- Email: [email protected]
Preimmunization and postimmunization pneumococcal antibody titers in children with recurrent infections
Preimmunization and postimmunization pneumococcal antibody titers in children with recurrent infections Hugo Hidalgo, MD*; Cleveland Moore, MD*; Lily E Leiva, PhD*; and Ricardo U Sorensen, MD*
Background: Patients with recurrent infections and normal IgG levels may have an abnormal response to pneumococcal polysaccharides. The ability to develop antibodies against different pneumococcal polysaccharides develops gradually in the first years of life, but the sequence of development and the influence of preexisting antibody titers has not been defined. Methods: Preimmunization and postimmunization IgG antibody titers against pneumococcal serotypes 3, 7F, 9N, and 14 were evaluated in a population of 100 1to 18-year-old children referred to a pediatric allergy-immunology clinic because of recurrent respiratory infections. None of the patients had a known immunodeficiency syndrome; all had normal total IgG levels. Postimmunization antibody levels were obtained 4 to 6 weeks after immunization. Patients ⱕ5 years of age who failed to develop antibody levels above 200 ng Ab N/mL against any serotype and older patients who failed to develop these levels against a second serotype in addition to serotype 3 were considered for IgG replacement therapy. Results: Prior to immunization, 50% of 51 patients did not have protective antibody levels against any of the serotypes tested. Immunization induced a high response to serotype 3 in all age groups, but responses to serotypes 7F and 14 increased with age. Five of 78 patients (6.4%) failed to develop protective antibody levels against any serotype tested. Three of these patients had clinical criteria that justified the use of IgG replacement therapy; all improved. Three patients were re-immunized 1 to 2 years after the first immunization and all developed protective levels of antibodies against serotype 3 after the second immunization. Conclusion: We conclude that, although measurement of antibody levels against pneumococcal serotype 3 allows a good differentiation of patients who are able to develop anti-polysaccharide antibodies from those who are not, further studies of the development of specific antibodies against other vaccine serotypes in normal populations of different ages are needed to define a normal response to pneumococcal polysaccharides. Ann Allergy Asthma Immunol 1996;76:341– 6.
INTRODUCTION Pneumococcal immunization and evaluation of the resulting antibody response against pneumococcal serotypes has been used increasingly to enhance protection against pneumococcal infection and to identify deficiencies in the devel-
* Division of Allergy and Immunology, Department of Pediatrics, Louisiana State University Medical Center, New Orleans, Louisiana. Received for publication July 14, 1995. Accepted for publication in revised form September 6, 1995.
VOLUME 76, APRIL, 1996
opment of specific antibodies against polysaccharide antigens. Absent or low specific antibody responses to polysaccharides are normally expected in children under 2 years of age. When the lack of an adequate response to pneumococcal polysaccharides persists past 2 years of age, it defines a deficiency of specific antibodies that may be present in patients with normal IgM, IgG, and IgA levels.1 The immunogenicity, however, of the various serotypes included in the available 23-valent vaccines varies and the ability to mount adult-like responses does not de-
velop uniformly for all serotypes at 2 years of age. Several immunization trials have clearly shown a stepwise, age-dependent development of the ability to produce antibodies against the various polysaccharide serotypes included in the vaccines used.2–5 In these early studies, the antibody response was determined with a radioimmunoassay which did not discriminate between IgM and IgG responses.6 Some of the observed responses tended to wane rapidly,2 suggesting that they were short-lived IgM responses. Long-term protective immunity against pneumococcal infection has been associated with the presence of serotypespecific IgG antibodies. Using a method capable of differentiating IgM from IgG responses, three types of responses to different serotypes were identified in young children: high IgM and IgG antibodies against serotype 3; intermediate IgM titers and low IgG titers against serotype 23; and little or no IgM or IgG response to serotypes 6, 18, and 19.7 The sequence of development of IgG antibodies against different serotypes of pneumococci at different ages and the effect of preexisting IgG antibodies to pneumococcal serotypes have not been defined well to date. Our objective was to evaluate the presence of preimmunization and postimmunization IgG antibodies against different pneumococcal serotypes in pediatric patients with a history of recurrent respiratory infections and normal levels of IgG. PATIENTS AND METHODS Patient Population We analyzed retrospectively 100 patients between 1 and 18 years of age referred to our allergy-immunology clinic between 1991 and 1994 with
341
Table 1. Pneumococcal Antibody Titers in Unimmunized and Immunized Children* Age Group
Unimmunized Serotype 3 7F 9N 14 P Value‡ Immunized Serotype 3 7F 9N 14 P Value‡
P Value†
1 (1–5 yr)
2 (6–10 yr)
3 (11–18 yr)
n ⫽ 23
n ⫽ 16
n ⫽ 12
102 ⫾ 4.19 60 ⫾ 2.61 35 ⫾ 3.03 71 ⫾ 2.78 .030650 n ⫽ 48
412 ⫾ 5.47 100 ⫾ 2.61 82 ⫾ 2.65 128 ⫾ 3.02 .006529 n ⫽ 19
715 ⫾ 4.68 122 ⫾ 1.93 84 ⫾ 2.09 225 ⫾ 2.27 .001601 n ⫽ 11
.001981 .052826 .007572 .004271
1764 ⫾ 3.35 264 ⫾ 3.35 241 ⫾ 3.59 97 ⫾ 2.15 ⬍.000001
1741 ⫾ 4.35 361 ⫾ 4.30 365 ⫾ 3.73 219 ⫾ 3.49 .000028
1895 ⫾ 2.59 484 ⫾ 5.31 208 ⫾ 3.90 284 ⫾ 4.23 .005877
.922600 .300863 .312861 .004497
* Serotype-specific IgG titers measured by enzyme-linked immunosorbent assay expressed as ng Ab N/mL. Values are expressed as mean ⫾ SD. † Comparison among age groups by Kruskal-Wallis One Way ANOVA. ‡ Comparison among serotypes by Kruskal-Wallis One Way ANOVA.
histories of recurrent respiratory infections. Only patients with normal levels of IgG and without known primary immunodeficiency syndromes were included in this analysis. Fiftytwo patients were 1 to 5 years old (group 1). Of these patients, 48 were over 2 years of age and 4 were between 18 and 24 months old. Twenty-eight were 6 to 10 years old (group 2) and 20 were 11 to 18 years old (group 3). There was a slight predominance of males in all age groups: 64%, 54%, and 55%, respectively. Thirty-five percent of the patients were African-American. Immunization with Pneumococcal Vaccine Immunization was performed with a 23-valent pneumococcal vaccine (PnuImmune, Lederle, Wayne, NJ or PneumovaxR, Merck, Sharp and Dome, West Point, PA). None of the patients had a febrile illness at the time of immunization. Immunologic Evaluation To decrease the likelihood of vaccine reactions, preimmunization antibody titers were obtained in patients with very frequent bacterial infections and in older patients who may have developed antibodies in response to
342
bacterial infections.2 Frequently, however, the vaccine was given at the time of the first visit and the immunological evaluation was performed 4 to 8 weeks after immunization. Immunoglobulins G, M, and A were determined by nephelometry. Age-adjusted normal immunoglobulin values were used to select patients with normal IgG levels and to determine whether IgM and IgA levels were within the normal range. Pneumococcal polysaccharide antibodies against serotypes 3, 7F, 9N, and 14 were determined commercially by an enzyme immunoassay (Specialty Laboratories, Los Angeles, CA). Only IgG antibodies are measured in this assay which uses an isotype-specific conjugate. Antibody levels are derived from comparison of the immunoassay signal to the standard curve generated for each pneumococcal serotype. The standard curve is calibrated against concentrations of antibody in nanograms of antibody nitrogen per mL (ng Ab N/mL). Maximum variation for repeat analyses of the same serum sample with this method is 15%. Levels above 200 ng Ab N/mL were considered protective.4,8 All four serotypes are included in the polyvalent pneumococcal vaccine given to these patients.
Criteria for IgG Replacement Therapy We used a modification of the criteria defined by Knutsen based on severity of infections and antibody response to pneumococcal immunization.1 IgG replacement therapy was considered when there was evidence of recurrent respiratory infections after adequate treatment of other risk factors, eg, gastroesophageal reflux, allergic rhinitis, asthma, or anatomical abnormalities of the airways. The infections considered were otitis media and/or sinusitis recurring at least 6 times/yr or pneumonia 3 times/yr. In addition, patients ⱕ5 years of age must have failed to develop protective antibody levels against any serotype, and older patients must have failed to develop protective antibody levels against a second serotype in addition to serotype 3. In some patients, IgG replacement therapy was discontinued after 1 to 2 years. These patients were immunized with a second dose of pneumococcal vaccine and retested 4 to 6 weeks later. Data Collection and Statistical Analysis The Epi-Info program was used to create a patient database and perform data analysis. All antibody titer data were
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
analyzed following logarithmic transformation in order to decrease the influence of a few extreme measurements in these positively skewed distributions. Geometric means of the data were obtained and comparisons of preimmunization and postimmunization antibody titers were made by the paired t test. Comparisons between age groups were made using the unpaired t test or analysis of variance for comparison of three or more age groups.9 RESULTS By definition, all patients had normal IgG levels. No IgM-deficient patients were found. Four patients had IgA levels below 2 SD for their age group: three patients in group 1 had IgA values of 8, 18, and 19 mg/dL (normal ⫽ 20 – 84 mg/dL); and one 16-year-old patient had an IgA level of 11 mg/dL. Preimmunization antibody titers were measured in 51 patients: 23 in group 1, 16 in group 2, and 12 in group 3. The mean antibody titers against each one of the four serotypes increased significantly with age (Table 1 and Fig 1, panels A–C). Twenty-six patients did not have protective antibody levels against any serotype tested. Of the remaining 25 patients, 14 had antibodies to serotype 3 only, 1 to serotype 7F only, 2 to serotype 14 only and 8 patients had protective antibodies against two or more serotypes. Pneumococcal immunization was well tolerated by most patients. Mild local and occasional febrile reactions were reported by some patients over 5 years of age. Postimmunization antibody titers were measured in 78 patients: 48 in group 1, 19 in group 2, and 11 in group 3. The mean titers of antibodies against the four serotypes were significantly different within each age group (Table 1). Serotype 3 induced a similarly high response in all age groups. Although the titers to serotype 7F and 14 increased with age, the differences among age groups were significant only for serotype 14. The percentages of patients who developed protective antibody titers against the four serotypes also in-
VOLUME 76, APRIL, 1996
Figure 1. Preimmunization (panels A–C) and postimmunization (panels D–E) pneumococcal antibody titers in patients in three different age groups. More patients in each age group had postimmunization titers than preimmunization antibody titers.
creased with age (Fig 1, panels D–F). Only three patients in the first group and two in the second group failed to develop protective antibody levels against serotype 3. The three patients in the first group under 2 years of age who had postimmunization titers responded to one or more serotypes and did not differ from the patients between 2 and 5 years of age. Lack of development of protective antibody levels to serotypes 7F, 9N, and 14 was common in patients younger than 11 years of age. Low antibodies to sero-
types 9N and 14 were also frequent in patients over 10 years of age. Five of 78 patients (6.4%) failed to develop protective antibody levels against any serotype tested. Eight patients (10%) of all age groups developed protective antibody levels to serotype 3 only. Preimmunization and postimmunization titers were measured in 29 patients: 19 in group 1 and 10 in groups 2 and 3. There was a significant increase in postimmunization titers to serotypes 3, 7F, and 9N in group 1 and to all serotypes in groups 2 and 3 (Table
343
Table 2. Preimmunization and Postimmunization Pneumococcal Antibody Titers* Age Groups, yr 1–5 (n ⫽ 19) Serotype 3 7F 9N 14 6–18 (n ⫽ 10) Serotype 3 7F 9N 14
Preimmunization Titers
Postimmunization Titers
P Value†
87 ⫾ 4.04 57 ⫾ 2.82 32 ⫾ 3.24 68 ⫾ 2.98
1313 ⫾ 5.04 152 ⫾ 4.57 135 ⫾ 4.21 94 ⫾ 3.53
.00160 .00880 .01290 .18160
348 ⫾ 5.64 77 ⫾ 2.33 65 ⫾ 2.35 143 ⫾ 3.67
1753 ⫾ 6.18 521 ⫾ 4.51 405 ⫾ 4.95 238 ⫾ 4.01
.01090 .01360 .02720 .08520
* Serotype-specific IgG titers measured by enzyme-linked immunosorbent assay and expressed as ng Ab N/mL. Values are expressed as mean ⫾ SD. † Comparison of preimmunization and postimmunization values by paired t test.
2). Four patients failed to have an increase in titer into the protective range against serotype 3, 11 against serotype 7F, 13 against 9N, and 20 against serotype 14 (Fig 2). Several patients had twofold or higher increases in titer against one or more serotypes without reaching the arbitrarily defined protective level of 200 ng Ab N/mL. High preimmunization titers against any serotype tended not to increase further after immunization, and a decrease in titer against serotype 3 was observed in one patient (Fig 2, panel A). The clinical characteristics of five patients who failed to develop protective antibody levels against any of the serotypes tested are listed in Table 3. Three of these patients had clinical criteria that justified the use of IgG replacement therapy. All showed significant improvement. IgG replacement could be discontinued after 6 to 12 months in two patients. These patients were re-immunized 4 to 6 months later and both developed protective levels of antibodies against serotype 3. Two additional patients, aged 7 and 15 years, had protective antibody levels to serotype 3 only and were judged to require IgG replacement therapy. Both patients had significant clinical improvement on this therapy, but symptoms recurred when IgG replacement was discontinued and treatment was reinstated.
344
None of the patients 1 to 5 years old who developed protective levels to serotype 3 and none of the older patients who developed protective levels to at least one additional serotype had a
clinical course that would have justified IgG replacement treatment. All four patients with low IgA levels developed protective antibody levels against one or more serotypes after immunization, and none had clinical criteria justifying IgG replacement therapy. DISCUSSION In our patient population evaluated because of recurrent respiratory infections, 50% had preimmunization protective antibody levels against the pneumococcal serotypes tested. The highest frequency of antibodies was found against serotype 3. This may reflect both the high immunogenicity of its polysaccharide2–5 and the frequency of infections with this serotype. While serotype 3 is an infrequent cause of invasive Streptococcus pneumoniae infections in children,10 it is a frequent isolate from the respiratory tract.11,12 The low percentage of pa-
Figure 2. Preimmunization and postimmunization titers in 29 patients: open circles ⫽ 1- to 5-year-old patients; solid triangles ⫽ patients ⱖ6 years of age; * ⫽ patients out of range of graph.
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
Table 3. Patients with No Response to Pneumococcal Immunization* Patient
Age
Clinical Presentation
1
2y
2
2y
Recurrent otitis media Recurrent sinusitis Allergic rhinitis Bronchial asthma Recurrent purulent otitis media, refractory to antibiotic treatment
3
4y
4
7y
5
8y
Recurrent otitis media Recurrent pneumonia Atopic dermatitis Allergic rhinitis Bronchial asthma Recurrent otitis media Allergic rhinitis Cranio-velo-facial syndrome Recurrent sinusitis Recurrent pneumonia Severe atopic dermatitis Allergic rhinitis Bronchial asthma
Rx & Clinical Course Antibiotic therapy only Lost to follow up
IgG (IVIG) replacement for 6 months with clinical improvement Reimmunization 6 months later with response to serotype 3 only Sustained clinical improvement afterwards IgG (IVIG) replacement for 1 year with clinical improvement Reimmunization 1 year later with response to serotype 3 only Sustained clinical improvement afterwards Antibiotic therapy only Reimmunization 1 year later with response to serotype 3 only Chronic IgG (IVIG) replacement with significant clinical improvement of infections & atopic dermatitis
* These patients failed to develop antibody levels ⬎200 ng Ab N/mL against any of the four serotypes tested.
tients who have developed protective antibody levels in response to clinical and subclinical infections appears to justify the use of pneumococcal vaccines in this population and indicates that the risk of reactions due to high preexisting antibody titers against vaccine serotypes is low. Our results clearly document that there is a stepwise development of the ability to develop IgG antibodies to various serotypes beyond the second year of life. Comparison of our pediatric population with asthmatic adults studied with the same methodology in the same area9 revealed that the mean postimmunization titers to serotypes 9N and 14 are significantly higher in 27- to 65-year-old adults than in all age groups included in this report (Table 1): 585 ⫾ 2.78 ng Ab N/mL for serotype 9N and 564 ⫾ 1.2 ng Ab N/mL for serotype 14. Since age is clearly one of the most important determinants of the specific antibody response to pneumococcal polysaccharides, pediatric patients who fail to develop protective antibody levels to pneumococcal immunization
VOLUME 76, APRIL, 1996
in the first years of life may be reimmunized at a later age. In the few patients in our study in whom re-immunization was performed, we always observed a higher response to at least one serotype after the second immunization. High levels of preexisting antibodies tend to decrease the increment in antibody titers after vaccination.2,4,13 Our results confirm this observation, therefore, we preferred to consider the presence of protective levels of antibodies rather than titer increases as indicative of an adequate immunity against pneumococci. While protective levels of antibodies have not been defined clearly for each of the pneumococcal serotypes, systemic pneumococcal infection rarely occurs in adults with serotype specific antibodies ⬎250 to 300 ng Ab N/mL.8 In children, antibody titers ⱖ200 ng Ab N/mL were correlated with decreased nasopharyngeal colonization with specific pneumococcal serotypes.4 Further studies are needed to better define what constitutes a protective antibody response to the various serotypes.
Conventional assessments of vaccine efficacy based on eradication of colonization and/or prevention of invasive disease are extremely difficult to perform in routine clinical settings due to the large number of pneumococcal serotypes included in the vaccine, the heterogeneity of infections caused by these serotypes, and the length of follow-up time needed to evaluate the effect of immunization on the recurrence of pneumococcal infections. While many parents and we were convinced of the benefits of pneumococcal immunization in this pediatric patient population, strict proof of effectiveness is impossible in a routine clinical setting like the one in which this work was performed. The most informative serotype-specific antibody responses for the differentiation of normal from abnormal responses still needs to be defined. Serotype 3, with its high immunogenicity, is clearly a very useful serotype for the evaluation of immunologic responses. The good response to IgG replacement therapy of those patients who failed to develop protective anti-
345
bodies against this serotype supports this assumption; however, it is also possible that some patients who respond only to this serotype may have a significant problem recognizing and responding to most other serotypes included in the vaccine and causing infections. Two of our patients in this category, who also had a good clinical response to IgG replacement, lend support to this possibility. Only randomized, double-blind, placebo-controlled studies like the one suggested by Knutsen could provide definitive support for any of these observations.1 For practical purposes, the criteria described in the “Methods” section above for defining abnormal responses to pneumococcal immunization and for considering IgG replacement therapy have proven useful to us. Based on the data presented here, these criteria do not appear to need modification at this time. While it is not clear which additional serotype responses would be the most important to evaluate, the development of experimental conjugate vaccines to seven serotypes has led to an effort sponsored by the Centers for Disease Control to standardize the measurement of IgG antibodies against these serotypes. Our laboratory is participating in this effort and it is hoped that, in the future, most laboratories will measure antibodies against the same serotypes with a standardized method validated by reference sera available from the US Food and Drug Administration. With this tool in hand, it should be possible to answer many as-yet-unresolved questions relevant to clinicians, including what constitutes a normal antibody response at different ages, how stable antibody titers are after immunization, which antibody
346
levels are protective against invasive diseases and nasopharyngeal colonization, whether evidence of significant titer increases after immunization may be associated with protection even if arbitrarily defined protective levels are not reached after one immunization, and what the value of re-immunization with either conventional or conjugated pneumococcal vaccines will be. ACKNOWLEDGMENTS We thank Mr. George Giannakos for developing the patient database, performing statistical calculations and preparing graphical material and Ms. Patricia Giangrosso for her editorial assistance in preparing this manuscript. REFERENCES
7.
8.
9.
10.
1. Knutsen AP. Patients with IgG subclass and/or selective antibody deficiency to polysaccharide antigens: initiation of a controlled clinical trial of intravenous immune globulin. J Allergy Clin Immunol 1989;84:640 –7. 2. Borgon˜o JM, McLean AA, Vella PP, et al. Vaccination and revaccination with polyvalent pneumococcal polysaccharide vaccines in adults and infants. Proc Soc Exp Biol Med 1978; 157:148 –54. 3. Douglas RM, Paton JC, Duncan SJ, Hansman DJ. Antibody response to pneumococcal vaccination in children younger than five years of age. J Infect Dis 1983;148:131–7. 4. Lawrence EM, Edwards KM, Schiffmann G, et al. Pneumococcal vaccine in normal children. Am J Dis Child 1983;137:846 –50. 5. Paton JC, Toogood IR, Cockington RA, Hansman D. Antibody response to pneumococcal vaccine in children aged 5 to 15 years. Am J Dis Child 1986;140:135– 8. 6. Schiffman G, Douglas RM, Bonner MJ, et al. Radioimmunoassay for im-
11.
12.
13.
munologic phenomena in pneumococcal disease and for the antibody response to pneumococcal vaccines. I. Method for the radioimmunoassay of anticapsular antibodies and comparison with other techniques. J Immunol Methods 1980;33:133– 44. Barret DJ, Lee CG, Amman AJ, Ayoub EM. IgG and IgM pneumococcal polysaccharide antibody responses in infants. Pediatr Res 1984;18: 1067–71. Landesman SH, Schiffman G. Assessment of the antibody response to pneumococcal vaccine in high-risk populations. Rev Infect Dis 1981;3: S184 –S97. Lahood N, Emerson SS, Kumar P, Sorensen RU. Antibody levels and response to pneumococcal vaccine in steroid-dependent asthma. Ann Allergy 1993;70:289 –94. Shapiro ED, Austrian R. Serotypes responsible for invasive Streptococcus pneumoniae infections among children in Connecticut. J Infect Dis 1994;169: 212– 4. Gray BM, Dillon HC. Clinical and epidemiological studies of pneumococcal infection in children. Pediatr Infect Dis 1986;5:201–7. Jorgensen JH, Howell AW, Maher LA, Facklam RR. Serotypes of respiratory isolates of Streptococcus pneumoniae compared with capsular types included in the current pneumococcal vaccine. J Infect Dis 1991;163:644 – 6. Baker CJ, Kasper DL, Edwards MS, Schiffman G. Influence of preimmunization antibody levels on the specificity of the immune response to related polysaccharide antigens. N Engl J Med 1980;303:173– 8.
Request for reprints should be addressed to: Ricardo U. Sorensen, MD Department of Pediatrics Louisiana State University Medical Center 1542 Tulane Avenue New Orleans, LA 70112-2822
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
Background: Patients with recurrent infections and normal IgG levels may have an abnormal response to pneumococcal polysaccharides. The ability to develop antibodies against different pneumococcal polysaccharides develops gradually in the first years of life, but the sequence of development and the influence of preexisting antibody titers has not been defined. Methods: Preimmunization and postimmunization IgG antibody titers against pneumococcal serotypes 3, 7F, 9N, and 14 were evaluated in a population of 100 1to 18-year-old children referred to a pediatric allergy-immunology clinic because of recurrent respiratory infections. None of the patients had a known immunodeficiency syndrome; all had normal total IgG levels. Postimmunization antibody levels were obtained 4 to 6 weeks after immunization. Patients ⱕ5 years of age who failed to develop antibody levels above 200 ng Ab N/mL against any serotype and older patients who failed to develop these levels against a second serotype in addition to serotype 3 were considered for IgG replacement therapy. Results: Prior to immunization, 50% of 51 patients did not have protective antibody levels against any of the serotypes tested. Immunization induced a high response to serotype 3 in all age groups, but responses to serotypes 7F and 14 increased with age. Five of 78 patients (6.4%) failed to develop protective antibody levels against any serotype tested. Three of these patients had clinical criteria that justified the use of IgG replacement therapy; all improved. Three patients were re-immunized 1 to 2 years after the first immunization and all developed protective levels of antibodies against serotype 3 after the second immunization. Conclusion: We conclude that, although measurement of antibody levels against pneumococcal serotype 3 allows a good differentiation of patients who are able to develop anti-polysaccharide antibodies from those who are not, further studies of the development of specific antibodies against other vaccine serotypes in normal populations of different ages are needed to define a normal response to pneumococcal polysaccharides. Ann Allergy Asthma Immunol 1996;76:341– 6.
INTRODUCTION Pneumococcal immunization and evaluation of the resulting antibody response against pneumococcal serotypes has been used increasingly to enhance protection against pneumococcal infection and to identify deficiencies in the devel-
* Division of Allergy and Immunology, Department of Pediatrics, Louisiana State University Medical Center, New Orleans, Louisiana. Received for publication July 14, 1995. Accepted for publication in revised form September 6, 1995.
VOLUME 76, APRIL, 1996
opment of specific antibodies against polysaccharide antigens. Absent or low specific antibody responses to polysaccharides are normally expected in children under 2 years of age. When the lack of an adequate response to pneumococcal polysaccharides persists past 2 years of age, it defines a deficiency of specific antibodies that may be present in patients with normal IgM, IgG, and IgA levels.1 The immunogenicity, however, of the various serotypes included in the available 23-valent vaccines varies and the ability to mount adult-like responses does not de-
velop uniformly for all serotypes at 2 years of age. Several immunization trials have clearly shown a stepwise, age-dependent development of the ability to produce antibodies against the various polysaccharide serotypes included in the vaccines used.2–5 In these early studies, the antibody response was determined with a radioimmunoassay which did not discriminate between IgM and IgG responses.6 Some of the observed responses tended to wane rapidly,2 suggesting that they were short-lived IgM responses. Long-term protective immunity against pneumococcal infection has been associated with the presence of serotypespecific IgG antibodies. Using a method capable of differentiating IgM from IgG responses, three types of responses to different serotypes were identified in young children: high IgM and IgG antibodies against serotype 3; intermediate IgM titers and low IgG titers against serotype 23; and little or no IgM or IgG response to serotypes 6, 18, and 19.7 The sequence of development of IgG antibodies against different serotypes of pneumococci at different ages and the effect of preexisting IgG antibodies to pneumococcal serotypes have not been defined well to date. Our objective was to evaluate the presence of preimmunization and postimmunization IgG antibodies against different pneumococcal serotypes in pediatric patients with a history of recurrent respiratory infections and normal levels of IgG. PATIENTS AND METHODS Patient Population We analyzed retrospectively 100 patients between 1 and 18 years of age referred to our allergy-immunology clinic between 1991 and 1994 with
341
Table 1. Pneumococcal Antibody Titers in Unimmunized and Immunized Children* Age Group
Unimmunized Serotype 3 7F 9N 14 P Value‡ Immunized Serotype 3 7F 9N 14 P Value‡
P Value†
1 (1–5 yr)
2 (6–10 yr)
3 (11–18 yr)
n ⫽ 23
n ⫽ 16
n ⫽ 12
102 ⫾ 4.19 60 ⫾ 2.61 35 ⫾ 3.03 71 ⫾ 2.78 .030650 n ⫽ 48
412 ⫾ 5.47 100 ⫾ 2.61 82 ⫾ 2.65 128 ⫾ 3.02 .006529 n ⫽ 19
715 ⫾ 4.68 122 ⫾ 1.93 84 ⫾ 2.09 225 ⫾ 2.27 .001601 n ⫽ 11
.001981 .052826 .007572 .004271
1764 ⫾ 3.35 264 ⫾ 3.35 241 ⫾ 3.59 97 ⫾ 2.15 ⬍.000001
1741 ⫾ 4.35 361 ⫾ 4.30 365 ⫾ 3.73 219 ⫾ 3.49 .000028
1895 ⫾ 2.59 484 ⫾ 5.31 208 ⫾ 3.90 284 ⫾ 4.23 .005877
.922600 .300863 .312861 .004497
* Serotype-specific IgG titers measured by enzyme-linked immunosorbent assay expressed as ng Ab N/mL. Values are expressed as mean ⫾ SD. † Comparison among age groups by Kruskal-Wallis One Way ANOVA. ‡ Comparison among serotypes by Kruskal-Wallis One Way ANOVA.
histories of recurrent respiratory infections. Only patients with normal levels of IgG and without known primary immunodeficiency syndromes were included in this analysis. Fiftytwo patients were 1 to 5 years old (group 1). Of these patients, 48 were over 2 years of age and 4 were between 18 and 24 months old. Twenty-eight were 6 to 10 years old (group 2) and 20 were 11 to 18 years old (group 3). There was a slight predominance of males in all age groups: 64%, 54%, and 55%, respectively. Thirty-five percent of the patients were African-American. Immunization with Pneumococcal Vaccine Immunization was performed with a 23-valent pneumococcal vaccine (PnuImmune, Lederle, Wayne, NJ or PneumovaxR, Merck, Sharp and Dome, West Point, PA). None of the patients had a febrile illness at the time of immunization. Immunologic Evaluation To decrease the likelihood of vaccine reactions, preimmunization antibody titers were obtained in patients with very frequent bacterial infections and in older patients who may have developed antibodies in response to
342
bacterial infections.2 Frequently, however, the vaccine was given at the time of the first visit and the immunological evaluation was performed 4 to 8 weeks after immunization. Immunoglobulins G, M, and A were determined by nephelometry. Age-adjusted normal immunoglobulin values were used to select patients with normal IgG levels and to determine whether IgM and IgA levels were within the normal range. Pneumococcal polysaccharide antibodies against serotypes 3, 7F, 9N, and 14 were determined commercially by an enzyme immunoassay (Specialty Laboratories, Los Angeles, CA). Only IgG antibodies are measured in this assay which uses an isotype-specific conjugate. Antibody levels are derived from comparison of the immunoassay signal to the standard curve generated for each pneumococcal serotype. The standard curve is calibrated against concentrations of antibody in nanograms of antibody nitrogen per mL (ng Ab N/mL). Maximum variation for repeat analyses of the same serum sample with this method is 15%. Levels above 200 ng Ab N/mL were considered protective.4,8 All four serotypes are included in the polyvalent pneumococcal vaccine given to these patients.
Criteria for IgG Replacement Therapy We used a modification of the criteria defined by Knutsen based on severity of infections and antibody response to pneumococcal immunization.1 IgG replacement therapy was considered when there was evidence of recurrent respiratory infections after adequate treatment of other risk factors, eg, gastroesophageal reflux, allergic rhinitis, asthma, or anatomical abnormalities of the airways. The infections considered were otitis media and/or sinusitis recurring at least 6 times/yr or pneumonia 3 times/yr. In addition, patients ⱕ5 years of age must have failed to develop protective antibody levels against any serotype, and older patients must have failed to develop protective antibody levels against a second serotype in addition to serotype 3. In some patients, IgG replacement therapy was discontinued after 1 to 2 years. These patients were immunized with a second dose of pneumococcal vaccine and retested 4 to 6 weeks later. Data Collection and Statistical Analysis The Epi-Info program was used to create a patient database and perform data analysis. All antibody titer data were
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
analyzed following logarithmic transformation in order to decrease the influence of a few extreme measurements in these positively skewed distributions. Geometric means of the data were obtained and comparisons of preimmunization and postimmunization antibody titers were made by the paired t test. Comparisons between age groups were made using the unpaired t test or analysis of variance for comparison of three or more age groups.9 RESULTS By definition, all patients had normal IgG levels. No IgM-deficient patients were found. Four patients had IgA levels below 2 SD for their age group: three patients in group 1 had IgA values of 8, 18, and 19 mg/dL (normal ⫽ 20 – 84 mg/dL); and one 16-year-old patient had an IgA level of 11 mg/dL. Preimmunization antibody titers were measured in 51 patients: 23 in group 1, 16 in group 2, and 12 in group 3. The mean antibody titers against each one of the four serotypes increased significantly with age (Table 1 and Fig 1, panels A–C). Twenty-six patients did not have protective antibody levels against any serotype tested. Of the remaining 25 patients, 14 had antibodies to serotype 3 only, 1 to serotype 7F only, 2 to serotype 14 only and 8 patients had protective antibodies against two or more serotypes. Pneumococcal immunization was well tolerated by most patients. Mild local and occasional febrile reactions were reported by some patients over 5 years of age. Postimmunization antibody titers were measured in 78 patients: 48 in group 1, 19 in group 2, and 11 in group 3. The mean titers of antibodies against the four serotypes were significantly different within each age group (Table 1). Serotype 3 induced a similarly high response in all age groups. Although the titers to serotype 7F and 14 increased with age, the differences among age groups were significant only for serotype 14. The percentages of patients who developed protective antibody titers against the four serotypes also in-
VOLUME 76, APRIL, 1996
Figure 1. Preimmunization (panels A–C) and postimmunization (panels D–E) pneumococcal antibody titers in patients in three different age groups. More patients in each age group had postimmunization titers than preimmunization antibody titers.
creased with age (Fig 1, panels D–F). Only three patients in the first group and two in the second group failed to develop protective antibody levels against serotype 3. The three patients in the first group under 2 years of age who had postimmunization titers responded to one or more serotypes and did not differ from the patients between 2 and 5 years of age. Lack of development of protective antibody levels to serotypes 7F, 9N, and 14 was common in patients younger than 11 years of age. Low antibodies to sero-
types 9N and 14 were also frequent in patients over 10 years of age. Five of 78 patients (6.4%) failed to develop protective antibody levels against any serotype tested. Eight patients (10%) of all age groups developed protective antibody levels to serotype 3 only. Preimmunization and postimmunization titers were measured in 29 patients: 19 in group 1 and 10 in groups 2 and 3. There was a significant increase in postimmunization titers to serotypes 3, 7F, and 9N in group 1 and to all serotypes in groups 2 and 3 (Table
343
Table 2. Preimmunization and Postimmunization Pneumococcal Antibody Titers* Age Groups, yr 1–5 (n ⫽ 19) Serotype 3 7F 9N 14 6–18 (n ⫽ 10) Serotype 3 7F 9N 14
Preimmunization Titers
Postimmunization Titers
P Value†
87 ⫾ 4.04 57 ⫾ 2.82 32 ⫾ 3.24 68 ⫾ 2.98
1313 ⫾ 5.04 152 ⫾ 4.57 135 ⫾ 4.21 94 ⫾ 3.53
.00160 .00880 .01290 .18160
348 ⫾ 5.64 77 ⫾ 2.33 65 ⫾ 2.35 143 ⫾ 3.67
1753 ⫾ 6.18 521 ⫾ 4.51 405 ⫾ 4.95 238 ⫾ 4.01
.01090 .01360 .02720 .08520
* Serotype-specific IgG titers measured by enzyme-linked immunosorbent assay and expressed as ng Ab N/mL. Values are expressed as mean ⫾ SD. † Comparison of preimmunization and postimmunization values by paired t test.
2). Four patients failed to have an increase in titer into the protective range against serotype 3, 11 against serotype 7F, 13 against 9N, and 20 against serotype 14 (Fig 2). Several patients had twofold or higher increases in titer against one or more serotypes without reaching the arbitrarily defined protective level of 200 ng Ab N/mL. High preimmunization titers against any serotype tended not to increase further after immunization, and a decrease in titer against serotype 3 was observed in one patient (Fig 2, panel A). The clinical characteristics of five patients who failed to develop protective antibody levels against any of the serotypes tested are listed in Table 3. Three of these patients had clinical criteria that justified the use of IgG replacement therapy. All showed significant improvement. IgG replacement could be discontinued after 6 to 12 months in two patients. These patients were re-immunized 4 to 6 months later and both developed protective levels of antibodies against serotype 3. Two additional patients, aged 7 and 15 years, had protective antibody levels to serotype 3 only and were judged to require IgG replacement therapy. Both patients had significant clinical improvement on this therapy, but symptoms recurred when IgG replacement was discontinued and treatment was reinstated.
344
None of the patients 1 to 5 years old who developed protective levels to serotype 3 and none of the older patients who developed protective levels to at least one additional serotype had a
clinical course that would have justified IgG replacement treatment. All four patients with low IgA levels developed protective antibody levels against one or more serotypes after immunization, and none had clinical criteria justifying IgG replacement therapy. DISCUSSION In our patient population evaluated because of recurrent respiratory infections, 50% had preimmunization protective antibody levels against the pneumococcal serotypes tested. The highest frequency of antibodies was found against serotype 3. This may reflect both the high immunogenicity of its polysaccharide2–5 and the frequency of infections with this serotype. While serotype 3 is an infrequent cause of invasive Streptococcus pneumoniae infections in children,10 it is a frequent isolate from the respiratory tract.11,12 The low percentage of pa-
Figure 2. Preimmunization and postimmunization titers in 29 patients: open circles ⫽ 1- to 5-year-old patients; solid triangles ⫽ patients ⱖ6 years of age; * ⫽ patients out of range of graph.
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
Table 3. Patients with No Response to Pneumococcal Immunization* Patient
Age
Clinical Presentation
1
2y
2
2y
Recurrent otitis media Recurrent sinusitis Allergic rhinitis Bronchial asthma Recurrent purulent otitis media, refractory to antibiotic treatment
3
4y
4
7y
5
8y
Recurrent otitis media Recurrent pneumonia Atopic dermatitis Allergic rhinitis Bronchial asthma Recurrent otitis media Allergic rhinitis Cranio-velo-facial syndrome Recurrent sinusitis Recurrent pneumonia Severe atopic dermatitis Allergic rhinitis Bronchial asthma
Rx & Clinical Course Antibiotic therapy only Lost to follow up
IgG (IVIG) replacement for 6 months with clinical improvement Reimmunization 6 months later with response to serotype 3 only Sustained clinical improvement afterwards IgG (IVIG) replacement for 1 year with clinical improvement Reimmunization 1 year later with response to serotype 3 only Sustained clinical improvement afterwards Antibiotic therapy only Reimmunization 1 year later with response to serotype 3 only Chronic IgG (IVIG) replacement with significant clinical improvement of infections & atopic dermatitis
* These patients failed to develop antibody levels ⬎200 ng Ab N/mL against any of the four serotypes tested.
tients who have developed protective antibody levels in response to clinical and subclinical infections appears to justify the use of pneumococcal vaccines in this population and indicates that the risk of reactions due to high preexisting antibody titers against vaccine serotypes is low. Our results clearly document that there is a stepwise development of the ability to develop IgG antibodies to various serotypes beyond the second year of life. Comparison of our pediatric population with asthmatic adults studied with the same methodology in the same area9 revealed that the mean postimmunization titers to serotypes 9N and 14 are significantly higher in 27- to 65-year-old adults than in all age groups included in this report (Table 1): 585 ⫾ 2.78 ng Ab N/mL for serotype 9N and 564 ⫾ 1.2 ng Ab N/mL for serotype 14. Since age is clearly one of the most important determinants of the specific antibody response to pneumococcal polysaccharides, pediatric patients who fail to develop protective antibody levels to pneumococcal immunization
VOLUME 76, APRIL, 1996
in the first years of life may be reimmunized at a later age. In the few patients in our study in whom re-immunization was performed, we always observed a higher response to at least one serotype after the second immunization. High levels of preexisting antibodies tend to decrease the increment in antibody titers after vaccination.2,4,13 Our results confirm this observation, therefore, we preferred to consider the presence of protective levels of antibodies rather than titer increases as indicative of an adequate immunity against pneumococci. While protective levels of antibodies have not been defined clearly for each of the pneumococcal serotypes, systemic pneumococcal infection rarely occurs in adults with serotype specific antibodies ⬎250 to 300 ng Ab N/mL.8 In children, antibody titers ⱖ200 ng Ab N/mL were correlated with decreased nasopharyngeal colonization with specific pneumococcal serotypes.4 Further studies are needed to better define what constitutes a protective antibody response to the various serotypes.
Conventional assessments of vaccine efficacy based on eradication of colonization and/or prevention of invasive disease are extremely difficult to perform in routine clinical settings due to the large number of pneumococcal serotypes included in the vaccine, the heterogeneity of infections caused by these serotypes, and the length of follow-up time needed to evaluate the effect of immunization on the recurrence of pneumococcal infections. While many parents and we were convinced of the benefits of pneumococcal immunization in this pediatric patient population, strict proof of effectiveness is impossible in a routine clinical setting like the one in which this work was performed. The most informative serotype-specific antibody responses for the differentiation of normal from abnormal responses still needs to be defined. Serotype 3, with its high immunogenicity, is clearly a very useful serotype for the evaluation of immunologic responses. The good response to IgG replacement therapy of those patients who failed to develop protective anti-
345
bodies against this serotype supports this assumption; however, it is also possible that some patients who respond only to this serotype may have a significant problem recognizing and responding to most other serotypes included in the vaccine and causing infections. Two of our patients in this category, who also had a good clinical response to IgG replacement, lend support to this possibility. Only randomized, double-blind, placebo-controlled studies like the one suggested by Knutsen could provide definitive support for any of these observations.1 For practical purposes, the criteria described in the “Methods” section above for defining abnormal responses to pneumococcal immunization and for considering IgG replacement therapy have proven useful to us. Based on the data presented here, these criteria do not appear to need modification at this time. While it is not clear which additional serotype responses would be the most important to evaluate, the development of experimental conjugate vaccines to seven serotypes has led to an effort sponsored by the Centers for Disease Control to standardize the measurement of IgG antibodies against these serotypes. Our laboratory is participating in this effort and it is hoped that, in the future, most laboratories will measure antibodies against the same serotypes with a standardized method validated by reference sera available from the US Food and Drug Administration. With this tool in hand, it should be possible to answer many as-yet-unresolved questions relevant to clinicians, including what constitutes a normal antibody response at different ages, how stable antibody titers are after immunization, which antibody
346
levels are protective against invasive diseases and nasopharyngeal colonization, whether evidence of significant titer increases after immunization may be associated with protection even if arbitrarily defined protective levels are not reached after one immunization, and what the value of re-immunization with either conventional or conjugated pneumococcal vaccines will be. ACKNOWLEDGMENTS We thank Mr. George Giannakos for developing the patient database, performing statistical calculations and preparing graphical material and Ms. Patricia Giangrosso for her editorial assistance in preparing this manuscript. REFERENCES
7.
8.
9.
10.
1. Knutsen AP. Patients with IgG subclass and/or selective antibody deficiency to polysaccharide antigens: initiation of a controlled clinical trial of intravenous immune globulin. J Allergy Clin Immunol 1989;84:640 –7. 2. Borgon˜o JM, McLean AA, Vella PP, et al. Vaccination and revaccination with polyvalent pneumococcal polysaccharide vaccines in adults and infants. Proc Soc Exp Biol Med 1978; 157:148 –54. 3. Douglas RM, Paton JC, Duncan SJ, Hansman DJ. Antibody response to pneumococcal vaccination in children younger than five years of age. J Infect Dis 1983;148:131–7. 4. Lawrence EM, Edwards KM, Schiffmann G, et al. Pneumococcal vaccine in normal children. Am J Dis Child 1983;137:846 –50. 5. Paton JC, Toogood IR, Cockington RA, Hansman D. Antibody response to pneumococcal vaccine in children aged 5 to 15 years. Am J Dis Child 1986;140:135– 8. 6. Schiffman G, Douglas RM, Bonner MJ, et al. Radioimmunoassay for im-
11.
12.
13.
munologic phenomena in pneumococcal disease and for the antibody response to pneumococcal vaccines. I. Method for the radioimmunoassay of anticapsular antibodies and comparison with other techniques. J Immunol Methods 1980;33:133– 44. Barret DJ, Lee CG, Amman AJ, Ayoub EM. IgG and IgM pneumococcal polysaccharide antibody responses in infants. Pediatr Res 1984;18: 1067–71. Landesman SH, Schiffman G. Assessment of the antibody response to pneumococcal vaccine in high-risk populations. Rev Infect Dis 1981;3: S184 –S97. Lahood N, Emerson SS, Kumar P, Sorensen RU. Antibody levels and response to pneumococcal vaccine in steroid-dependent asthma. Ann Allergy 1993;70:289 –94. Shapiro ED, Austrian R. Serotypes responsible for invasive Streptococcus pneumoniae infections among children in Connecticut. J Infect Dis 1994;169: 212– 4. Gray BM, Dillon HC. Clinical and epidemiological studies of pneumococcal infection in children. Pediatr Infect Dis 1986;5:201–7. Jorgensen JH, Howell AW, Maher LA, Facklam RR. Serotypes of respiratory isolates of Streptococcus pneumoniae compared with capsular types included in the current pneumococcal vaccine. J Infect Dis 1991;163:644 – 6. Baker CJ, Kasper DL, Edwards MS, Schiffman G. Influence of preimmunization antibody levels on the specificity of the immune response to related polysaccharide antigens. N Engl J Med 1980;303:173– 8.
Request for reprints should be addressed to: Ricardo U. Sorensen, MD Department of Pediatrics Louisiana State University Medical Center 1542 Tulane Avenue New Orleans, LA 70112-2822
ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY