in thirteen HIV-exposed cord bloods (1.50 ODR [SD ±
1.10]) compared to their twenty-seven HIV-unexposed (2.41 [SD ± 1.52)])
counterparts. Protective antibody (anti-MS >1.09 ODR) was prevalent
in nine of thirteen (59%) HIV-exposed compared to twenty-one of twenty-seven
(78%) of the HIV-unexposed cord bloods.
MS immunity in matched mother-infant pairs: Mean anti-MS IgG was slightly higher in both eleven HIV- exposed
(1.65 ODR [SD + 1.11) and twenty-three HIV-unexposed cord bloods (2.53
ODR [SD + 1.56] compared to their HIV-infected (1.39 ODR [SD +
0.77]) and their HIV-uninfected mothers (2.06 ODR [SD + 1.05]),
respectively.
DISCUSSION
This manuscript describes a review and
analyses from a different perspective, of data collected from past passive
immunity studies of the author. The primary
objective of this data analysis was to assess if differences existed in the
placental transfer of antibodies specific for polysaccharide versus protein
antigens and between HIV- exposed and unexposed neonates. The secondary
objective was to assess if results from this analysis would support maternal
immunization during pregnancy of not only HIV-infected women but also
HIV-uninfected women in resource limited settings, with ultimate goals to boost
passive immunities in their neonates against vaccine preventable diseases.
As shown in Table
2, it is interesting to note that the placental transfer of antibodies
(ratio of cord blood/maternal blood) was significantly lower for the
polysaccharide antigens (PCP and PRP) compared to the protein antigens (tetanus
toxoid and measles) within the HIV- uninfected mother-infant pairs (Figure
1).
However,
similar discrepancy was not noted within the HIV- infected group of
mother-infant pairs, except for between PCP of SPN and tetanus toxoid antigens.
This finding reinforces findings in literature that antibodies to
polysaccharide antigens (IgG2) may have significantly lower avidity for
placental transfer compared to protein antigens (IgG1, IgG3 or IgG4) [4,5]. Our
study results further suggest that, among the antigens analyzed in this study,
antibody against tetanus toxoid may have the strongest avidity and antibody against
PCP may have the weakest avidity for trans-placental transfer.
One of the objectives of our study was to
examine impact of HIV infection in pregnant mothers on passive immunities in
their newborns (Figures 2 and 3).
While no significant differences in mean
antibody levels were noted in our study between HIV-infected and HIV-uninfected
groups of mothers or infants, except those for measles, mean antibody levels
were consistently lower for both polysaccharide and protein antigens in the
HIV-infected group, compared to their HIV-uninfected counterparts. Although
protective antibody levels prevailed for tetanus toxoid and SPN in all mothers and infants of both
groups, suboptimal protection was noted for measles and Hemophilus influenzae type b in the HIV group compared to their
HIV-uninfected counterparts (Tables 3 and 4). Indeed, a high level of
protective immunity to SPN and TT in
this population may be attributable to subclinical infections and herd immunity
to SPN and TT, respectively.
Although, our study underlined interesting differences that support the contention of lower passive immunities in
HIV- infected mothers and their exposed infants, further efforts to increase
sample size should allow for improved confidence in significant differences
between groups.
We have further noted that antibody levels
to the protein antigens were higher in the cord blood compared to those in
their maternal blood samples in both HIV-infected and uninfected groups (Table
5). This finding may be explained by that immune response during the
neonatal period is T cell dependent for protein antigens only [2,16]. Thus,
trans-placental transfer of antibody only against tetanus toxoid was influenced
by maternal CD4 counts in our study population. The above finding re-emphasizes
the recommendation to maintain optimal CD4 counts and undetectable viral loads
during pregnancy in HIV-infected mothers with use of highly active
antiretroviral therapy to augment placental antibody transfer, in addition to
maintaining optimal levels of specific antibodies during pregnancy.
It
is well documented that pregnant women and newborns are more vulnerable to
infectious diseases than the overall population; nevertheless, vaccination
rates are often low in pregnant women, particularly in LMIC. This may suggest
why approximately 2.6 million children had died during the neonatal period
(0-27 days of age) and five and a half million children died before 5 years of
age worldwide in 2015 [24]. Among the serious infections during neonatal
period, influenza and pertussis are associated with significant morbidities and
mortalities [25]. Thus, maternal immunization during pregnancy against these
infectious diseases, have been implemented in resource limited settings [26].
Fortunately, the immune response against these infections, are T-cell dependent
with favorable avidity of these antibodies for trans-placental transfer.
However, secondary bacterial infections following infection with influenza are
not uncommon, accounting for a considerable number of morbidities and
mortalities in neonates and younger children. Most common bacteria known to
cause supra-infections are Streptococcus
pneumoniae and Hemophilus influenzae type
b [27], which are not isolated in most cases [28]. Therefore, it may be
reasonable to conclude that the incidence of supra-infection (bacterial
pneumonia) caused by these bacteria may be underestimated. The immune response
to these bacteria is T-cell independent, with poor avidity of these antibodies
for trans-placental transfer indeed. Therefore, concern for supra-infection
(bacterial pneumonia) with these encapsulated bacteria in patients following an
influenza infection should remain a priority.
It
is well known that neonates are unable to mount a fully protective immune
response to many pathogens, notably the intracellular pathogens and bacterial
polysaccharides [4,5]. As mentioned earlier, this is a T-cell independent
phenomenon in neonates [4,5]. This Th2-skewed response suppresses cytotoxic
T-lymphocytes and stimulates B lymphocytes to increase the production of
antibodies [25]. During this period, infants rely on these passively
transferred maternal antibodies for their protection. It has been suggested
that the levels of antibodies present in infants at birth correlate to the
levels of maternal antibodies. However, maternal specific antibody levels are
often suboptimal and therefore may not be sufficient to confer full protective
immunity to the infants or may protect them for only a short period of time
(approximately 6 months). Hypothetically, maternal immunization during
pregnancy can increase specific antibody concentrations to augment passive
transfer to their fetuses. This will reduce the window of vulnerability for the
infants until the appropriate time for infant vaccinations or the period of
greatest susceptibility has passed.
CONCLUSIONS
This
study has demonstrated significantly lower placental antibody transfer against SPN and Hib antigens in not only HIV-exposed, but also
HIV-unexposed-infants. Therefore, we strongly support maternal immunization
during pregnancy against vaccine preventable infections in vulnerable
populations, including HIV-infected women and pregnant women in LMIC, particularly
against the encapsulated bacteria. Additionally, we recommend pneumococcal
polysaccharide vaccine (PPV23) to mitigate invasive pneumococcal disease in
infants against a wider spectrum of serotypes, for immunizing mothers during
pregnancy. We further recommend that larger prospective studies should examine
specific serotypes of SPN not only
quantitatively, but also qualitatively by assessing functional abilities of the
opsonic antibodies through killing assays, and also examine duration of their immune protection.
This research is being supported by NIH grant numbers P20RR11792 and
RR03032.