- Email: [email protected]
The validity of the criteria for primary infection of Chlamydophila pneumoniae in children by measuring ELISA IgM antibodies
J Infect Chemother (2012) 18:308–312 DOI 10.1007/s10156-011-0327-x
ORIGINAL ARTICLE
The validity of the criteria for primary infection of Chlamydophila pneumoniae in children by measuring ELISA IgM antibodies Ayako Kamata • Kaoru Obinata • Takahiro Niizuma Keiji Kinoshita • Toshiaki Shimizu
•
Received: 25 April 2011 / Accepted: 3 October 2011 / Published online: 19 October 2011 Ó Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases 2011
Abstract As IgM antibody measurement by enzymelinked immunosorbent assay (ELISA) has become possible for the serological diagnosis of Chlamydophila pneumoniae (C. pn) infection, the HITAZYME-ELISA method has become widely employed in Japan. However, in children, when the diagnostic criterion of primary infection is set at ID C1.1, the positive rate is higher than expected, and the potential for inaccurate reflection of the prevalence has been raised. In this study, we performed ROC analysis involving 136 pediatric patients with acute airway symptoms (0–14 years of age), considering a 32-fold or higher microimmunofluorescence IgM antibody titer against C. pn as positive. Setting the cut-off value for ELISA C. pn IgM antibody ID at 2.0, the specificity was 100%, with no false positivity. The maximum (sensitivity ? specificity)/2 was obtained when the cut-off value was set at 1.5. Therefore, IgM ID C2.0 was regarded as definitely positive and an IgM ID between 1.5 and 2.0 was regarded as indeterminate as diagnostic criteria for the primary infection. When the prevalence was investigated in 3,108 children (0–15 years of age) with airway symptoms based on these criteria, 542 cases (17.4%) were positive, and the median duration of IgM antibody positivity was five months. Long-term positivity
A. Kamata K. Obinata T. Niizuma K. Kinoshita Department of Pediatrics, Koshigaya Municipal Hospital, Saitama, Japan A. Kamata T. Shimizu Department of Pediatrics, Juntendo University Faculty of Medicine, Tokyo, Japan K. Obinata (&) Department of Pediatrics, Juntendo University Urayasu Hospital, 2-1-1 Tomioka, Urayasu, Chiba 279-0021, Japan e-mail: [email protected]
123
(ten cases) for more than 12 months and recurrent positivity (eight cases) were also observed, but it may be appropriate to set a new criterion of IgM antibody ID C2.0 for the diagnosis of primary Chlamydophila pneumoniae infection in children. Keywords Chlamydophila pneumoniae Serological diagnosis Enzyme-linked immunosorbent assay Immunoglobulin M Children Primary infection
Introduction Chlamydophila pneumoniae (C. pn) is an important causative pathogen of airway infection [1] that has been suggested to be associated with asthma and arteriosclerosis. Diagnostic methods include detection of the pathogen, i.e., isolation culture and polymerase chain reaction (PCR) test, and measurement of the serum specific antibody titer [2]. Regarding antibody titer measurement, the micro-immunofluorescence (MIF) method is considered to be the world standard [3], but it requires skill to perform the technique and to accurately judge the results. On the other hand, enzyme-linked immunosorbent assay (ELISA) has recently been employed in clinical practice because it requires no complex techniques, and judgments can be made objectively [4]. In Japan, a kit using a purified chlamydial outer membrane complex as the antigen (HITAZYME C. pneumoniae; Hitachi Chemical Co. Ltd., Tokyo, Japan) has been widely used [5], and IgM antibody measurement has been covered by the national health insurance since 2005. However, the positive rate is unexpectedly high, and negative conversion does not occur for a prolonged period when the current diagnostic criteria of C. pn primary infection in children (IgM ID C1.1) is applied, which can
J Infect Chemother (2012) 18:308–312
be problematic [6]. In this study, we re-evaluated the diagnostic criteria of primary infection involving children with airway symptoms, and investigated the prevalence of the infection employing the new criteria.
Patients and methods Patients 1.
2.
The subjects were 136 children (0–3 years of age, 68; 4–14 years of age, 68; 77 boys and 59 girls) who visited the Department of Pediatrics, Koshigaya Municipal Hospital, for acute airway symptoms between July 2005 and June 2010. Samples that were positive for anti-mycoplasma IgM antibody (ImmunoCard M. pneumoniae-specific IgM; Meridian Bioscience Inc., Cincinnati, OH, USA) were excluded. ELISA C. pn IgM ID and MIF C. pn IgM were measured, ROC analysis was performed with MIF considered the standard method, and the candidate values for the ELISA C. pn IgM diagnostic criteria of primary infection were investigated. The prevalence of primary infection was investigated in 3,108 children (0–15 years of age, 1,688 boys and 1,420 girls) who visited our department for airway symptoms (cough with/without sputum, rhinorrhea, wheezing, and fever) between January 2004 and December 2010, based on the criteria established above. When several samples were collected from the same subject, the maximum value was adopted.
Informed consent was obtained from the family of each patient. The study protocol was approved by the institutional Ethics Committee at Koshigaya Municipal Hospital.
309
of p-nitrophenol was measured at a wavelength of 405 nm, which corresponded to the antibody level in the sample. MIF The antibody titers against various Chlamydia strains were measured employing the MIF method reported by Wang and Grayston [7]. Five antigens, i.e., C. pn (TW-183 strain), mixed antigens of three C. trachomatis groups (B ? D ? E, C ? J ? H ? I, and G ? F ? K strains), and C. psittaci (Budgerigar no. 1 strain), were spotted on a multislide, dried at room temperature, and fixed in acetone for 15 min. In the primary reaction, eight- to twofold serial serum dilutions were reacted at 37°C for 2 h. After washing, the slide was dried sufficiently at room temperature, and then the secondary reaction with fluorescein isothiocyanateconjugated goat-labeled anti-human IgM antibody was performed at 37°C for 1 h. Finally, the slide was mounted with 50% glycerin–PBS and a cover glass, and the fluorescence of EB particles for each of the Chlamydia strains was observed under a fluorescence microscope at 4009 magnification. When a fluorescence particle was observed at 32-fold or higher dilution, the serum was judged to be positive. In the statistical analysis, the chi-square test was employed to analyze gender differences and the annual frequency. The t test was employed to analyze differences in the mean, and p values of less than 5% were regarded as significant. ANOVA was employed to compare three groups.
Results 1.
Methods ELISA Antibodies were measured according to the test procedure of HITAZYME C. pneumoniae Ab-IgM. The serum sample was diluted 21-fold with sample processing solution and used as an IgM measurement sample. In the chlamydial outer membrane protein complex antigen-fixed plate, 100 ll of blank, negative, and positive control sera and diluted samples were applied and incubated at 37°C for 1 h. After washing three times, 100 ll of alkaliphosphataselabeled anti-human IgM antibody were added and the mixture incubated at 37°C for 1 h. After washing another three times, 100 ll of substrate solution were added and reacted at 20–25°C for 10 min, followed by the addition of 25 ll of the reaction-stopping solution, and the absorbance
The distribution of ELISA C. pn IgM ID by age is shown in Fig. 1. The geometric means of the ELISA C. pn IgM ID were 1.02 and 1.64 in zero- to three- and four- to 14-year-old children, respectively. The number of patients with upper respiratory infection, bronchitis, pneumonia, and bronchial asthma were 13, 59, 40, and 41, respectively. Upon analyzing of the correlation between ELISA C. pn IgM ID and MIF C. pn IgM (Fig. 2), the correlation coefficient was 0.66. MIF C. pn IgM was positive in 38 and negative in 30 of the 68 zeroto three-year-old patients. Of the 68 four- to 14-year-old patients, MIF C. pn IgM was positive in 42 and negative in 26. In the ROC analysis, when the cut-off value of the ELISA C. pn IgM antibody ID was set at 2.0, the specificity was 100% in both groups, and no false positivity was observed. In the 136 zero- to 14-year-old patients, the maximum (sensitivity ? specificity)/2 was obtained when the cut-off value was set at 1.5. The
123
310
J Infect Chemother (2012) 18:308–312 Table 1 Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of ELISA Chlamydophila pneumoniae IgM, with MIF used as the reference standard MIF C. pn IgM ?
Total -
ELISA C. pn IgM C1.5
57
6
63
\1.5
23
50
73
80
56
136
Total
Sensitivity 71%, specificity 89% Fig. 1 Age distribution of Chlamydophila pneumoniae IgM antibody titers (ELISA)
PPV 90%, NPV 68% ELISA C. pn IgM C2.0
35
0
35
\2.0
45
56
101
80
56
136
Total
Sensitivity 44%, specificity 100% PPV 100%, NPV 55%
Fig. 2 Correlation between titers of Chlamydophila pneumoniae IgM determined by ELISA and by MIF
2.
consistency at each cut-off value is shown in Table 1. Based on these findings, IgM ID C2.0 was regarded as definitely positive and between 1.5 and 2.0 was considered indeterminate as the diagnostic criteria of primary infection. Table 2 presents the number of patients with primary C. pn infection in each age group when ELISA C. pn IgM ID C2.0 was regarded as positive, and the geometric mean ELISA C. pn IgM ID. Of the 3,108 patients, 542 (17.4%) were judged as being C. pn IgMpositive: 247 boys and 295 girls, the mean age was 4.8 years, the geometric mean WBC was 9,084/ll, and the geometric mean CRP level was 1.69 mg/dl. Indeterminate cases (ELISA C. pn IgM ID between 1.5 and 2.0) accounted for 13.2% of all cases. By year, the positive rate was significantly higher in 2008 and 2009 compared with the other years (Fig. 3).
Anti-Mycoplasma IgM antibody was measured in 2,960 patients, and found to be positive in 1,004 (33.9%) and negative in 1,812 (indeterminate in 144). Both ELISA C. pn IgM ID C2.0 and anti-Mycoplasma IgM antibody were positive in 225, ELISA C. pn IgM alone was positive in 255, and anti-Mycoplasma IgM alone was positive in 779. Upon comparing these three groups, there were no
123
significant differences in the age, WBC, or CRP level (Table 3). No significant difference was noted in the ELISA C. pn IgM antibody titer between the patients who were positive for both IgM antibodies and those positive for only ELISA C. pn IgM. Among the patients from whom samples were collected several times (1,550 samples from 596 patients), the antibody titer was measured once or more following the judgment of ELISA C. pn IgM positivity in 107 patients; among these, C. pn IgM converted to negative after three days in the fastest case and after a median of five months. Long-term positivity was observed in ten patients in whom the antibody was positive upon re-testing after more than 12 months. A five-year-boy had ELISA C. pn IgM ID 3.2- and MIF C. pn IgM 128-fold at the first examination. After 21 months, these titers were still positive (2.48- and 32-fold). Eight patients reverted to positive after becoming negative once. IgG or IgA was measured in 2,671 patients, and the diagnostic criterion— an increase of 3 or more—was met by 61 patients (2.3%) and only ten (0.6%) of 1,492 patients aged three years or younger. In paired serum samples, a significant increase was noted in only one patient (a four-year-old girl).
Discussion In a study based on MIF C. pn IgG antibody reported by Ouchi et al. [8], the prevalence of anti-C. pn antibody rose rapidly from about four years of age, and was maintained at about 50% after eight years of age. In a study based on the MIF C. pn IgM antibody [9], C. pn was the causative microorganism in 13.5% of the acute lower airway infection cases in children, showing that the primary infection
J Infect Chemother (2012) 18:308–312
311
Table 2 Age-related seroprevalence of Chlamydophila pneumoniae, as determined by ELISA Age (years)
0
1
2–3
4–6
7–9
10–15
Number of cases
287
678
800
793
296
254
Number of positive cases (%)
9 (3.1)
85 (12.5)
132 (16.5)
173 (21.8)
74 (25.0)
69 (27.2)
Average ELISA C. pn IgM ID (geometric mean)
0.48
0.88
1.14
1.32
1.39
1.43
Fig. 3 Annual occurrence of primary Chlamydophila pneumoniae infections, as determined by ELISA. *p \ 0.01
Table 3 Comparison of the cases judged to be positive with ELISA Chlamydophila pneumoniae IgM and/or ImmunoCard Mycoplasma IgM Both positive
ELISA C. pn IgM-only positive
ImmunoCard Mycoplasma IgM-only positive
Number of patients
225
255
779
Mean age (years)
5.2
4.5
5.0
WBC (/ll) (geometric mean)
8,938
9,079
8,791
CRP (mg/dl) (geometric mean)
1.72
1.63
1.62
C. pn IgM ID (geometric mean)
2.92
2.84
1.05
may be overlooked in children when IgM antibodies are not measured. As the ELISA method, which is simple and correlates well with MIF, has recently entered widespread use, epidemiological surveys of C. pn infection employing ELISA have been performed [3, 5, 10, 11]. In the ELISA method, consistency varies depending on the antigen used [12]. Miyashita et al. [13] reported that ELISA C. pn IgM antibody testing using Ani-Lab-EIA was superior in a multicenter study. HITAZYME-ELISA is widely used in Japan, but problems with the diagnostic criteria, which consider ELISA C. pn IgM ID C1.1 to indicate acute infection, have been pointed out [5, 14]. Niizuma et al. [15] compared the ELISA (Hitazyme C. pneumoniae), MIF, and Western blot methods using the same sera from children with acute airway infection, and observed that the optimum cut-off value of ELISA C. pn IgM ID was 1.4. In our study, to reduce the number of false positive cases, which is a
problem with the current criteria, and to increase the specificity, the cut-off value of ELISA C. pn IgM ID was set at 2.0, and the specificity and positive predictive value improved to 100% in both age groups. Kishimoto et al. [16] recently proposed new criteria in which a value of 2.0 or greater, close to the mean ELISA C. pn IgM ID ?3.0 SD in the criteria for healthy children and adults, was regarded as positive. Since an ELISA C. pn IgM ID C2.0 was regarded as positive based on the above criteria, and the best balance of sensitivity and specificity was obtained at a cut-off value of 1.5 in the ROC curve, an ID of between 1.5 and 2.0 was regarded as indeterminate. Applying these criteria, 17.4% of the children with airway symptoms had primary C. pn infection, and their number increased with age, accounting for 23.5% at four years of age or older. Regarding a gender difference, the prevalence is reported to be higher in males in adults [17], but the positivity rate was significantly higher in girls than in boys in this study. This rate was also close to the frequency of C. pn infection in the study reported by Block et al. [18], in which 48 of 260 pediatric pneumonia patients (18.5%) were MIF C. pn antibody positive. In addition, the positive rate varied between years in this seven-year study. Mycoplasma pneumoniae is recognized as a common pathogen of respiratory infection in children. The ImmunoCard test that detects anti-Mycoplasma IgM antibody by ELISA has found widespread use in Japan. Inutsuka and Yamada [19] reported that ImmunoCard Mycoplasmapositive patients had significantly higher C. pn IgM indices than Mycoplasma-negative patients. In order to clarify the frequency of co-infection with Mycoplasma pneumoniae, we measured anti-Mycoplasma IgM antibody using ImmunoCard. About half of the patients with primary C. pn infection were positive for anti-Mycoplasma IgM antibody, but no significant difference was noted in the ELISA C. pn IgM ID value between those positive and negative for antiMycoplasma IgM antibody. It is considered that crossreactivity between the two pathogens is insignificant. Reportedly, MIF C. pn IgM antibodies appear about three weeks after the primary infection, the titer starts to decrease after two months, it disappears after 4–6 months, and it does not rise during re-infection [20]. Since infantile C. pn infection can be protracted in many cases, although symptoms are mild [21], it is difficult to determine the disease stage on the first test day. In our study, the median duration of positivity was about five months, and long-term
123
312
positivity and reversion to positivity were observed in only a few cases. Very few cases showed significant elevation of IgG and IgA antibody titers, and this may be due to low antibody productivity in childhood [22], suggesting that IgM antibody measurement is essential for the diagnosis of C. pn infection by ELISA in children. The presence of false positive cases is possible even though the cut-off value of the ELISA C.pn IgM antibody titer is elevated, but it may be appropriate to regard HITAZYME ELISA C. pn IgM ID C2.0 as positive for the primary diagnosis of C. pn infection. Acknowledgments We are grateful to Ms. Chikako Mori and Mr. Toshihide Morikawa, Hitachi Chemical Co., Ltd., for their assistance with the ELISA and MIF antibody measurements.
J Infect Chemother (2012) 18:308–312
10.
11.
12.
13.
14.
References 1. Hammerschlag MR. Pneumonia due to Chlamydia pneumoniae in children: epidemiology, diagnosis, and treatment. Pediatr Pulmonol. 2003;36:384–90. 2. Burillo A, Bouza E. Chlamydophila pneumoniae. Infect Dis Clin N Am. 2010;24:61–71. 3. Wang SP. The microimmunofluorescence test for Chlamydia pneumoniae infection: technique and interpretation. J Infect Dis. 2000;181(Suppl 3):S421–5. 4. Verkooyen RP, Willemse D, Hiep-van Casteren SC, Joulandan SA, Snijder RJ, van den Bosch JM, et al. Evaluation of PCR, culture, and serology for diagnosis of Chlamydia pneumoniae respiratory infections. J Clin Microbiol. 1998;36:2301–7. 5. Kishimoto T, Kubota Y, Matsushima T, Izutsu H, Matsumoto A, Soejima R, et al. Assay of specific anti-Chlamydia pneumoniae antibodies by ELISA method. 1. Evaluation of ELISA kit using outer membrane complex. Kansenshogaku Zasshi. 1996;70(8): 821–9 (in Japanese). 6. Miyashita N, Obase Y, Fukuda M, Shoji H, Mouri K, Yagi S, et al. Evaluation of serological tests detecting Chlamydophila pneumoniae-specific immunoglobulin M antibody. Intern Med. 2006;45:1127–31. 7. Wang SP, Grayston JT. Immunologic relationship between genital TRIC, lymphogranuloma venereum, and related organisms in a new microtiter indirect immunofluorescence test. Am J Ophthalmol. 1970;76:367–74. 8. Ouchi K, Kanamoto Y, Ushio M. Prevalence of antibodies to Chlamydia pneumoniae (Strain TWAR) and other Chlamydia in Japan. Kansenshogaku Zasshi. 1991;65(1):19–25 (in Japanese). 9. Ouchi K, Komura H, Fujii M, Matsushima H, Maki T, Hasegawa K, et al. Chlamydia pneumoniae infection and Mycoplasma
123
15.
16.
17. 18.
19.
20. 21.
22.
pneumoniae infection in pediatric patients. Kansenshogaku Zasshi. 1999;73(12):1177–82 (in Japanese). Bamba M, Jozaki K, Sugaya N, Tamai S, Ishihara J, Kori T, et al. Prospective surveillance for atypical pathogens in children with community-acquired pneumonia in Japan. J Infect Chemother. 2006;12(1):36–41. Ciarrocchi G, De Benedetto F, Fogliani V, Magliano E, Del Prete R, Miragliotta G. Serological study on Chlamydophila pneumoniae in patients with community-acquired pneumonia. New Microbiol. 2004;27(4):335–43. Persson K, Boman J. Comparison of five serologic tests for diagnosis of acute infections by Chlamydia pneumonia. Clin Diagn Lab Immunol. 2000;7:739–44. Miyashita N, Ouchi K, Kawasaki K, Nomura H, Kawai Y, Tsumura N, et al. Comparison of serological tests for detection of immunoglobulin M antibodies to Chlamydophila pneumoniae. Respirology. 2008;13:427–31. Miyashita N, Ouchi K, Kawasaki K, Komura H, Kawai Y, Obase Y, et al. Evaluation of enzyme-linked immunosorbent assay for Chlamydophila pneumoniae-specific immunoglobulin M in acute respiratory tract infection. Respirology. 2008;13:299–302. Niizuma T, Kamata A, Kinoshita K, Hisata K, Obinata K. Discussion of positive primary infection criteria for Chlamydophila (Chlamydia) pneumoniae in children: IgM antibody titers by ELISA compared to micro-IF and Western blotting. Kansenshogaku Zasshi. 2008;82(6):619–23 (in Japanese). Kisihimoto T, Ando S, Numazaki K, Ouchi K, Yamazaki T, Nakahama C. Assay of Chlamydia pneumoniae specific IgM antibodies by ELISA method—reduction of non-specific reaction and resetting of serological criteria by measuring IgM antibodies. Jpn J Infect Dis. 2009;62(4):260–4. Leinonen M. Pathogenetic mechanisms and epidemiology of Chlamydia pneumoniae. Eur Heart J. 1993;14(Suppl K):57–61. Block S, Hedrick J, Hammerschlag, Cassell GH, Craft JC. Mycoplasma pneumoniae and Chlamydia pneumoniae in pediatric community-acquired pneumonia: comparative efficacy and safety of clarithromycin vs. erythromycin ethylsuccinate. Pediatr Infect Dis J. 1995;14:471–7. Inutsuka M, Yamada K. The clinical features and the diagnostic problems of Chlamydia pneumoniae infection in pediatric patients. Nihon Syounikagakkai Zasshi. 2008;112(11):1674–9 (in Japanese). Kuo CC, Jackson LA, Campbell LA, Grayston JT. Chlamydia pneumoniae (TWAR). Clin Microbiol Rev. 1995;8:451–61. Normann E, Gnarpe J, Gnarpe H, Wettergren B. Chlamydia pneumoniae in children with acute respiratory tract infection. Acta Paediatr. 1998;87:23–7. Schmidt SM, Mu¨ller CE, Mahner B, Wiersbitzkey SK. Prevalence, rate of persistence and respiratory tract symptoms of Chlamydia pneumoniae infection in 1211 kindergarten and school age children. Pediatr Infect Dis J. 2002;21:758–62.
ORIGINAL ARTICLE
The validity of the criteria for primary infection of Chlamydophila pneumoniae in children by measuring ELISA IgM antibodies Ayako Kamata • Kaoru Obinata • Takahiro Niizuma Keiji Kinoshita • Toshiaki Shimizu
•
Received: 25 April 2011 / Accepted: 3 October 2011 / Published online: 19 October 2011 Ó Japanese Society of Chemotherapy and The Japanese Association for Infectious Diseases 2011
Abstract As IgM antibody measurement by enzymelinked immunosorbent assay (ELISA) has become possible for the serological diagnosis of Chlamydophila pneumoniae (C. pn) infection, the HITAZYME-ELISA method has become widely employed in Japan. However, in children, when the diagnostic criterion of primary infection is set at ID C1.1, the positive rate is higher than expected, and the potential for inaccurate reflection of the prevalence has been raised. In this study, we performed ROC analysis involving 136 pediatric patients with acute airway symptoms (0–14 years of age), considering a 32-fold or higher microimmunofluorescence IgM antibody titer against C. pn as positive. Setting the cut-off value for ELISA C. pn IgM antibody ID at 2.0, the specificity was 100%, with no false positivity. The maximum (sensitivity ? specificity)/2 was obtained when the cut-off value was set at 1.5. Therefore, IgM ID C2.0 was regarded as definitely positive and an IgM ID between 1.5 and 2.0 was regarded as indeterminate as diagnostic criteria for the primary infection. When the prevalence was investigated in 3,108 children (0–15 years of age) with airway symptoms based on these criteria, 542 cases (17.4%) were positive, and the median duration of IgM antibody positivity was five months. Long-term positivity
A. Kamata K. Obinata T. Niizuma K. Kinoshita Department of Pediatrics, Koshigaya Municipal Hospital, Saitama, Japan A. Kamata T. Shimizu Department of Pediatrics, Juntendo University Faculty of Medicine, Tokyo, Japan K. Obinata (&) Department of Pediatrics, Juntendo University Urayasu Hospital, 2-1-1 Tomioka, Urayasu, Chiba 279-0021, Japan e-mail: [email protected]
123
(ten cases) for more than 12 months and recurrent positivity (eight cases) were also observed, but it may be appropriate to set a new criterion of IgM antibody ID C2.0 for the diagnosis of primary Chlamydophila pneumoniae infection in children. Keywords Chlamydophila pneumoniae Serological diagnosis Enzyme-linked immunosorbent assay Immunoglobulin M Children Primary infection
Introduction Chlamydophila pneumoniae (C. pn) is an important causative pathogen of airway infection [1] that has been suggested to be associated with asthma and arteriosclerosis. Diagnostic methods include detection of the pathogen, i.e., isolation culture and polymerase chain reaction (PCR) test, and measurement of the serum specific antibody titer [2]. Regarding antibody titer measurement, the micro-immunofluorescence (MIF) method is considered to be the world standard [3], but it requires skill to perform the technique and to accurately judge the results. On the other hand, enzyme-linked immunosorbent assay (ELISA) has recently been employed in clinical practice because it requires no complex techniques, and judgments can be made objectively [4]. In Japan, a kit using a purified chlamydial outer membrane complex as the antigen (HITAZYME C. pneumoniae; Hitachi Chemical Co. Ltd., Tokyo, Japan) has been widely used [5], and IgM antibody measurement has been covered by the national health insurance since 2005. However, the positive rate is unexpectedly high, and negative conversion does not occur for a prolonged period when the current diagnostic criteria of C. pn primary infection in children (IgM ID C1.1) is applied, which can
J Infect Chemother (2012) 18:308–312
be problematic [6]. In this study, we re-evaluated the diagnostic criteria of primary infection involving children with airway symptoms, and investigated the prevalence of the infection employing the new criteria.
Patients and methods Patients 1.
2.
The subjects were 136 children (0–3 years of age, 68; 4–14 years of age, 68; 77 boys and 59 girls) who visited the Department of Pediatrics, Koshigaya Municipal Hospital, for acute airway symptoms between July 2005 and June 2010. Samples that were positive for anti-mycoplasma IgM antibody (ImmunoCard M. pneumoniae-specific IgM; Meridian Bioscience Inc., Cincinnati, OH, USA) were excluded. ELISA C. pn IgM ID and MIF C. pn IgM were measured, ROC analysis was performed with MIF considered the standard method, and the candidate values for the ELISA C. pn IgM diagnostic criteria of primary infection were investigated. The prevalence of primary infection was investigated in 3,108 children (0–15 years of age, 1,688 boys and 1,420 girls) who visited our department for airway symptoms (cough with/without sputum, rhinorrhea, wheezing, and fever) between January 2004 and December 2010, based on the criteria established above. When several samples were collected from the same subject, the maximum value was adopted.
Informed consent was obtained from the family of each patient. The study protocol was approved by the institutional Ethics Committee at Koshigaya Municipal Hospital.
309
of p-nitrophenol was measured at a wavelength of 405 nm, which corresponded to the antibody level in the sample. MIF The antibody titers against various Chlamydia strains were measured employing the MIF method reported by Wang and Grayston [7]. Five antigens, i.e., C. pn (TW-183 strain), mixed antigens of three C. trachomatis groups (B ? D ? E, C ? J ? H ? I, and G ? F ? K strains), and C. psittaci (Budgerigar no. 1 strain), were spotted on a multislide, dried at room temperature, and fixed in acetone for 15 min. In the primary reaction, eight- to twofold serial serum dilutions were reacted at 37°C for 2 h. After washing, the slide was dried sufficiently at room temperature, and then the secondary reaction with fluorescein isothiocyanateconjugated goat-labeled anti-human IgM antibody was performed at 37°C for 1 h. Finally, the slide was mounted with 50% glycerin–PBS and a cover glass, and the fluorescence of EB particles for each of the Chlamydia strains was observed under a fluorescence microscope at 4009 magnification. When a fluorescence particle was observed at 32-fold or higher dilution, the serum was judged to be positive. In the statistical analysis, the chi-square test was employed to analyze gender differences and the annual frequency. The t test was employed to analyze differences in the mean, and p values of less than 5% were regarded as significant. ANOVA was employed to compare three groups.
Results 1.
Methods ELISA Antibodies were measured according to the test procedure of HITAZYME C. pneumoniae Ab-IgM. The serum sample was diluted 21-fold with sample processing solution and used as an IgM measurement sample. In the chlamydial outer membrane protein complex antigen-fixed plate, 100 ll of blank, negative, and positive control sera and diluted samples were applied and incubated at 37°C for 1 h. After washing three times, 100 ll of alkaliphosphataselabeled anti-human IgM antibody were added and the mixture incubated at 37°C for 1 h. After washing another three times, 100 ll of substrate solution were added and reacted at 20–25°C for 10 min, followed by the addition of 25 ll of the reaction-stopping solution, and the absorbance
The distribution of ELISA C. pn IgM ID by age is shown in Fig. 1. The geometric means of the ELISA C. pn IgM ID were 1.02 and 1.64 in zero- to three- and four- to 14-year-old children, respectively. The number of patients with upper respiratory infection, bronchitis, pneumonia, and bronchial asthma were 13, 59, 40, and 41, respectively. Upon analyzing of the correlation between ELISA C. pn IgM ID and MIF C. pn IgM (Fig. 2), the correlation coefficient was 0.66. MIF C. pn IgM was positive in 38 and negative in 30 of the 68 zeroto three-year-old patients. Of the 68 four- to 14-year-old patients, MIF C. pn IgM was positive in 42 and negative in 26. In the ROC analysis, when the cut-off value of the ELISA C. pn IgM antibody ID was set at 2.0, the specificity was 100% in both groups, and no false positivity was observed. In the 136 zero- to 14-year-old patients, the maximum (sensitivity ? specificity)/2 was obtained when the cut-off value was set at 1.5. The
123
310
J Infect Chemother (2012) 18:308–312 Table 1 Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of ELISA Chlamydophila pneumoniae IgM, with MIF used as the reference standard MIF C. pn IgM ?
Total -
ELISA C. pn IgM C1.5
57
6
63
\1.5
23
50
73
80
56
136
Total
Sensitivity 71%, specificity 89% Fig. 1 Age distribution of Chlamydophila pneumoniae IgM antibody titers (ELISA)
PPV 90%, NPV 68% ELISA C. pn IgM C2.0
35
0
35
\2.0
45
56
101
80
56
136
Total
Sensitivity 44%, specificity 100% PPV 100%, NPV 55%
Fig. 2 Correlation between titers of Chlamydophila pneumoniae IgM determined by ELISA and by MIF
2.
consistency at each cut-off value is shown in Table 1. Based on these findings, IgM ID C2.0 was regarded as definitely positive and between 1.5 and 2.0 was considered indeterminate as the diagnostic criteria of primary infection. Table 2 presents the number of patients with primary C. pn infection in each age group when ELISA C. pn IgM ID C2.0 was regarded as positive, and the geometric mean ELISA C. pn IgM ID. Of the 3,108 patients, 542 (17.4%) were judged as being C. pn IgMpositive: 247 boys and 295 girls, the mean age was 4.8 years, the geometric mean WBC was 9,084/ll, and the geometric mean CRP level was 1.69 mg/dl. Indeterminate cases (ELISA C. pn IgM ID between 1.5 and 2.0) accounted for 13.2% of all cases. By year, the positive rate was significantly higher in 2008 and 2009 compared with the other years (Fig. 3).
Anti-Mycoplasma IgM antibody was measured in 2,960 patients, and found to be positive in 1,004 (33.9%) and negative in 1,812 (indeterminate in 144). Both ELISA C. pn IgM ID C2.0 and anti-Mycoplasma IgM antibody were positive in 225, ELISA C. pn IgM alone was positive in 255, and anti-Mycoplasma IgM alone was positive in 779. Upon comparing these three groups, there were no
123
significant differences in the age, WBC, or CRP level (Table 3). No significant difference was noted in the ELISA C. pn IgM antibody titer between the patients who were positive for both IgM antibodies and those positive for only ELISA C. pn IgM. Among the patients from whom samples were collected several times (1,550 samples from 596 patients), the antibody titer was measured once or more following the judgment of ELISA C. pn IgM positivity in 107 patients; among these, C. pn IgM converted to negative after three days in the fastest case and after a median of five months. Long-term positivity was observed in ten patients in whom the antibody was positive upon re-testing after more than 12 months. A five-year-boy had ELISA C. pn IgM ID 3.2- and MIF C. pn IgM 128-fold at the first examination. After 21 months, these titers were still positive (2.48- and 32-fold). Eight patients reverted to positive after becoming negative once. IgG or IgA was measured in 2,671 patients, and the diagnostic criterion— an increase of 3 or more—was met by 61 patients (2.3%) and only ten (0.6%) of 1,492 patients aged three years or younger. In paired serum samples, a significant increase was noted in only one patient (a four-year-old girl).
Discussion In a study based on MIF C. pn IgG antibody reported by Ouchi et al. [8], the prevalence of anti-C. pn antibody rose rapidly from about four years of age, and was maintained at about 50% after eight years of age. In a study based on the MIF C. pn IgM antibody [9], C. pn was the causative microorganism in 13.5% of the acute lower airway infection cases in children, showing that the primary infection
J Infect Chemother (2012) 18:308–312
311
Table 2 Age-related seroprevalence of Chlamydophila pneumoniae, as determined by ELISA Age (years)
0
1
2–3
4–6
7–9
10–15
Number of cases
287
678
800
793
296
254
Number of positive cases (%)
9 (3.1)
85 (12.5)
132 (16.5)
173 (21.8)
74 (25.0)
69 (27.2)
Average ELISA C. pn IgM ID (geometric mean)
0.48
0.88
1.14
1.32
1.39
1.43
Fig. 3 Annual occurrence of primary Chlamydophila pneumoniae infections, as determined by ELISA. *p \ 0.01
Table 3 Comparison of the cases judged to be positive with ELISA Chlamydophila pneumoniae IgM and/or ImmunoCard Mycoplasma IgM Both positive
ELISA C. pn IgM-only positive
ImmunoCard Mycoplasma IgM-only positive
Number of patients
225
255
779
Mean age (years)
5.2
4.5
5.0
WBC (/ll) (geometric mean)
8,938
9,079
8,791
CRP (mg/dl) (geometric mean)
1.72
1.63
1.62
C. pn IgM ID (geometric mean)
2.92
2.84
1.05
may be overlooked in children when IgM antibodies are not measured. As the ELISA method, which is simple and correlates well with MIF, has recently entered widespread use, epidemiological surveys of C. pn infection employing ELISA have been performed [3, 5, 10, 11]. In the ELISA method, consistency varies depending on the antigen used [12]. Miyashita et al. [13] reported that ELISA C. pn IgM antibody testing using Ani-Lab-EIA was superior in a multicenter study. HITAZYME-ELISA is widely used in Japan, but problems with the diagnostic criteria, which consider ELISA C. pn IgM ID C1.1 to indicate acute infection, have been pointed out [5, 14]. Niizuma et al. [15] compared the ELISA (Hitazyme C. pneumoniae), MIF, and Western blot methods using the same sera from children with acute airway infection, and observed that the optimum cut-off value of ELISA C. pn IgM ID was 1.4. In our study, to reduce the number of false positive cases, which is a
problem with the current criteria, and to increase the specificity, the cut-off value of ELISA C. pn IgM ID was set at 2.0, and the specificity and positive predictive value improved to 100% in both age groups. Kishimoto et al. [16] recently proposed new criteria in which a value of 2.0 or greater, close to the mean ELISA C. pn IgM ID ?3.0 SD in the criteria for healthy children and adults, was regarded as positive. Since an ELISA C. pn IgM ID C2.0 was regarded as positive based on the above criteria, and the best balance of sensitivity and specificity was obtained at a cut-off value of 1.5 in the ROC curve, an ID of between 1.5 and 2.0 was regarded as indeterminate. Applying these criteria, 17.4% of the children with airway symptoms had primary C. pn infection, and their number increased with age, accounting for 23.5% at four years of age or older. Regarding a gender difference, the prevalence is reported to be higher in males in adults [17], but the positivity rate was significantly higher in girls than in boys in this study. This rate was also close to the frequency of C. pn infection in the study reported by Block et al. [18], in which 48 of 260 pediatric pneumonia patients (18.5%) were MIF C. pn antibody positive. In addition, the positive rate varied between years in this seven-year study. Mycoplasma pneumoniae is recognized as a common pathogen of respiratory infection in children. The ImmunoCard test that detects anti-Mycoplasma IgM antibody by ELISA has found widespread use in Japan. Inutsuka and Yamada [19] reported that ImmunoCard Mycoplasmapositive patients had significantly higher C. pn IgM indices than Mycoplasma-negative patients. In order to clarify the frequency of co-infection with Mycoplasma pneumoniae, we measured anti-Mycoplasma IgM antibody using ImmunoCard. About half of the patients with primary C. pn infection were positive for anti-Mycoplasma IgM antibody, but no significant difference was noted in the ELISA C. pn IgM ID value between those positive and negative for antiMycoplasma IgM antibody. It is considered that crossreactivity between the two pathogens is insignificant. Reportedly, MIF C. pn IgM antibodies appear about three weeks after the primary infection, the titer starts to decrease after two months, it disappears after 4–6 months, and it does not rise during re-infection [20]. Since infantile C. pn infection can be protracted in many cases, although symptoms are mild [21], it is difficult to determine the disease stage on the first test day. In our study, the median duration of positivity was about five months, and long-term
123
312
positivity and reversion to positivity were observed in only a few cases. Very few cases showed significant elevation of IgG and IgA antibody titers, and this may be due to low antibody productivity in childhood [22], suggesting that IgM antibody measurement is essential for the diagnosis of C. pn infection by ELISA in children. The presence of false positive cases is possible even though the cut-off value of the ELISA C.pn IgM antibody titer is elevated, but it may be appropriate to regard HITAZYME ELISA C. pn IgM ID C2.0 as positive for the primary diagnosis of C. pn infection. Acknowledgments We are grateful to Ms. Chikako Mori and Mr. Toshihide Morikawa, Hitachi Chemical Co., Ltd., for their assistance with the ELISA and MIF antibody measurements.
J Infect Chemother (2012) 18:308–312
10.
11.
12.
13.
14.
References 1. Hammerschlag MR. Pneumonia due to Chlamydia pneumoniae in children: epidemiology, diagnosis, and treatment. Pediatr Pulmonol. 2003;36:384–90. 2. Burillo A, Bouza E. Chlamydophila pneumoniae. Infect Dis Clin N Am. 2010;24:61–71. 3. Wang SP. The microimmunofluorescence test for Chlamydia pneumoniae infection: technique and interpretation. J Infect Dis. 2000;181(Suppl 3):S421–5. 4. Verkooyen RP, Willemse D, Hiep-van Casteren SC, Joulandan SA, Snijder RJ, van den Bosch JM, et al. Evaluation of PCR, culture, and serology for diagnosis of Chlamydia pneumoniae respiratory infections. J Clin Microbiol. 1998;36:2301–7. 5. Kishimoto T, Kubota Y, Matsushima T, Izutsu H, Matsumoto A, Soejima R, et al. Assay of specific anti-Chlamydia pneumoniae antibodies by ELISA method. 1. Evaluation of ELISA kit using outer membrane complex. Kansenshogaku Zasshi. 1996;70(8): 821–9 (in Japanese). 6. Miyashita N, Obase Y, Fukuda M, Shoji H, Mouri K, Yagi S, et al. Evaluation of serological tests detecting Chlamydophila pneumoniae-specific immunoglobulin M antibody. Intern Med. 2006;45:1127–31. 7. Wang SP, Grayston JT. Immunologic relationship between genital TRIC, lymphogranuloma venereum, and related organisms in a new microtiter indirect immunofluorescence test. Am J Ophthalmol. 1970;76:367–74. 8. Ouchi K, Kanamoto Y, Ushio M. Prevalence of antibodies to Chlamydia pneumoniae (Strain TWAR) and other Chlamydia in Japan. Kansenshogaku Zasshi. 1991;65(1):19–25 (in Japanese). 9. Ouchi K, Komura H, Fujii M, Matsushima H, Maki T, Hasegawa K, et al. Chlamydia pneumoniae infection and Mycoplasma
123
15.
16.
17. 18.
19.
20. 21.
22.
pneumoniae infection in pediatric patients. Kansenshogaku Zasshi. 1999;73(12):1177–82 (in Japanese). Bamba M, Jozaki K, Sugaya N, Tamai S, Ishihara J, Kori T, et al. Prospective surveillance for atypical pathogens in children with community-acquired pneumonia in Japan. J Infect Chemother. 2006;12(1):36–41. Ciarrocchi G, De Benedetto F, Fogliani V, Magliano E, Del Prete R, Miragliotta G. Serological study on Chlamydophila pneumoniae in patients with community-acquired pneumonia. New Microbiol. 2004;27(4):335–43. Persson K, Boman J. Comparison of five serologic tests for diagnosis of acute infections by Chlamydia pneumonia. Clin Diagn Lab Immunol. 2000;7:739–44. Miyashita N, Ouchi K, Kawasaki K, Nomura H, Kawai Y, Tsumura N, et al. Comparison of serological tests for detection of immunoglobulin M antibodies to Chlamydophila pneumoniae. Respirology. 2008;13:427–31. Miyashita N, Ouchi K, Kawasaki K, Komura H, Kawai Y, Obase Y, et al. Evaluation of enzyme-linked immunosorbent assay for Chlamydophila pneumoniae-specific immunoglobulin M in acute respiratory tract infection. Respirology. 2008;13:299–302. Niizuma T, Kamata A, Kinoshita K, Hisata K, Obinata K. Discussion of positive primary infection criteria for Chlamydophila (Chlamydia) pneumoniae in children: IgM antibody titers by ELISA compared to micro-IF and Western blotting. Kansenshogaku Zasshi. 2008;82(6):619–23 (in Japanese). Kisihimoto T, Ando S, Numazaki K, Ouchi K, Yamazaki T, Nakahama C. Assay of Chlamydia pneumoniae specific IgM antibodies by ELISA method—reduction of non-specific reaction and resetting of serological criteria by measuring IgM antibodies. Jpn J Infect Dis. 2009;62(4):260–4. Leinonen M. Pathogenetic mechanisms and epidemiology of Chlamydia pneumoniae. Eur Heart J. 1993;14(Suppl K):57–61. Block S, Hedrick J, Hammerschlag, Cassell GH, Craft JC. Mycoplasma pneumoniae and Chlamydia pneumoniae in pediatric community-acquired pneumonia: comparative efficacy and safety of clarithromycin vs. erythromycin ethylsuccinate. Pediatr Infect Dis J. 1995;14:471–7. Inutsuka M, Yamada K. The clinical features and the diagnostic problems of Chlamydia pneumoniae infection in pediatric patients. Nihon Syounikagakkai Zasshi. 2008;112(11):1674–9 (in Japanese). Kuo CC, Jackson LA, Campbell LA, Grayston JT. Chlamydia pneumoniae (TWAR). Clin Microbiol Rev. 1995;8:451–61. Normann E, Gnarpe J, Gnarpe H, Wettergren B. Chlamydia pneumoniae in children with acute respiratory tract infection. Acta Paediatr. 1998;87:23–7. Schmidt SM, Mu¨ller CE, Mahner B, Wiersbitzkey SK. Prevalence, rate of persistence and respiratory tract symptoms of Chlamydia pneumoniae infection in 1211 kindergarten and school age children. Pediatr Infect Dis J. 2002;21:758–62.