b r a z j i n f e c t d i s .

2 0 1 4;1 8(2):124–128

The Brazilian Journal of

Original article

Clinical significance of different bacterial load of
Mycoplasma pneumoniae in patients with Mycoplasma
pneumoniae pneumonia

Wujun Jiang, Yongdong Yan ∗ , Wei Ji, Yuqing Wang, Zhengrong Chen
Department of Respiratory Medicine, Children’s Hospital Affiliated to Soochow University, China

a r t i c l e i n f o a b s t r a c t

Article history: Objective: This retrospective study was conducted to investigate the clinical significance of
Received 3 April 2013 different Mycoplasma pneumoniae bacterial load in patients with M. pneumoniae pneumonia
Accepted 12 June 2013 (MP) in children.
Available online 26 October 2013 Methods: Patients with MP (n = 511) were identified at the Children’s Hospital Affiliated to
Soochow University database during an outbreak of MP between January 2012 and February
Keywords: 2013.
Mycoplasma pneumoniae Results: Comparing patients with high and low bacterial load those with higher loads were
Pneumonia significantly older (p < 0.01) and had fever significantly more frequently (p = 0.01). Presence
Polymerase chain reaction of wheezing at presentation was associated with low bacterial load (p = 0.03). Baseline posi-
Serology tive IgM was present in 93 (56.4%) patients with high bacterial load compared to 46 (27.8%)
patients with low bacterial load (p < 0.001). Co-infection with viruses was found signifi-
cantly more frequent among patients with low bacterial load (24.2%) than those with high
bacterial load (8.5%) [p < 0.001]. Bacterial co-infection was also more frequently detected
among patients with low bacterial load (22.4%) than in those with high bacterial load (12.1%)
[p = 0.01].
Conclusion: M. pneumoniae at a high bacterial load could be an etiologic agent of respira-
tory tract disease, whereas the etiologic role of MP at a low bacterial load remains to be
© 2013 Elsevier Editora Ltda. All rights reserved.

M. pneumoniae pneumonia (MP) has been reported in 10–40%
Introduction of community-acquired pneumonia cases.1–4 The mainstay
for diagnosing of M. pneumoniae infection presently relies on
Mycoplasma pneumoniae is an important pathogen respon-
a rise in paired titer by serological methods.5 Polymerase
sible for community-acquired pneumonia in children.
chain reaction (PCR) has been applied for the early detection

Corresponding author at: The Department of Respiratory Medicine, Children’s Hospital Affiliated to Soochow University, No. 303 Jing De
Road, Suzhou 215003, China.
E-mail address: yyd3060@126.com (Y. Yan).
1413-8670/$ – see front matter © 2013 Elsevier Editora Ltda. All rights reserved.

and 100 centrifuged at 12. 1 – Age distribution of our patients. b r a z j i n f e c t d i s . pneumoniae bacterial load IgM antibodies was used as criteria for current M.5%). 60 followed by 30 cycles of 93 ◦ C for 30 s and 55 ◦ C for 45 s. incubated at 100 ◦ C for 10 min. disease onset. pneumoniae infection. influenza viruses A and B.10 age was 35 months (range. the bacterial load in consecutive samples gradually declined in relation to Diagnosis of MP was based on serology or PCR findings.9% NaCl. pneumoniae is of much interest. Real-time PCR was per- formed on the resulting supernatant with 2 ␮L. China) for M.3%). washed twice with 0.. pneumoniae bacterial load using nasopharyngeal aspirates (NPAs) during a recent outbreak of Review of medical records MP in children. A quantitative diagnostic kit (DaAn Gene 511 (35. and distribution of the patients is shown in Fig.000 rpm for five min. nificance of different M. (Virion-Serion.8 Previous study demonstrated that PCR can detect persistent M.1 8(2):124–128 125 of M. and parainfluenza viruses 1. The specimens were centrifuged and were stored at Among 1429 patients with lower respiratory tract infections. Briefly. All nasopharyngeal swabs were tested for antigen detection by 40 immunofluorescence for seven common viruses (RSV. Ltd.0) software was used for all statistical analysis. pneumoniae DNA was findings. clinical features and laboratory data. considered M. characteristics associated with different bacterial loads. pneumoniae infection. pneumoniae IgG titer or the presence of clinical significance of different M.8%) were diagnosed with MP based on serology or PCR Co.9 Thus. 918 patients who were not diagnosed with MP did not differ 1 mL of nasopharyngeal aspirates diluted with 4% NaOH was significantly from patients with MP in terms of sex and age at centrifuged at 12. pneumoniae infection by antigen detec- tion.7 The role of PCR assays per. The age The method is based on the TaqMan PCR technology. pneumoniae Diagnosis of MP infection up to 7 months after disease onset. Data compiled from a retrospective chart review of all study patients were extracted through the use of specially pre- Materials and methods pared data forms without knowledge of serologic status or PCR results. The information extracted included demograph- Patients and specimens ics. We calculated 2012 and February 2013. adeno- virus. USA) as follows: 93 ◦ C for 2 min. Patients hospitalized with clinically and radiologically We used n (%) for categorical variables and median (IQR) for confirmed lower respiratory tract infections (n = 1429) were continuous variables with non-normal distribution or mean identified at the Children’s Hospital Affiliated to Soochow Uni- (SD) for those with normal distribution.1 or if there was ≥4-fold rise in IgG titer. The assay was considered positive formed on upper respiratory tract samples for the diagnosis if IgM ≥ 1. Logis- from these patients. lected. The remaining the target is 16S rRNA gene specific for MP genome. Of the 511 patients with MP. patients without stored serum tic regression analysis was performed to identify clinical samples or results of NPAs were excluded from our study. 1. This study was a retrospective analysis of real-time PCR for Statistical analysis diagnostic testing of M. 0 Serology 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Ages Detection of serum M. 10 cycles of 93 ◦ C for 45 s and 55 ◦ C for 60 s. the significant rise in M. . SPSS Paired serum samples taken on admission and at least seven (version 17. Germany). Therefore. Results Real-time PCR Demographic findings Nasopharyngeal swabs were obtained within 24 h of admis- sion. 304 (59. 2 0 1 4. We assessed differ- versity database during an outbreak of MP between January ences in categorical variables with the 2 test..5. Guangzhou. California. Likewise. DNA detection by real-time PCR was also This study was conducted to investigate the clinical sig. as previously reported. and PCR mix with 43 ␮L (supplied with the kits) using the DA 7600 real-time 80 Number of cases PCR system (Applied Biosystems. blended with 50 ␮L of DNA extraction solution.6.000 rpm for five min. A the time interval from onset of illness to sampling. pneumoniae infection is controversial. 20 2 and 3). the mean used to determine the load of MP. days after the first collection were available for 733 patients (51. one month to 156 months). pneumoniae-specific antibody was performed using enzyme-linked immunosorbent assay Fig. The sediment was col. infection. Both serum and NPAs were obtained 95% CI for differences in medians with an exact test. −80 ◦ C until tested.8 of M.

01) and tive in 330 (64.2% vs. 2 0 1 4. Wheezing at presenta- load in NPAs ranged from <2500 to >2. Fewer Children younger than one year of age had between onset of symptoms and the time of obtaining the first less high bacterial load (17. Patients with high bacterial load also had geometric March 18 (15%) MP patients had high bacterial load. 0 0 0 0 ≤1 –1 –1 >1 –1 Genomal bacterial loads (copies/ml) Fig. 8. p < 0.01). In patients with high bacterial load. geometric mean titers of second IgG significantly increased Outbreaks began in summer months and peaked during compared with the initial IgG in patients more than five years August and September (73.001] (Fig. 12. According to the distribution of genomal bacterial loads (Fig. patients were classified into two groups: Comparison of serological tests by PCR status high bacterial load group (>107 copies/mL) (n = 165) and low bacterial load (<107 copies/mL) (n = 165).1%. PCR for M. sampled material. 0 <1y 1-2y 3-4y >5y Age of patients Discussion Fig. 2). 4.3%) from July patients compared with patients with low bacterial load. 2 – Distribution of different bacterial loads in our patients. with high bacterial load were significantly older (p < 0. significance of real-time PCR in children during an epidemic of . Patients Of the 511 patients with MP. 3). 3 – Different bacterial loads individuals categorized This study investigated the diagnostic values and clinical according to age group. Comparison of clinical characteristics by PCR status Clinical characteristics according to PCR status are summa- Quantitative analysis of M. 93 (56. To assess the independent association of nasopharyngeal aspirates evaluated parameters.1 8(2):124–128 200 120 High bacterial load Low bacterial load No.126 b r a z j i n f e c t d i s .001). Co-infection with viruses was 80 Low bacterial load significantly more frequent among patients with low bacte- PCR negative Percentage 60 rial load than those with high bacterial load (24. titers of second IgG significantly increased in younger patients mer and peaked in August and September. From January to (Fig. fever. 137 (43.7%) from October to December.001). geometric mean of patients with high bacterial load also increased during sum.5% of all MP cases). Out of 330 PCR positive children. Genomal bacterial had significantly higher fever (p = 0. load is shown in Fig. com. No difference was found in the interval (p < 0. Comparison of co-infection by PCR status 100 Co-infections were detected in 204 patients with MP (102 with High bacterial load virus and 102 with bacteria).4%) patients (53. 5).5 × 107 copies/mL of tion was associated with low bacterial load (p = 0. to September. with high bacterial load 46 (27. and 96 (38.8%) had more initial positive IgM compared to patients with low bacterial load (p < 0. mean age.8%) [p < 0.03). pneumoniae was posi. of tested samples 80 PCR negative 150 Number of patients 40 100 0 50 1 2 3 4 5 6 7 8 9 10 11 12 1 2 2012 2013 Fig.7%) than those older than five years serum sample or second serum sample.01). 4 – Monthly occurrence of total episodes of Mycoplasma 0 pneumoniae pneumonia and cases with different bacterial 4 6 5 7 7 0 loads during the outbreaks in 2012. p = 0.001). The frequency old. In patients with low bacterial load.4% vs. mean titers of second IgM significantly increased in younger pared to 52 (27. Epidemiology of MP outbreak Geometric mean titers of initial IgG and IgM of patients with high bacterial load were not significantly different from those Seasonal distribution of MP according to genomal bacterial of patients with low load.5%) from April to June. Bacterial co-infections were also more frequently 40 detected among patients with low bacterial load than in those 20 with high bacterial load (22.6%) nasopharyngeal samples.5%. 179 patients had paired The genomal bacterial load increased significantly with age serum samples. and wheezing at pre- sentation were tested by logistic regression analysis. pneumoniae DNA in rized in Table 1.

5 How. However. the above finding may also suggest that the the less than one-year and one to four-year-old group when presence of respiratory viruses may reduce M.5) 7.8)c.5 ◦ C).d Wheezing at presentation. pneumoniae col- serum paired samples were available. Infections associated with ever. pneumoniae involves cytoad. pneumoniae infection in 2012. high-throughput and rapid method. pneumoniae is weak or deferred high bacterial load was associated with increased humoral in young children because of poor antigenic stimulation from response. infections in age group (3. but inversely associated with mation in human tissues. The Nilsson et al.10. Besides.5 (5. we were unable to confirm whether this From culture based studies.4% in patients greater than three herence between M. In addition. onization. In addition. patients moniae replicates slowly and has limited capacity for protein with low bacterial load did neither have IgM or IgG increase in biosynthesis. Our study also found ing.01 compared with low genomal bacterial load. it is known that the first step happened in older children due to the low proportion of mixed in the pathogenic process of M.7% of the children with MP younger than one MP.4% vs. the data presented here demonstrated that the load was noted mainly in the absence of other respiratory number of children with high bacterial load increased with viral or bacterial infections. particularly in low bacterial load may reflect M. d Mean age. and wheezing at presentation were tested by logistic regression analysis.1 8(2):124–128 127 Table 1 – Clinical characteristics according to PCR status in patients with MP.8% of the chil. suggesting that a low bacterial load in the nasopharynx that positive initial IgM were detected more often in patients seems to be of little clinical significance.8) Fever (Tmax ≥ 38. our study. 2 0 1 4.9) 59 (35. pneumoniae infection presently tive samples gradually declined in relation to the time interval relies on a rise in Ig titers in paired serum samples. 60 Initial titler Initial titler Second titler 60 Second titler Geometric mean IgG Geometric mean IgG 40 40 20 20 0 0 <1y 1-4y >5y <1y 1-4y >5y Age of patients with low bacterial loads Age of patients with high bacterial loads Fig. Various tests have been used to detect MP infection. . pneumoniae at a high load After adherence. M.0) 7. Filled circles indicate geometric mean titers (GMTs) of initial antibody titers and unfilled circles indicate GMTs of second antibody titers in each age group and bars indicate 95% confidence intervals.14 The above findings suggest that wheezing. High bacterial Interestingly. n (%) 56 (30. pneumoniae persistence. c p < 0.0)c. n (%) 111 (61. years).1% of the PCR positive children were seroposi- cal methods have low sensitivity. positive initial IgM and significantly in young children M. serologi. In young children and immunocompromised patients.2 (3. pneu- with high bacterial load (56. b p < 0. 42. viral co-infection and more strongly associated with wheez- dren older than five years had high load.d Duration of prior symptoms (days)a 7. 66. M. b r a z j i n f e c t d i s . Most of these seropositive diagnoses (93 out of 139. fever. involving colonization and inflam- strongly associated with fever. as M. pneumoniae needs to replicate in order was more frequently detected in older children and more to establish an infection.11 In contrast. low bacterial load was detected more often in year had high genomal bacterial load.9 found that the bacterial load in consecu- mainstay for diagnosing of M.8%). 27.d a Data expressed as means (SD).13 occurred in children with high bacterial load.05 compared with low genomal bacterial load. n (%) 141 (77.12. pneumoniae colonization is usually weak increased second antibody titers were detected more often which leads to poor antigenic stimulation.6 (2.4) 135 (81.9%) to be a sensitive. another prospective. which indicated that why antibody response to M. M. from onset of illness to sampling. Only 17. 5 – Correlation between initial and second antibody titers among patients with different bacterial loads. These could explain in patients with high bacterial load.3) 97 (58. suggesting a causative role of the age.8) 26 (15. but 53. Characteristics Genomal bacterial load PCR negative (n = 181) Low load (n = 165) High load (n = 165) Male:female (n) 109:72 100:65 95:70 Mean age (months)a 26 (29) 29 (31) 49 (54)b. However.7) 127 (77. PCR seemed tive.7) Antibiotic therapy within 2 weeks preceding admission. during the early phase of MP infection. pneumoniae and respiratory epithelium.9) 81 (84.

Carquin J. particle agglutination antibody test for the diagnosis of Mycoplasma pneumoniae pneumonia in children during two outbreaks.15:1630–6.52:125–9. therefore. Etiology of community-acquired Hardy RD. bacterial load remains to be determined. Persson K. Thurman KA. Warner AK. 2008. 11. et al. prescribing Mycoplasma pneumoniae pneumonia in pediatric patients. 2004. Indian J Med Res. Eun BW. culture & serological tests for the diagnosis of community-acquired lower respiratory tract infections. Diagn Microbiol Infect Dis. 2006. Steiner F. et al. Assessment of real-time PCR for diagnosis of sary use of macrolides. Blasi F. J Microbiol Methods. et al. considered. 14. The patients were is superior to serology for the diagnosis of acute Mycoplasma also more likely to have received antibiotic treatment due to pneumoniae infection and reveals a high rate of persistent infection. 2009.8:93. Pediatr Infect Dis J. Kashyap B. Roivainen M. population from 12. Tan JJ. Kumar S. Lozano J. Harrison C. The management of community-acquired pneumonia in infants and children older than 3 months of The findings of the present study have some limitations. Olsen K. Rapid detection of Mycoplasma pneumoniae in clinical samples by potential inability to distinguish between respiratory car- real-time PCR. bacterial load and serologically negative should be carefully 8. Epidemiological and clinical 16. According to our study. respiratory illnesses. McCracken GH. et al. 2011. concerns exist about the 6. Park KW. age: clinical practice guidelines by the Pediatric Infectious The patients included were hospitalized cases of pneumo. et al. Diagnostic utility and clinical significance of naso- pneumonia in 254 hospitalized children. Mitchell SL. sis may. Saigal SR. Baughman AL. 2008. 2008.17 Prescribing macrolides in the 7. Michelow IC. Carter ER. British Thoracic Society guidelines for the management 2004. Dillon MT. Coote N. procedures for detection of Mycoplasma pneumoniae in et al. Expert Rev Anti Infect Ther.41:45–51. Ziegler T. Clin Infect Dis. Nilsson AC. 4.39:15–9. in clinical specimens using a single-tube multiplex real-time PCR assay. Leinonen M. Peng B. 12. Lee KY. Dutly F. Diseases Society and the Infectious Diseases Society of nia. Fletcher P. Walter ND. Bessaci-Kabouya K.26: Conflicts of interest 897–903. Zhang R. Rollins NK.53:e25–76. Thurman KA. Villenet N. 2008. Kim NH. McKean M. Wang XB. Michelow IC.70:1–9. whereas the etiologic role of MP at a low Diagn Microbiol Infect Dis. and Legionella spp. be important for studies of M. Morozumi M. Shin SH. Olsen K. Das BC. Meurman O. Principi N. Polymerase chain reaction symptoms may have been overrepresented. 2000. Quantitative analy. Yuan ZJ.15–17 Also. et al. Murayama SY. Hardegger D.66 Suppl 2:ii1–23. Iwata S. et al.32:1281–9. Lee JA. 2011. Chung EH. Waris M.128:134–9. 17. Ito A. They could not of community acquired pneumonia in children: update 2011. patients with more severe 9. riage vs. Therefore. He XY. Clark J. Winchell JM. Cowart KC. Alverson B. . Clin Mycoplasma pneumoniae in community-acquired lower Infect Dis. 2001. Nolevaux G. community-acquired pneumonia. J Clin Microbiol. Bjorkman P. Mertsola J. tory tract disease. Zucol F. Bossart W. pneumoniae at Investigation of Mycoplasma pneumoniae infection in pediatric a high bacterial load could be an etiologic agent of respira. Comparison of laboratory diagnostic 1. Lips U.025 cases with respiratory infection. community-acquired pneumonia in hospitalized children. Berger C. our results suggest that M. 2013. Harris M.48:1244–9. Pediatr Infect Dis J. distinguish between viable and nonviable organisms after Thorax.6:509–21.1 8(2):124–128 PCR tests also have their drawbacks. Detection of Mycoplasma pneumoniae. references 2011.75:22–7. 2007. Benitez AJ. Diagn Microbiol Infect Dis. Byington CL. Clin Infect Dis. In conclusion. Role of Mycoplasma 15. Chlamydia pneumoniae. Laplanche D. Duffy LB. 2001. 13. Pediatric respiratory infections by Mycoplasma Pediatrics. Nadal D. Shah SS. Schwartz SB.19:293–8. Arch Pediatr. Harnden A. pneumoniae.113:701–7. The authors declare no conflicts of interest. Esposito S. Lozano J. pneumoniae Comparison of polymerase chain reaction and the indirect infection. Nadal D. Mycoplasma pneumoniae infection in children with 5. Allegra L. 2. and oropharyngeal samples used in a PCR assay to diagnose 2000. 3. the clinical manifestations. Andreoletti L. Comparison pneumoniae and Chlamydia pneumoniae in children with of PCR.42:3339–41. BMC Microbiol. seminested 16S rDNA-PCR. Bradley JS.128 b r a z j i n f e c t d i s . 2 0 1 4. Sethi GR. Duffy LB. study of Mycoplasma pneumoniae respiratory infections in Community-acquired pneumonia in children due to children hospitalized in a pediatric ward between 1999 and Mycoplasma pneumoniae: diagnostic performance of a 2005 at the Reims University Hospital. presence of a positive PCR assay may increase the unneces. active disease. respiratory tract infections in children. Epidemiology and clinical characteristics of community outbreaks. et al. Can macrolides based on a positive PCR in patients with low J Microbiol. antibacterial therapy. et al. We did not enroll patients with other spectrum of acute America. Hasegawa K. 10. Bossart W. Kobayashi R. Altwegg M. Juven T.