FETAL
AND
MEDICINE
NEONATAL RichardE. Behrman,Editor
Prevalence of type-specific group B streptococcal antibody in pregnant women Immunoglobulin G antibody against the four major serotypes o f group B streptococcus was measured by indirect immunofluoreseence in the sera of 200 conseeutive pregnant women seen in the obstetric screening clinic o f an 'urban teaching hospital Antibody was detectable in 26% of undiluted sera against serotype Ia, 52% against serotype Ib, 82% against serotype II, and 45% against serotype I l L Only had antibody against all four GBS types. When serotype-speeific antibody prevalences in 108 women with GBS vaginal eolonization were compared with prevalences in noncolonized women, only women colonized with GBS type Ia were more likely to have antibody against Ia than noncolonized women. Antibody prevalences in sera from 54 mothers whose infants developed invasive GBS disease were significantly lower than those in colonized or noneolonized women. Since low titers o f I F antibody to GBS 111 were present in some sera from mothers of infected infants, the data were analyzed based on I F antibody titers associated with passive protection in a chick embryo model o f GBS septicemia. None o f the sera from mothers of infected infants had antibody levels associated with chick embryo protection. Less than 10% o f women had titers associated with chick embryo protection. These data suggest that the majority of pregnant women lack immunity to GBS, regardless o f colonization status.
Lawrence C. Vogel, M.D., Kenneth M. Boyer, M.D., Cecile A. Gadzala, R.N., and Samuel P. Gotoff, M.D.,* Chicago, Ill.
GROUP B STREPTOCOCCI are a major cause of bacterial infections in the perinatal period? Subdivision of these organisms by type is based on specific polysaccharide and protein antigens. Serotype III is isolated from most cases of neonatal GBS meningitis; serotypes Ia, Ib, lc, and II are important causes of septicemia and focal infections in neonates. Immunity to GBS appears to be mediated by antibodydependent phagocytosis, as demonstrated by animal protection tests '-'-~and opsonophagocytic assays?. 5-7 A quantitative test for antibody to the type-specific native antigen of GBS III has shown that low concentrations o f antibody
From the Department o f Pediatrics, Michael Reese Hospital and Medical Center, Pritzker School of Medieine, University o f Chicago. Supported by Public Health Service grants RO-I-HD09700 and RO-1-HD-11576 from the National Institute o f ChiM Health and Human Development. *Reprint address: Michael Reese Hospital and Medical Center, 29th St. and Ellis Ave., Chicago, IL 60616.
0022-3476/80/061047+05500.50/0 9 1980 The C. V. Mosby Co.
in maternal sera are associated with susceptibility to GBS infection in the newborn i n f a n t ? 9 We have recently developed an indirect immunofluorescent antibody test for type-specific IgG antibody against the five serotypes of GBS? 9 In that study we showed that a minimum I F antibody titer was necessary to protect chick embryos against lethal infections with Abbreviations used GBS: group B-streptococcus IF: immunofluorescent MRHMC: MichaelReese Hospital and Medical Center RABA: radioactiveantigen-binding assay
/
homologous GBS serotypes. Moreover, sera from mothers &infected infants uniformly had IF antibody titers below levels associated with chick embryo protection, However, the concentration of antibody providing protection in man has not been defined. Limited data are available on the prevalence of typespecific antibody in human populations, and these are primarily based on qualitative assays. Hemming et aP The Journal of P E D l A T R I C S Vol. 96, No. 6, pp. 1047-1051
1047
10 4 8
Vogel et al.
The Journal of Pediatrics June 1980
Io
8 o/=
,~-,o5%4.5
r[
%
- ,.
Tr I I
O,fl%
3.5*,
6.5% I KS'
5*/*
Fig. 1. Group B streptococcal antibody titers in 200 pregnant women. Numbers within shaded areas indicate titers.
noted serum opsonic activity for types Ia, II, and 1II in 12%, 40%, and 74% of unselected adult sera, respectively. Stewardson-Krieger et al'~ found that 25% of sera from adult women had opsonic activity and protected mice against lethal intraperitoneal challenge by GBS type Ia. Using a quantitative radioactive antigen binding assay, Beachler et a111 reported that pregnant women colonized with GBS type III had significantly higher concentrations of antibody to a native type IlI antigen than women who were not colonized. The present study attempts to determine the prevalence of antibody to GBS in a population of pregnant women. MATERIALS
AND M E T H O D S
Study populations. Sera were obtained from 200 consecutive pregnant women attending an obstetrical screening clinic at Michael Reese Hospital and Medical Center. These subjects were predominantly black and of lower socioeconomic status. The mean age was 22 years and the mean number of previous pregnancies was 2.7. Sera were obtained during the first trimester in 35% of the women, 59% during the second trimester, and 6% during the third trimester. From 143 of these 200 women, vaginal cultures were obtained at the time of serum collection; a selective broth medium w~s used? 2 Fourteen women were colo-
nized with GBS. Sera were also obtained from 84 pregnant women who had GBS isolated from vaginal cultures at the time of delivery at MRHMC, .and from ten additional colonized women attending the obstetrical screening clinic. In the 108 women colonized with GBS, the mean age was 25.1 years and mean number of pregnancies was 2.8. Twenty-two strains were serotype Ia, 28 serotype.Ib, 22 serotype II, and 36 serotype III by capillary precipitation using monospecific rabbit antisera. 13 None of the infants born to these colonized women developed invasive GBS infections. A third study population consisted of 54 mothers of neonates with invasive GBS infection. These women had a mean age of 24 years and mean number of pregnancies of 2.8. Sera were obtained within five days of delivery. Blood or cerebrospinal fluid isolates from the infected infants were 2 serotype Ia, 8 serotype Ib, 4 serotype II, and 40 serotype III. Indirect immunofluoreseent assay. As previously described, the IF assay for measuring type-specific lgG antibody utilizes acetone-fixed whole bacteria on slides and monospecific fluorescein-conjugated rabbit antihuman IgG. TM Specificity was demonstrated bY absorption experiments with whole bacteria and with type Ia polysac-
Volume 96 Number 6
Prevalence o f GBS antibody in pregnant women
Ia
160-
OO
80-
O
40--
E3E]
O
20-
]
o
IO-
E~
O
5-
Zb
Z
10 4 9
I~
O
[]
ooo
KID
ooo
88~ 88
OO
ooo
5~E] ooo
[]
NC
C
I
NC
C
I
NC
C
I
NC
C
Z
Fig. 2. IF antibody in colonized and noncolonized pregnant women, and mothers of infected infants. Antibody titers in noncolonized pregnant women (n) to types Ia, Ib, I1, and III-GBS, colonized women (o) to the homologous colonizing type, and mothers of infected infants (A) to the homologous infecting GBS type.
charide antigen. Human sera were tested undiluted and in serial twofold dilutions, beginning at 1:5. Antibody titers measured by IF that are associated with protection of chick embryos ~ against lethal intravenous challenge by GBS types Ia, Ib, II, and III are 1:20, 1:40, 1:80, and 1:40, respectively.TM Antibody to type Ic was not determined, since IF antibody titers and chick embryo protection are nearly identical for types Ia and Ic. TM RESULTS
The prevalence of detectable IgG type-specific GBS antibody measured by IF in sera from 200 consecutively sampled pregnant women is shown in Fig. 1. Antibody against types Ia, Ib, II, and III was present in sera from 53,
103, 164, and 89 women, respectively. Only 19 subjects had antibody against all four GBS types tested. The distribution of antibody titers in colonized and noncolonized women and in mothers of infected infants is shown Fig. 2. Of the 129 women without vaginal colonization, 26% had antibody to type la, 54% to Ib, 82% to 11, and 47% to lII. Among the 108 colonized women, antibody was detected against the homologous colonizing type in 59%, 57%, 96%, and 50% of sera for GBS types Ia, Ib, II, and III, respectively. A significant difference in antibody prevalence between colonized and noncolonized women was found only in women colonized with GBS type Ia ('X2 = 10.0; P < 0,01). Neither of the two mothers of infants with type Ia
10 5 0
Vogel et al.
infections had antibody, to Ia present in undiluted serum. Only one of eight sera from mothers of infants infected with type lb and none of four sera from mothers with infants infected with type II had antibody against the homologous infecting GBS type detected in undiluted sera. Eight of the 40 sera from mothers of infants with GBS type III infections had detectable type-specific antibody. The prevalence of detectable antibody in this latter group was significantly lower than the prevalence of antibody in noncolonized w o m e n (X2 = 8.9. P < 0.01) or in w o m e n colonized with type IIl ( ) 2 = 7.6. P < 0.01). Next we analyzed the data in order to determine the prevalence of IF titers associated with passive protection of chick embryos against lethal challenges with GBS. C o m p a r e d to prevalences o f detectable antibody in undiluted sera, prevalences of protective antibody titers were strikingly low in all of the groups of w o m e n studied. A m o n g the 200 consecutively sampled pregnant women. protective antibody titers were found to types Ia. Ib, II, and III in 17. 11. 4. and 11 sera. respectively. N o n e of these 200 sera had chick embryo protective antibody titers against all four GBS types tested. Protective antibody titers in sera from noncolonized pregnant women were found in 7%. 4%. 1%, and 5% for types la. l b . I1, and III, respectively. A m o n g colonized women, protective antibody titers to the homologous colonizing type were found in 23%. 11%, 14%. and 14% of sera for types Ia. lb. II. and III. respectively. The increased rates in colonized women were significant for type Ia (X ~ = 3.9. P < 0 . 0 5 ) , type II (X2 = 12.13. P < 0.001), and type III (X2 = 3.9. P < 0.05). N o n e of the sera from the 54 mothers with infected infants had protective antibody titers to the homologous infecting GBS type. Prevalence of protective antibody titers in mothers of infants infected with type Ill GBS was less than in w o m e n asymptomatically colonized with type IIl (X~- = 5.9. P < 0.05), However. there was no difference between protective antibody titers in mothers of infants with type III GBS infections and those in noncolonized women. DISCUSSION Humoral immunity to an infectious disease is ultimately characterized by the concentration and specificity of protective antibody. Such information is difficult to obtain directly in man for an infection with an attack rate of 3 to 4/1.000. The present study provides data on the prevalence of IgG antibody measured by IF against the four major serotypes of GBS in an inner-city population of pregnant women. Protective levels of type-specific antibody for h u m a n beings could not be defined in this
The Journal of Pediatrics June 1980
study because of insufficient sera from mothers of infected newborn infants. However. it is apparent that low titers of IF antibody to GBS III in maternal sera failed to protect eight newborn infants from GBS infections. Therefore. we analyzed our data using IF antibody titers associated with chick embryo protection, which may ultimately approximate protective titers in man. Less than 9% of sera from 200 pregnant w o m e n had antibody titers by IF associated with protection of chick embryos for any one of the four GBS type studied; none had chickprotective activity agamst all four serotypes. Thus. if these titers ultimately are shown to approximate titers necessary for protection in man, most newborn infants are susceptible to GBS infections. Similar conclusions have been drawn for susceptibility to GBS Ill. ~ 14 W o m e n colonized with type Ia GBS were more likely to have antibody and to have antibody titers associated with chick embryo protection than women who were not colonized. In addition, w o m e n colonized with type II or type IIl GBS had a higher prevalence of chick embryo protective antibody titers compared to noncolonized women. Beachler et a111 found that the concentration of antibody to type III GBS, measured by RABA. was significantly higher in 13 women colonized with type III bacteria than in w o m e n without GBS colonization. Our data suggest that vaginal colonization does not greatly increase the likelihood of detecting type-specific antibody by 1F in undiluted sera. but may affect the titer. The latter point is consistent with Beachler's study. Based on continuing surveillance since 1973. the incidence of early-onset GBS sepsis at M R H M C has been 1.8/1.000 live births. 15 while the annual prevalences of GBS vaginal colonization have been in the rang e of 15 to 20% (K.M. Boyer, unpublished data). This ratio of 100:1 between rates of maternal colonization and neonatal early-onset invasive disease has been noted previously Our present data suggest that susceptibility to GBS infection is common to most newborn infants. Thus. the absence of humoral immunity is not sufficient to account for the wide discrepancy between rates o f asymptomatie colonization and invaslve neonatal infection Rather. the critical importance of other host and microbial factors in the pathogenesis of neonatal GBS sepsis ts emphasized. REFERENCES
1
Baker CJ: Summary of the workshop on perinatal infections due to group B streptococcus. J Infect Dis 136:137. 1977. 2. Lancefield RC. McCarthy M. and Everly WN: Multiple mouse-protection antibodies directed against group B streptococci: special reference to antibodies effective against protein antigens, J Exp Med 142:165. 1975. 3. Stewardson-Krieger PB. Albrandt K. Nevin T. Kretschmer RR. and Gotoff SP: Perinatal immunity to group B B-
Volume 96 Number 6
4.
5.
6.
7.
8.
9.
hemolytic streptococcus type la, J Infect Dis 136:649, 1977. Tieffenberg J, Vogel L, Kretschmer R R, Padnos D, and Gotoff SP: A chick embryo model for type III group B beta hemolytic streptococcal septicemia, Infect Immun 19:481, 1978. Hemming VG, Hall RT, Rhodes PG, Shigeoka AO, and Hill HR: Assessment of group B streptococcal opsonins in human and rabbit serum by neutrophil chemiluminescence, J Cliri Invest 58:1379, 1976. Anthony BF: Immunity to the group B streptococci: interaction of serum and macrophages with types Ia, Ib, and Ic, J Exp Med 143:1186, 1976. Baltimore RC, Kasper DL, Baker CJ, and Goroff DK: Antigenic specificity of opsonophagocytic antibodies in rabbit antisera tO group B streptococci, J lmmunol !18:673, 1977. Baker CJ, and Kasper DL: Correlation of maternal antibody deficiency with susceptibility to neonatal group B streptococcal infection, N Engl J Med 294:752, 1976. Baker CJ, Kasper DL, Tager iB, Paredes A, Alpert S, McCormack WM, and Goroff D: Quantitative determination of antibody to capsular polysaccharide in infection with type iii strains of group B streptococcus, J Clin Invest 59:810. 1977.
Prevalence o f GBS antibody in pregnant women
10 5 1
10. Vogel LC, Kretschmer RR, Boyer KM, Padnos DM, Gadzala CA; and Gotoff SP: Human immunity to group B streptococci measured by indirect immunofluorescence: correlatiofi with chick embryo protection, J Infect Dis 140:682, 1979. 11. Beachler CW, Baker CJ, Kasper DL, Fleming DK, Webb BJ, and Yow MD: Grou p B streptococcal colonization and antibody status in lower socioeconomic parturient women, Am J Obstet Gynecol 133:171, 1979. 12. Baker CJ, Clark DJ, and Barrett FF: Selective broth medium for isolation Of group B streptococci, Appl Microbiol 26:884, 1973. 13. Facklam RR: Streptococci, in Lennette Ett, Spaulding Eli, and Truant JP, editors: Manual of clinical microbiology, Washington, DC, 1974, American Society for Microbiology, p 96. 14. Hamerschlag MR, Baker CJ, Alpert S, Kasper DL, Rosner !, Thurston P, Webb BJ, and McCormack WM: Colonization with group B streptococci in girls under 16 years of age, Pediatrics 60:473, 1977. 15. Stewardson-Krieger PB, and Gotoff SP: Risk factors in early-onset neonatal group B streptococcal infections, Infection 6:50, 1978.