ABSTRACT
Mycoplasma pneumoniae is an important causative pathogen of community-acquired pneumonia in children. Rapid and reliable laboratory diagnosis of M. pneumoniae infection is important so that appropriate antibiotic treatment can be initiated to reduce the misuse of drugs and resistance rates. Anti-M. pneumoniae immunoglobulin M (IgM) is an indicator of recent primary infection but can persist for several months after initial infection. It has been suggested that anti-M. pneumoniae immunoglobulin A (IgA) can be a reliable indicator for recent M. pneumoniae infection in adults. We investigated the clinical diagnostic value of M. pneumoniae IgA in school-age children and adolescents with M. pneumoniae-related pneumonia. Eighty children with pneumonia and seropositive for M. pneumoniae IgM or with a 4-fold increase of anti-M. pneumoniae immunoglobulin G (IgG) were enrolled from May 2015 to March 2016. The titers of M. pneumoniae IgA, IgM, and IgG, the clinical features, and laboratory examinations of blood, C-reactive protein, and liver enzymes were analyzed. The initial positivity rates for M. pneumoniae IgM and IgA upon admission to the hospital were 63.6 and 33.8%, respectively. One week after admission, the cumulative positivity rates for M. pneumoniae IgM and IgA increased to 97.5 and 56.3%, respectively. Detection of M. pneumoniae IgM was more sensitive than detection of M. pneumoniae IgA for the diagnosis of M. pneumoniae-related pneumonia in school-age children and adolescents; however, paired sera are necessary for a more accurate diagnosis.
INTRODUCTION
Mycoplasma pneumoniae is one of the most common atypical pathogens causing pneumonia, especially in school-age children and adolescents. Cases occur in all seasons but predominantly in the spring and autumn. Although a significant cough without rhinorrhea is the typical manifestation, the symptoms can vary from an asymptomatic respiratory infection to severe pneumonia (1). M. pneumoniae is generally susceptible to macrolide or tetracycline antibiotics, but these treatments are not often immediately used for cases of community-acquired pneumonia if no optimal detecting tool for M. pneumoniae exists. Although M. pneumoniae is also susceptible to fluoroquinolone, its utilization in children is limited due to its side effect of joint toxicity (2). In recent years, macrolide-resistant M. pneumoniae has become more and more prevalent across the world, partly due to the wide use of macrolides (3). The macrolide resistance rate of M. pneumoniae is about 10 to 20% in Taiwan (4, 5), and upward of 90 to 100% in China, the neighboring country of Taiwan (3). Therefore, an optimal detection method is needed to allow for a more precise diagnosis, and thus the appropriate treatment of patients with M. pneumoniae-related pneumonia, and reduce the misuse of antibiotics.
M. pneumoniae is difficult to culture and has a very low growth rate, and thus culture-based approaches are usually not helpful clinically. The clinical use of PCR assays for the diagnosis of M. pneumoniae respiratory infections has been reported in both adults (6–9) and children (10–14). However, inducing sputum samples is difficult in young children, and some reports have shown no obvious differences between M. pneumoniae PCR positivity rates of throat swabs from individuals with suspected cases of M. pneumoniae infection and those of healthy subjects (15). This suggests that M. pneumoniae may colonize the upper airway in humans without inducing disease. Since the pathogen-induced immune response results in the production of antibodies, serological testing is a major diagnostic tool for the detection of M. pneumoniae infection. The complement fixation test has also been used; however, the complex nature of the technique and its high false-positivity rate limit its use (16, 17). The detection of M. pneumoniae IgM has been used to determine the presence of an acute infection. However, some patients do not have sufficient levels of M. pneumoniae IgM antibodies to allow for detection during the early stages of acute infection or reinfection (18), and some patients may have detectable levels of M. pneumoniae IgM for several months after M. pneumoniae infection (19). Both of these situations can lead to difficulties in making a clinical diagnosis. Only a few studies have reported on the use of M. pneumoniae IgA antibodies for the diagnosis of M. pneumoniae infection in the past 2 decades. One study reported that detection of M. pneumoniae IgA is more sensitive than that of IgM in diagnosing acute M. pneumoniae-related pneumonia in adults (20), whereas another reported lower sensitivity in detecting M. pneumoniae IgA in young children (21). However, M. pneumoniae-related pneumonia often occurs in school-age children and adolescents. To date, no study has focused on the diagnostic value of M. pneumoniae IgA in school-age children. Therefore, the aim of this study was to investigate the clinical significance and usefulness of detecting M. pneumoniae IgA in school-age children and adolescents with community-acquired pneumonia caused by M. pneumoniae.
RESULTS
Clinical characteristics of enrolled patients.A total of 89 children with M. pneumoniae IgM-positive serum were enrolled in this study, of whom nine (10.1%) were excluded due to persistently low positive M. pneumoniae IgM titers without a 2- or 4-fold increase in M. pneumoniae IgG titers after 2 weeks of follow-up. The remaining 80 children (seroconversion, n = 26; IgM 2-fold increase: n = 72; IgG 4-fold increase: n = 31 [Fig. 1]) (31 males, 49 females; mean age, 6.6 ± 3.3 years old [range, 3 to 17 years old]) were entered into the analysis, and 215 serum samples were obtained. The duration of fever (≥38°C) before admission was 4.3 ± 2.7 days, as reported by the patient or a family member. The clinical characteristics, including age, duration of fever, hospital stay, and laboratory analysis of blood, liver enzymes, and C-reactive protein (CRP) are presented in Table 1.
Venn diagram showing serologically diagnostic criteria for the 80 children with M. pneumoniae-related pneumonia. The intersection areas among these three circles show cases that simultaneously fit the different diagnostic criteria.
Clinical characteristics and laboratory results for blood, liver function, and C-reactive protein for 80 patients with M. pneumoniae infectiona
Detection of M. pneumoniae IgM was more sensitive than M. pneumoniae IgA in school-age children with M. pneumoniae-related pneumonia.Among the 80 cases, the positivity rates for M. pneumoniae IgA, IgM, and IgG upon admission were 33.8% (27/80), 63.8% (51/80), and 32.5% (26/80), respectively (Fig. 2). Of note, 21 cases (26.2%) were initially serum negative for M. pneumoniae IgA, IgM, and IgG. A chi-square test showed significant differences between the initial positivity rates of M. pneumoniae IgA and IgM (P = 0.019). Then, we analyzed the correlation between initial M. pneumoniae IgA, IgM, and IgG values and the duration of fever before hospitalization, and the results showed that M. pneumoniae IgM titers, but not IgA or IgG titers, were significantly positively correlated with the duration of fever before hospitalization (r = 0.377, P = 0.002) (Fig. 3). In order to determine the relevance between the M. pneumoniae antibody titer and fever duration before admission, we divided the patients into three groups according to the duration of fever before hospitalization to compare the positivity rates of M. pneumoniae IgA and IgM as follows: (i) 0 to 3 days, (ii) 4 to 6 days, and (iii) 7 to 14 days. Most of the patients (40% [32/80]) had a fever duration of fewer than 3 days before hospitalization. Twenty-seven patients (33.8%) and 21 patients (26.3%) had a fever duration of 4 to 6 days and 7 to 14 days before hospitalization, respectively. As shown in Fig. 4, both M. pneumoniae IgA and IgM positivity rates were higher in patients with a longer fever duration before hospitalization. In addition, the serum positivity rate of M. pneumoniae IgM was higher than that for M. pneumoniae IgA in all groups (Fig. 4), suggesting that the detection of M. pneumoniae IgM is more sensitive than M. pneumoniae IgA detection at any stage of disease for school-age children with M. pneumoniae-related pneumonia.
Venn diagram showing the case numbers initially positive for Mycoplasma pneumoniae IgA (27, 33.8%), IgM (51, 63.8%), and IgG (26, 32.5%) upon hospital admission. Twenty-one cases were negative for M. pneumoniae IgA, IgM, and IgG upon admission.
Correlation of initial M. pneumoniae IgA, IgM, and IgG values with duration of fever upon admission. M. pneumoniae IgM but not IgA or IgG values were significantly positively correlated with the duration of fever (Pearson's correlation coefficient: IgA, r = 0.240, P = 0.052; IgM, r = 0.377, P = 0.002; and IgG, r = −0.016, P = 0.902).
Positivity rates for M. pneumoniae IgA, IgM, and IgG with different durations of fever upon admission. The patients were divided into three groups based on the duration of fever upon admission (0 to 3 days, 4 to 6 days, and 7 to 14 days); all of them had a higher positivity rate for M. pneumoniae IgM than for M. pneumoniae IgA.
Paired serum follow-up is needed for patients who are initially M. pneumoniae IgM negative but clinically symptomatic.We next compared the cumulative positivity rates for M. pneumoniae IgA, IgM, and IgG within 1 month of hospitalization. Since the patients were at different time points in the disease course on admission to the hospital, blood samples were taken according to their clinical condition but not on the same schedule. In order to organize the serological data of the patients, blood sample data were divided into five groups according to the hospitalization stay after admission: (i) baseline (the day of admission), (ii) 2 to 4 days, (iii) 5 to 7 days, (iv) 8 to 14 days, and (v) 15 to 28 days. The initial M. pneumoniae IgM positivity rate for all 80 patients was 63.8% (51/80) (Fig. 5A). The cumulative positivity rates for M. pneumoniae IgM for groups 2 and 3 after admission were 85.0 and 97.5%, respectively. Twenty-six of the 80 cases (32.5%) showed seroconversion of M. pneumoniae IgM 1 week after admission. These findings suggested that paired serum follow-up samples are needed for patients who are initially M. pneumoniae IgM negative but are clinically symptomatic. The cumulative positivity rate of M. pneumoniae IgM was always higher than that of M. pneumoniae IgA, even though 25 of the 80 patients were still serum negative for M. pneumoniae IgA 15 to 28 days after admission.
Cumulative positivity rate of M. pneumoniae IgA, IgM, and IgG. (A) The cumulative positivity rate for M. pneumoniae IgA, IgM, and IgG among all 80 patients was calculated for five time intervals after admission: 0, 2 to 4, 5 to 7, 8 to 14, and 15 to 28 days. (B) Cumulative positivity rates for M. pneumoniae IgA, IgM, and IgG for 26 cases with seroconversion. Twenty-six cases were initially negative for M. pneumoniae IgM and then experienced seroconversion after 2 weeks. Only four cases (15.4%) were initially positive for M. pneumoniae IgA. The cumulative positivity rate for M. pneumoniae IgM was higher than for IgA after 2 days (61.5% versus 46.2%, respectively).
Among the 80 patients with pneumonia due to M. pneumoniae, 26 (32.5%) who were initially negative for M. pneumoniae IgM upon admission showed seroconversion after 2 weeks of follow-up (Fig. 5B). These findings showed the natural course of M. pneumoniae antibodies in these patients. Among the 26 patients who were initially negative for M. pneumoniae IgM, four (15.4%) were initially positive for M. pneumoniae IgA; however, the cumulative positivity rate for M. pneumoniae IgM was also higher than that for M. pneumoniae IgA after 2 to 4 days of follow-up (61.5% versus 46.2%) in all enrolled cases (Fig. 5B). These findings also suggest the clinical importance of paired serum follow-up.
We also analyzed the levels of M. pneumoniae IgA, IgM, and IgG after 1 month of follow-up, and all of the patients were still positive for all antibodies after 15 to 28 days of follow-up (Fig. 6).
Titers of M. pneumoniae IgA, IgM, and IgG after follow-up for 1 month. The means and standard errors of the M. pneumoniae IgA, IgM, and IgG titers were classified into five time intervals: 0, 2 to 4, 5 to 7, 8 to 14, and 15 to 28 days after admission. None of the titers showed any significant decline after 1 month of follow-up.
Fever duration before admission and the initial platelet count were positively correlated with the initial M. pneumoniae IgM value.In order to establish what factors can influence the M. pneumoniae IgM titer, we analyzed the correlation between the patients' initial M. pneumoniae IgM value, their clinical features, and the results of the laboratory analysis of blood, liver function, and CRP using Pearson's correlation analysis. We found that fever duration before admission and initial platelet count were positively correlated with initial M. pneumoniae IgM titers (Table 1 and Fig. 7A and B). The associations between the clinical characteristics, laboratory analysis of blood cells, liver enzymes and CRP, and M. pneumoniae IgM titer were then tested again using multivariate logistic regression. This result also indicated that fever duration before admission and the platelet count were positively correlated with M. pneumoniae IgM (P = 0.005 and P = 0.001, respectively). However, the correlation between the fever duration before admission and platelet count was not significant (r = 0.075, P = 0.638) (Fig. 7C). This suggests that both fever duration before admission and the initial platelet count are independent factors correlated with M. pneumoniae IgM titer.
Correlation between the duration of fever before admission, the platelet count, and the level of M. pneumoniae IgM. (A) There was no significant correlation between fever duration before admission and platelet count (P = 0.638). (B) There was significant correlation between M. pneumoniae IgM and platelet count (P = 0.004). (C) There was significant correlation between M. pneumoniae IgM and fever duration before admission (P = 0.014).
DISCUSSION
In this study, we investigated longitudinal changes in M. pneumoniae IgA, IgM, and IgG titers in school-age children with pneumonia caused by M. pneumoniae infection. We found that the initial positivity rates for M. pneumoniae IgM and IgA upon admission to hospital were 63.8 and 33.8%, respectively. The duration of fever before admission and platelet count were positively correlated with initial M. pneumoniae IgM titers, and the cumulative positivity rates for M. pneumoniae IgM and IgA increased to 97.5 and 56.3%, respectively, 1 week after admission. Detection of M. pneumoniae IgM was more sensitive than detection of M. pneumoniae IgA in the diagnosis of school-age children and adolescents with pneumonia caused by M. pneumoniae.
The progression of Mycoplasma infection is often gradual and slow and, in adults, the illness is usually present for more than 1 week before the patient seeks medical assistance (20). However, our results showed that children with pneumonia caused by M. pneumoniae were mostly admitted to hospital within the first week of infection (mean, 4.3 ± 2.7 days). This may be because fever is the major symptom in children rather than the prolonged cough typically seen in adults infected with M. pneumoniae. Children are therefore brought to the hospital sooner than adults due to the concerns of the caregiver. For this reason, we found that many cases (21/80 [26.3%]) were negative for M. pneumoniae-specific antibodies upon admission. In order to avoid missing false-negative cases, we collected paired sera for M. pneumoniae IgA, IgM, and IgG after admission. However, since fever duration before admission was reported by the patients or family members, there is a limitation for the accuracy of this value. Beersma et al. reported high background rates of seropositivity for M. pneumoniae IgM in some adult populations, ranging from 0 to 51% depending on the assay used (22). In the present study, nine children were excluded due to persistently low positive M. pneumoniae IgM titers with no obvious increase in the paired sera. The background seropositivity rate for M. pneumoniae IgM was 10.1% in the children in this study.
In previous studies, the detection of M. pneumoniae IgA has been reported to be a good indicator of recent M. pneumoniae infection in predominantly adult populations (23–25). Detection of M. pneumoniae IgA has been reported to be more sensitive than detection of M. pneumoniae IgM for the diagnosis of M. pneumoniae infection in adults (20). However, conflicting results have been reported in toddlers, for whom M. pneumoniae IgA was reported to be a poor indicator of M. pneumoniae infection (21). In this study, we investigated the diagnostic value of M. pneumoniae IgA in diagnosing pneumonia caused by M. pneumoniae in school-age children, the most common age for infection with this pathogen. Our results showed positivity rates for M. pneumoniae IgM and IgA on the first day of hospital admission of 63.8 and 33.8%, respectively, and these rates were positively correlated with the duration of fever before admission (Fig. 2). We also found that the positivity rate for M. pneumoniae IgM was higher than that of M. pneumoniae IgA for each duration of fever before admission group (Fig. 3). This suggests that detection of M. pneumoniae IgA is less sensitive than detection of M. pneumoniae IgM for the diagnosis of M. pneumoniae-related pneumonia in school-age children and adolescents. The reason that M. pneumoniae IgA is a good marker in adults but not in children with M. pneumoniae-related pneumonia is unclear. The developing immune system may partly explain this phenomenon. Human serum IgA and IgM concentrations change with age, with lower concentrations at birth. Serum IgM reaches an adult concentration at 4 to 5 years of age, whereas serum levels of IgA mature slowly until puberty (26, 27). Additional studies are needed to investigate this issue further.
In the present study, the positivity rate for M. pneumoniae IgM upon hospital admission was 63.8%, with a mean duration of illness of 4.3 ± 2.7 days prior to admission. This is consistent with another study in which the M. pneumoniae IgM positivity rate in children was 62.2% in the first week (28) and ca. 70.9 to 81.8% in the second week (29, 30) of the illness. The sensitivity of serological tests can be affected by the timing of specimen collection, the standard of diagnosis, and the assay used (31). This could explain the higher sensitivity of M. pneumoniae IgM in patients with a longer duration of illness prior to admission (Fig. 2B and Fig. 3). We found a positive correlation between the M. pneumoniae IgA positivity rate and the duration of illness prior to admission, although this correlation was less significant than that for M. pneumoniae IgM (Fig. 2A).
A 4-fold increase in M. pneumoniae IgG titers in acute- and convalescent-phase sera is considered to be the gold standard for the diagnosis of acute M. pneumoniae respiratory infection (32). The rate of 4-fold increase in M. pneumoniae IgG in paired sera varies in different studies. Medjo et al. reported that 90% of cases had a 4-fold increase in M. pneumoniae IgG in patients with positive throat swabs via PCR (29). In comparison, Ma et al. reported that only 2.4% of cases had a 4-fold increase in M. pneumoniae IgG in patients with positive IgM or positive nasopharyngeal swabs via PCR (30). In the present study, 31 cases (38.8%) had a 4-fold increase in M. pneumoniae IgG in paired sera. Since M. pneumoniae IgG in paired sera cannot provide an instant diagnosis and as obtaining second blood samples from children can be difficult in the convalescence phase, M. pneumoniae IgG does not seem to be an optimal tool for the early diagnosis of M. pneumoniae infection.
In addition to the duration of fever before admission, we found a positive correlation between M. pneumoniae IgM titers and platelet count upon hospital admission. This may be explained by the production of immunoglobulin after an illness. However, this correlation has not previously been reported in the literature. It has been reported that chronic inflammation is often associated with reactive thrombocytosis (33). Youn et al. reported a correlation between platelet count and serum status of M. pneumoniae IgM, where the platelet count was higher in a subgroup with increased M. pneumoniae IgM than in those with seroconversion (34). In the present study, the platelet count was also higher in the subgroup with increased M. pneumoniae IgM levels than in that of the seroconverted cases (292,090 ± 73,530/μl versus 227,850 ± 63,730/μl, respectively; P < 0.001). Furthermore, we found that the platelet count was significantly positively correlated with the M. pneumoniae IgM titer, which could explain the lower platelet count in the seroconverted cases.
In conclusion, detection of M. pneumoniae IgM was more sensitive than detection of M. pneumoniae IgA for the diagnosis of M. pneumoniae-related pneumonia in school-age children and adolescents. If only one serum sample for M. pneumoniae IgM was tested at initial admission, there was a 10.1% false-positivity rate and a 32.5% false-negativity rate. For a more accurate diagnosis of M. pneumoniae-related pneumonia in school-age children, testing of paired sera of M. pneumoniae IgM is still necessary.
MATERIALS AND METHODS
Patients.We enrolled patients aged 3 to 18 years who were admitted to Kaohsiung Chang Gung Memorial Hospital between May 2015 and March 2016. All of the patients had a cough, fever, and chest X-ray-confirmed bronchopneumonia or lobar pneumonia, suggesting a lower respiratory tract infection. Fever was defined as a body temperature of ≥38.0°C, as measured by infrared tympanic membrane thermometers. All patients were diagnosed serologically, and exhibited seroconversion or an increase in antibody titers (2-fold increase of serum M. pneumoniae IgM or a 4-fold increase in M. pneumoniae IgG) for at least 2 weeks after admission (35). The patients enrolled in this study were otherwise healthy without any underlying conditions that might have altered their clinical course of disease.
Methods.All patients received macrolide antibiotic treatment, after which all of the patients improved clinically. The serum levels of M. pneumoniae IgA, IgM, and IgG were determined upon admission and at various time points during or after admission depending on their clinical condition. Serum M. pneumoniae IgM and IgG were assayed using ELISA ImmunoWELL kits (Ben-Bio, San Diego, CA), and serum M. pneumoniae IgA was assayed using CHORUS kits (Diesse Diagnostica Senese, Siena, Italy) according to the manufacturers' instructions. The positive cutoff values for M. pneumoniae IgA, IgM, and IgG were 18 arbitrary units (AU)/ml (with upper and lower limits of 10 and 100 AU/ml), 950 AU/ml, and 320 AU/ml, respectively. The Institutional Review Board of Kaohsiung Chang Gung Memorial Hospital approved the entire study (103-7061A3). The values and positivity rates for M. pneumoniae IgA, IgM, and IgG were correlated with the clinical features. Continuous variables are expressed as means ± standard errors. The chi-square test was used to determine differences in initial positivity rates of M. pneumoniae IgA and IgM. Statistical correlations between M. pneumoniae IgM titers, clinical characteristics, and the results of laboratory analysis of blood, liver function, and C-reactive protein (CRP) were determined using Pearson's correlation test and multivariate logistic regression. All statistical tests were performed using SPSS version 19.0 for Windows (SPSS, Inc., Chicago, IL). A P value of <0.05 indicated statistical significance.
ACKNOWLEDGMENTS
This study was supported in part by grants CMRPG8E0011 (W.-J.L.) and MOST 105-2314-B-182-051-MY2 (H.-R.Y.) from the Ministry of Science and Technology, Taiwan. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
FOOTNOTES
- Received 1 October 2016.
- Returned for modification 4 October 2016.
- Accepted 7 October 2016.
- Accepted manuscript posted online 19 October 2016.
- Copyright © 2017 American Society for Microbiology.
REFERENCES
