374
March 1981 TheJournalofPEDIATRICS
Antibody response to group B streptococcus type 111 and AB blood group antigens induced by pneumococcal vaccine The effect of pneumococcal vaccination on antibodies to Streptococcus pneumoniae type 14, group B streptococcus type HI, and A B blood group antigens was studied in 40 vaccinated adults. Fourfold or greater increases in type-specific IgG antibody to Pn-14 were found in 26 of 40 vaccinees (mean increase 6.4-fold) and against GBS-II1 in 16 of the 40 (mean increase 2.9-fold) by an indirect immunofluorescence assay. However, only six of the 26 vaccinees with low levels (titers <<-20)of GBS-HI antibody in pre~eaccination sera developed titers >20 after vaccination. Thus, vaccination with polyvalent pneumococcal vaccine does not reliably induce high levels o f I F antibody to GBS-III. Fourfold or greater increases in IgG isohemagglutinins against blood group A cells were also found m 22 of 27 vaccinees (mean increase 4.5-fold) and against blood group B cells in nine of 34 (mean increase 1.7-fold) using the indirect anti-human globulin test. Chromatographic fractionation o f selected sera confirmed that lhe anti-A isohemagglutinins stimulated in group 0 subjects were o f the IgG class. Thus, pneumococcal vaccination during incompatible pregnancy could potentiate A 0 hemolytic disease o f the newborn infant.
Kenneth M. Boyer, M.D.,* Jutharat Theeravuthichai, M.D., Lawrence C. Vogel, M.D., Armando Orlina, M.D., and S a m u e l P. Gotoff, M . D . , C h i c a g o , Ill.
PASSlV~ IMMUNIZATION of the fetus by means of maternal vaccination has been proposed by Baker' as an approach to preventing perinatal group B streptococcal disease. Fischer et al ~ recently described an antigenic cross-reaction between the capsular polysaccharides of group B streptococcus serotype Ill and Streptococcus
From the Division o f Infectious Diseases, Department of Pediatrics, and the Blood Center, Michael Reese Research Foundation, Michael Reese Hospital and Medical Center, Pritzker School of Medicine, University of Chicago. Supported by grants RO-1-HD-09700 and RO-1-HD11576,from the National Institute o f Child Health and Human Development. Presented in part at the 19th lnterscience Conference on Antimicrobial Agents and Chemotherapy, Boston, Massachusetts, 1979. *Reprint address: Department of Pediatrics, Michael Reese Hospital and Medical Center, 29th St. and Ellis Ave., Chicago, IL 60616.
Vol. 98, No. 3, pp. 374-378
pneumoniae serotype 14, and demonstrated protection
against lethal GBS-III infections in neonatal rats by passive inoculation with hyperimmune Pn-14 rabbit antisera. Kasper et al 3 have clarified this important antigenic relationship by showing tha~ the type-specific antigen of Pn- 14 differs from "native" GBS-III polysaccharide, but is identical to a more degraded "core" polysaccharide of Abbreviations used Pn- 14: Streptococcus pneumoniae serotype 14 GBS-III: group B streptococcus serotype III IF: immunofluorescence DEAE: diethylaminoethyl RABA: radioactive antigen binding assay GBS-III from which sialic acid has been cleaved. Since the Pn-14 antigen is a constituent of the licensed polyvalent pneumococcal vaccine, 4 the question of whether pneumococcal vaccination will induce protective immunity against GBS-III in human subjects is of considerable interest.
0022-3476/81/030374+05500,50/0 9 1981 The C. V. Mosby Co.
Volume 98 Number 3
Antibody response to pneumococcal vaccine
Antigenic similarity between the Pn-14 polysaccharide and the ABO blood group antigens has been recognized since the late 1930s. The relationship was first noted when hemolytic reactions occurred in occasional patients with pneumonia during treatment with monospecific Pn-14 horse antisera? A disaccharide subunit of the Pn-14 capsular antigen has more recently been shown to serve as a biochemical precursor of the ABO blood group antigens, and accounts for the observed cross-reaction.6. r This immunologic relationship is of potential importance in the pregnant woman, since the administration of blood group antigens by vaccination or mismatched transfusion may result in hemolysis in the fetus. 8 To assess pneumococcal vaccine as a means of immunization against GBS-III and its effect on AB isohemagglutinins, we have measured antibodies against Pn-14, GBS-III, and the A and B blood group antigens in a group of adult vaccine recipients. MATERIALS
AND METHODS
The study population consisted of 40 adult patients receiving pneumococcal vaccine because of chronic pulmonary disease or congestive heart failure. Eightyeight percent of the patients were female. Their ages ranged from 28 to 80 with a median age of 64 years. Each participant received 0.5 ml intramuscularly of 14-valent pneumococcal vaccine (Pneumovax, Merck Sharp & Dohme, West Point, Pa.). Prevaccination sera were collected on the day of vaccination; postvaccination sera were collected at three to four weeks. Type-specific IgG antibodies against Pn-14 (strain CDC-PS14, kindly supplied by Dr. John Anhalt, Mayo Clinic Foundation, Rochester, Minn.) and GBS-III (strain Ill-Bell, a clinical isolate from a Chicago infant with neonatal meningitis) were measured by an indirect immunofluorescence technique developed in our laboratories.6 Fixed slides of log-phase whole bacteria grown in Todd-Hewitt broth and diluted to a concentration of 108-107 cfu/ml were used as test antigens. Fluoresceinated rabbit anti-human IgG was used as the secondary antibody. Sera were tested undiluted and in serial twofold dilutions from 1:5 to 1: 1,280, with the results read blindly. Antibody titers were defined as the greatest dilution of test serum yielding --> 2 + fluorescence. ABO grouping of vaccinees was carried out by standard forward and reverse techniques. 1~After screening of sera v~,ith human group O cells of defined antigenic composition (Spectrogen-Duo, Spectra Biologicals, Oxnard, Calif.) to exclude the presence of unexpected blood group antibodies, isohemagglutinin titers were measured in serial twofold saline dilutions from 1:2 to 1:8,192 against Rh-negative human A and B red blood cells (Gamma
3 75
Biologicals, Inc., Houston, Texas). Titers were measured by the indirect anti-human globulin technique ~~ with monospecific rabbit anti-human lgG (Ortho Diagnostics, Raritan, N.J.) after incubation at 37~ Pneumococcal vaccine was tested for the presence of blood group A- and B-like substances by competitive inhibition of specific hemagglutination of human A cells by anti-A serum and human B cells by anti-B serum using the standard technique for determination of salivary "secretor" status. TM In order to confirm that the anti-A isohemagglutinin response among group O vaccinees was IgG, we selected four paired sera with eightfold or greater increases in anti-A antibody for fractionation by ion-exchange chromatography. H Sera were fractionated by elution from diethylaminoethyl cellulose (Pharmacia Fine Chemicals, Uppsala, Sweden) using 0.018M sodium phosphate buffer at pH 7.6, and the eluates were concentrated to original serum volumes. Immunoglobulins were quantitated by immunodiffusion in agar plates containing goat antihuman IgG and IgM (Malloy Laboratories, Springfield, Va.). I A H G titers were then measured in the reconstituted fractions and compared with titers in corresponding whole sera. RESULTS All 40 vaccinees had detectable antibody to Pn-14 in prevaccination sera (Fig. 1) and 26 (65%) had _> fourfold increases in antibody titer after vaccination. The geometric mean antibody titer of prevaccination sera was 13.7; the geometric mean antibody titer of postvaccination sera was 88.3. The mean increase in antibody titer was 6.4fold. Prevaccination antibody titers against GBS-III consistently were lower than for Pn-14. Eleven vaccinees had undetectable antibody in undiluted sera. Fifteen had detectable antibody levels with titers of <20. Fourteen vaccinees had prevaccination antibody titers of >20. Fourfold or greater increases in IF antibody titer against GBS-III occurred in 16 vaccinees (40%). The geometric mean antibody titer of prevaccination sera was 6.4; the geometric mean antibody titer of postvaccination sera was 18.7. The mean increase in antibody titer was 2.9-fold. After vaccination, four of 11 initially seronegative recipients developed detectable IF antibody to GBS-III; none achieved titers of >20. Of the 15 recipients initially seropositive but with titers --<20, six developed titers of > 2 0 postvaccination. Prevaccination titers and antibody response to Pn-14 and GBS-III were not significantly affected by age or blood group. The study population was found to consist of 22
376
Boiler et aL
The Journal of Pediatrics March 1981
t~ao. 8192
640-
ooo 9
520-
oooeo
160 80-
9
9 9 oooo
eoooo
ooo
eeoc
4o2
"0~
fill
20-
$00$
go 9
eoooe
9
ooeoe
ee
oe
eeoc
oo41
4096. 20481024-
9
.o .o .9. . . . .
oll
lifO 9
512-
9 oeooo
o9
256-
cocoa 9
~
128-
9
9
9149
lille
9 4111
oeooo
64~b
I0.
ooo
9
9
eeoeee
me o9
32-
A& 04O
oeeo
9149
eeee 9149
9149
&AA 9
9
Ill 9
Nee
A 9149
16-
A&&
5-
8-
|mm
oo
42I
ooe
9
e ee eooc
oooo 9
;;0";;
....
~2( I P R E -V A C
POST-VAC
Pn - 14
PR E - VAC
oee
P O S T - VAC
GBS - I T /
Fig. 1. Antibody titers against S. pneumoniae type 14 and group B streptococcus type III measured by indirect immunofluorescence in prevaccination and postvaccination sera of 40 adult recipients of polyvalent pneumococcal vaccine. individuals of blood group 0, 12 of group A, five of group B, and one of group AB. Fourfold or greater increases in anti-A antibody occurred in 17 of 22 paired sera (77%) from vaccinees of blood group O and five of five paired sera (100%) from vaccinees of blood group B (Fig. 2). The overall geometric mean increase in anti-A titer was 4.5-fold. Fourfold or greater increases in anti-B antibody occurred in nine of 22 paired sera (41%) from vaccinees of blood group O and none of 12 paired sera from vacinees of blood group A. The overall geometric mean increase in titer was 1.7-fold. Pneumococcal vaccine was found to inhibit hemagglutination of human A cells by anti-A serum, but had no effect on hemagglutination of B ceils by anti-B serum, indicating that the vaccine contains antigenic material resembling human A substance. The first fractions of sera chromatographed on DEAE cellulose and reconstituted to original serum volumes contained undetectable IgM ( < 2 mg/dl) and IgG in concentrations ranging from 375 to 1,080 mg/dl. The fold increases and postvaccination anti-A titers in the fractionated sera were comparable to those measured in the whole serum pairs (Table), confirming that the anti-A antibody stimulated by vaccination was largely of the IgG class.
PRE-VAC POST-VAC ANTI-A
PRE -VAC POST- VAC ANTI- B
Fig. 2. Isohemagglutinin titers against human A and B red cells measured by indirect saline titration at 37~ with added anti-IgG serum in prevaccination and postvaccination sera of 40 adult recipients of polyvalent pneumococcal vaccine. 9 = Type O vaccinees. & = Type A vaccinees. 9 = Type B vaccinees. DISCUSSION Our findings indicate that polyvalent pneumococcal vaccine does not reliably induce type-specific IgG antibody against GBS-III in human subjects. Moreover, the relatively greater increase in isohemagglutinin antibodies against blood group A cells induced in vaccine recipients of blood group O suggests that pneumococcal vaccination during incompatible pregnancy could potentiate AO hemolytic disease of the newborn infant. The conclusions of the present study are necessarily influenced by the composition of the study population. Although the majority of participants were female, few were of childbearing age and all had underlying pulmonary or cardiovascular conditions that justified administration of vaccine. The distribution of prevaccination titers against GBS-III was higher in this population than we have observed in serologic surveys of younger, healthy women attending an inner-city obstetric screening clinic? 2 Although anti-GBS-III antibody response s to vaccination were not significantly affected by age, they were influenced by prevaccination antibody status. Of the 11 recipients with undetectable prevaccination antibody, only four developed detectable levels of antibody in postvaccination sera and none developed titers >20.
Volume 98 Number 3
A n t i b o d y response to pneumococcal vaccine
377
Table. Pre- and postvaccination immunoglobulin concentrations and indirect anti-human globulin anti-A isohemagglutinin titers in whole sera and the first fraction of DEAE cellulosefractionated sera of selected recipients of polyvalent pneumococcal vaccine IgG concentration*
IgM concentration
Vaccinee
Serum
Whole
DEAE~
Whole
A
Prevaccination Postvaccination Prevaccination Postvaccination Prevaccination Postvaccination Prevaccination Postvaccination
1,420 NDw 1,310 ND 1,280 ND !,420 ND
730 660 410 600 375 410 600 1,080
124 ND 86 ND 132 ND 94 ND
B C D
Anti-A titer~
D EA E
< < < < < < < <
2 2 2 2 2 2 2 2
Whole
DEAE
128 1,024 64 2,048 256 4,096 256 2,048
16 256 32 512 64 1,024 64 2,048
*Concentration in mg/dl. tTiter expressedas reciprocaldilution. ~DEAE = First fraction after DEAE chromatography. w = Not done. Thus, the higher proportion of subjects with undetectable prevaccination GBS-III antibody expected in a population of younger, healthy women would most probably have favored a poorer overall immunologic response than we observed. We are aware of two other limited studies of GBS-III antibody response to vaccination with polyvalent pneumococcal vaccine. Kasper et aF vaccinated 12 adult volunteers selected on the basis of low prevaccination antibody levels detected by radioimmunoassay against the "native" polysaccharide of GBS-III. Although eight of 12 vaccinees had -->1.4-fold increases in antibody directed against Pn-14 polysaccharide, only four developed a comparable response against the GBS-III "native" antigen. Only one postvaccination serum had opsonophagocytic activity against GBS-III whole organisms. In contrast, Lowell et a113have recently reported that administration of pneumococcal vaccine to a similar population of presumably susceptible adult volunteers uniformly resulted in increases in opsonophagocytic antibody against GBS-III. Although the antibody assay used in the present work differs from those utilized in either previous study, IF correlates with RABA for antibody to GBS-III. TM Human sera with GBS-III IF antibody titers _<20 fail to passively protect 12-day chick embryos against lethal challenge with GBS-III? Moreover, maternal sera of 40 infants with invasive GBS-III infection have all had titers -< 10 by this technique. 1~ Although a critical level of antibody is likely to be required for protection against neonatal infection, the exact line of demarcation between "protective" and "nonprotective" levels of antibody in human neonates by any of the available techniques remains uncertain. In contrast to the other available assays, however, IgG
antibody is specifically identified by IF and thus is directly relevant to the question of whether fetal immunity would be induced by maternal immunization. Since only six of 26 subjects had increases in IF antibody from --<20 to > 2 0 after vaccination, our results suggest that pneumococcal vaccine will not reliably induce high levels of IgG antibody to GBS-III, Since we did not vaccinate pregnant women in the present study, the clinical impact of stimulating maternal isohemagglutinins by pneumococcal vaccination remains theoretical. It is clear that AB hemagglutinins of IgG class are required to induce ABO hemolytic disease of the newborn infant. 1~ In previous studies, however, it has proven difficult to predict the degree of clinical erythroblastosis in the offspring of group O mothers on the basis of maternal antibody titers alone. Voak et al TM found that maternal incompatible IgG antibody titers of >256, as determined after 2-mercaptoethanol treatment of test sera, had a predictive value of 66% for development of AO hemolytic disease and of 90% for BO hemolytic disease. Comparable titers occurred in 13 and 7%, respectively, of maternal sera from AO and BO incompatible pregnancies in which hemolytic disease did not occur. Using the I A H G titration method, we found that anti-A titers of >256 occurred regularly in postvaccination sera of vaccinees of blood group O. Further, three of the four vaccinees in whom ion-exchange chromatography was used to detect isohemagglutinins of IgG class had postvaccination anti-A titers >256. Thus, it would seem reasonable to infer from our data that vaccination with polyvalent pneumococcal vaccine has the potential to accentuate hemolytic disease of the newborn in AO incompatible pregnancy. Since approximately 45% of the general population is of blood group O, 1~ this possible risk provides
378
Boyer et al.
support for the manufacturer's recommendation that polyvalent pneumococcal vaccine not b e administered to pregnant women. Immunoprophylaxis of perinatal GBS-III infections would appear to await development of an effective and safe polysaccharide vaccine based on a type-specific antigen of this organism. 17 Since polysaccharide antigens may be degraded to lower molecular weight products in vivo, TM the possibility that type-specific GBS candidate vaccines may stimulate isohemagglutinin antibodies should be investigated prior to their use in pregnant women.
The Journal of Pediatrics March 1981
5.
6.
7. 8.
ADDENDUM Recently, Baker et al reported that GBS-III antibody concentrations were not significantly increased after pneumococcal vaccination of adults with low prevaccination levels (_< 2 ~g/ml) by the R A B A test. Significant increases occurred only in recipients with moderateto-high prevaccination levels ( > 2 ~g/ml). (Baker CJ, Kasper DL, Edwards MS, and Schiffman G: Influence of preimmunization antibody levels on the specificity of the immune response to related polysaccharide antigens, N Engl J Med 303"173, 1980.) The authors acknowledge the assistance of Dr. Allen Sutow and Erva Cobb of the Mandel Clinic, Michael Reese Hospital, in patient recruitment and serum specimen collection. We appreciate the assistance of Marsha Durkin and Phyllis unger of the Michael Reese Hospital Blood Center in performance of isohemagglutinin determinations. We also thank Madeline Murphy for her excellent secretarial work. REFERENCES 1. Baker C J: Summary of the workshop on perinatal infections due to group B streptococcus, J Infect Dis 136"137, 1977. 2. Fischer GW, Lowell MH, Crumrine MH, and Bass JW: Demonstration of opsonic activity and in vivo protection against group B streptococci type III by Streptococcus pneumoniae type 14 antiserum, J Exp Med 148:776, 1978. 3. Kasper DL, Baker CJ, Baltimore RS, Crabb JH, Schiffman G, and Jennings HJ: Immunodeterminant specificity of human immunity to type III group B Streptococcus, J Exp Med 149:327, 1979. 4. Austrian R, Douglas RM, Schiffman G, Coetzee AM,
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Kornhoff H J, Hayden-Smith S, and Reid RDW: Prevention of pneumococcal pneumonia by vaccination, Trans Assoc Am Physicians 89"184, 1976. Finland M, and Curnen FC: Agglutinins for human erythrocytes in type XIV anti-pneumococcic horse serums, Science 87:417, 1938. Beeson PB, and Goebel WF: The immunological relationship of the capsular polysaccharide of type XIV pneumococcus to the blood group A specific substance, J Exp Med 70:239, 1939. Watkins WM: Blood-group substances, Science 152:172, 1966. Abelson NM, and Rawson AJ: Studies of blood group antibodies. V. Fractionation of examples of anti-B, anti-A, anti-M, anti-P, anti-JK a, anti-LCL anti-D, anti-CD, anti-K~ anti-FyL anti-S, and anti-Good, Transfusion 1:116, 1961. Vogel LC, Kretschmer RR, Boyer KM, Padnos DM, Gadzala CA, and Gotoff SP: Human immunity to group B streptococci measured by indirect immunofluorescence: correlation with chick embryo protection, J Infect Dis 140;682, 1979. American Association of Blood Banks: Technical Manual of the American Association of Blood Banks, Washington, DC, 1977. Abelson NM, and Rawson AJ: Studies of blood group antibodies. I. Fraetionation of anti-A and anti-B isohemagglutinins by anion-cation cellulose exchange chromatography, J Immunol 82:435, 1959. Vogel LC, Boyer KM, Gadzala CA, and Gotoff SP: Prevalence of type-specific group B streptococcal antibody in pregnant women, J PEDIATR96:1047, 1980. Lowell GH, Fischer GW, Wilson SR, and Crumrine MH: Serum from adults immunized with pneumococcal vaccine is opsonic in vitro and protective in vivo for group B type III streptococci, Pediatr Res 13:463, 1979. Vogel LC, Boyer KM, Gotoff SP, Kasper DL, and Baker CJ: Comparison of assays for antibody to group B streptococcus, type III, J Infect Dis 141:530, 1980. Kochwa S, Rosenfield RE, Tallal L, and Wasserman LR: Isoagglutinins associated with ABO erythroblastosis, J Clin Invest 40:874, 1961. Voak D, and Bowley CC: A detailed serologic study on the prediction and diagnosis of ABO hemolytic disease of the newborn, Vox Sang 17:321, 1969. Baker CJ, Edwards MS, and Kasper DL: Immunogenicity of polysaccharides from type III, group B streptococcus, J Clin Invest 61:1107, 1978. Coonrod JD: Evidence of breakdown of ~pneumococcal polysaccharides in vivo, Proc Soc Exp Biol Med 162:249, 1979.