The Incidence and Risk Factors of Extrapulmonary Manifestations in Mycoplasma Pneumoniae Pneumonia

Background: Mycoplasma pneumoniae pneumonia (MP) is a major cause of community acquired pneumonia (CAP) in children and it is known to be associated with extrapulmonary manifestations (EPM). The incidence and risk factors of EPM in children are not known. Methods: This is a retrospective study involving 65,243 pediatric CAP patients between 2010 and 2015 at 23 nationwide hospitals was conducted in South Korea. The medical records were reviewed to collect the information regarding the clinical characteristics, radiological results, and laboratory ndings. In total, 9,190 children with MP were identied and included in the analysis. Logistic regression with multivariable analysis was performed to evaluate the risk factors associated with EPM in MP. Results: The mean age of the enrolled patients with MP was 64.3±39.8 months, and the proportion of male patients was 49.5%. The incidence of EPM was 23.9% and included elevation of liver enzymes (18.1%), mucocutaneous manifestations (4.4%), proteinuria (4.1%), cardiovascular and neurologic manifestation (0.4%), hematologic manifestation (0.2%) and arthritis (0.2%). Statistical analysis showed that mucocutaneous manifestations were signicantly increased with elevated alanine aminotransferase (adjusted odds ratio [aOR] 3.623, 95% condence intervals [CI] 1.933-6.790) and atopic sensitization (aOR 2.973, 95% CI 1.615-5.475) and decreased with respiratory virus co-infection (aOR 0.273, 95% CI 0.084-0.887). Elevated liver enzymes was signicantly associated with the elevation of lactate dehydrogenase (aOR 3.055, 95% CI 2.257-4.137) and presence of pleural effusion (aOR 2.635, 95% CI 1.767-3.930) and proteinuria with respiratory virus co-infection (aOR 2.245, 95% CI 1.113-4.527). Conclusions: About 24% of pediatric MP patients were identied with various EPM.


Introduction
Mycoplasma pneumoniae is one of the most common pathogens responsible for community acquired pneumonia (CAP) in children, which is characterized by fever, cough, and sputum and considerable portion of children need hospitalization [1,2]. Mycoplasma pneumoniae pneumonia (MP) epidemic occurs in Korea about every 3 to 4 years [3], and could lead to increased burden in pediatric community. The characteristics of MP are various from self-limiting to having long-term sequelae such as bronchiolitis obliterans and bronchiectasis [4]. In addition, macrolide-resistant MP (MRMP) is emerging issue these days [5][6][7].
Besides respiratory symptoms, MP pathogenesis can also involve extrapulmonary manifestations (EPM) that affect almost all the body organs and are associated with mucositis, Stevens-Johnson syndrome (SJS), rash, tendinitis, and central nerve system infection. Mycoplasma infection can lead to several EPM such as proteinuria, acute hepatitis, myocarditis, Kawasaki disease, peripheral neuropathy, Guillain-Barré syndrome and hemophagocytic syndrome [8-10]. The incidence of EPM and the associated risk factors have not been extensively studied. Furthermore, although MP is a common cause for CAP, the mechanism of MP-driven periodic epidemics and etiology of various clinical characteristics, including EPM, are not yet understood leading to di cult treatment.
Macrolides are the rst-choice antibiotic agents for the treatment of MP in children. However, MRMP are increasing abruptly [5][6][7] resulting in clinical deterioration despite treatment [11][12][13][14]. MRMP could be treated with tetracycline and quinolones; however, their use in pediatric patients is limited due to safety concerns, such as permanent tooth discoloration by tetracycline and side effects involving muscles, tendons or joints by quinolones. It has also been reported that MRMP could be a risk factor for the development of EPM [11]; therefore, careful observation is necessary in MRMP patients.
In this study, we aimed to study clinical aspects of pediatric MP patients and to evaluate the incidence of EPM and associated risk factors.

Study population
This study was conducted based on the data collected from 23 medical centers on pediatric CAP patients hospitalized between January 1, 2010 and December 31, 2015. The data were collected in cooperation with secondary and tertiary medical centers under the 'Pneumonia and Respiratory Diseases study group' of the Korean Academy of Pediatric Allergy and Respiratory Disease [3]. Data from a total of 65,243 pediatric CAP patients under the age of 18 years were collected by a retrospective chart review. Of these, patients of 9,190, diagnosed with MP, were included in this analysis.
To identify differences in incidence of EPM according to age, two age groups were considered; (1) preschool children (< 60 months old) and (2) schoolchildren (≥ 60 months old).
Clinical characteristics, including respiratory symptoms with or without intensive treatment, underlying diseases, prescribed drugs, and laboratory and radiological ndings were collected from a retrospective chart review of medical records.
The study protocol was approved by the Institutional Review Boards of all participating medical centers. This study was approved by the Institutional Review Board and Ethics Committee of Soonchunhyang University Seoul Hospital (SCHUH201-309013001).
De nition of Mycoplasma pneumoniae pneumonia Pneumonia was diagnosed by pediatricians based on both physical examinations and radiological assessments [1]. MP was considered when: (1) a 4-fold or greater increase in IgM and/or IgG antibody titers between acute and convalescent stages was observed and/or (2) polymerase chain reaction (PCR) showed positive results for mycoplasma in nasopharyngeal aspiration or sputum samples [5,15].
The response of patients with MP to macrolide treatment was divided into three categories based on the fever duration in each pneumonia episode after the initiation of macrolide treatment, regardless of the macrolide sensitivity test results. We de ned clinical macrolide-sensitive MP (MSMP), macroliderefractory MP (MRMP), and macrolide-less effective MP (MLMP) as fever for ≤ 3 days after the macrolide treatment, > 7 days, and >3 days but ≤ 7 days, respectively [3].

Laboratory and radiologic studies
Real-time PCR (RT-PCR) analyses were performed to identify the causative respiratory viruses; adenovirus, human rhinovirus, in uenza virus, parain uenza virus, human metapneumovirus, respiratory syncytial virus, bocavirus, and human coronavirus were identi ed.
Chest X-ray ndings were classi ed as bronchopneumonia and lobar pneumonia with or without pleural effusion and atelectasis in the worst condition.

Statistical analysis
Continuous variables were presented as mean ± standard deviation and frequency (percentage) for categorical variables. Differences in baseline characteristics between groups were explored using independent two sample t-test and chi-square test (or Fisher's exact test) where appropriate.
To identify the independent risk factors for EPM, a multivariable logistic regression model was performed using a enter method that included variables with a probability value < 0.05 in the univariable analysis.
Odd ratios (OR) and their 95% con dence intervals (CI) were also calculated.
A two-tailed P-value of less than 0.05 was considered statistically signi cant. All analyses were conducted using SPSS Statistics version 26.0 (IBM Corp., Armonk, NY, USA).

Result
Characteristics of the study participants Page 7/22 The clinical characteristics of the pediatric patients of this study are presented in Table 1. The mean age of the 9,190 enrolled children with MP was 64.3 months. EPM were observed in older children (P < 0.001), however there was no difference in EPM occurrence between schoolchildren and preschool children (P = 0.141). The portion of male and female patients was similar (49.5% vs. 50.5%), although EPM occurred more frequently in male patients (51.9% vs 48.1%, P = 0.010). ). EPM were more common in children with MRMP than that with MSMP (P < 0.001). Children with respiratory di culty, such as cyanosis and dyspnea, were more prone to develop EPM (P < 0.001) and therefore oxygen therapy and ventilator care were more accompanied in children with EPM (P < 0.001). Children with asthma history showed less EPM (P < 0.001), however, atopic sensitization was not associated with EPM (P = 0.326).
In addition to the clinical characteristics, laboratory and radiological ndings were signi cantly different between children with and without EPM ( Table 2). Liver enzymes and LDH were higher in children with EPM (P < 0.001, respectively), but CRP levels and ESR were not statistically different (P = 0.111 and P = 0.958, respectively). EPM were developed more in children with lobar pneumonia and pleural effusion (P < 0.001, respectively). Values are presented as a mean ± standard deviation or number (%).

Risk factors associated with mucocutaneous manifestations
The univariable logistic regression model showed that the main parameters associated with mucocutaneous manifestation include schoolchildren, male patients, MRMP and MLMP, elevated ALT and LDH levels, respiratory virus co-infection, pleural effusion, and atopic sensitization (  To assess the relationships between more than one predictor and the outcome, the multivariable logistic regression was also evaluated. The elevated ALT levels and atopic sensitization were found to be positively associated with mucocutaneous manifestation (aOR 3.623, 95% CI   In multivariable analysis, elevated LDH levels and pleural effusion were signi cantly associated with elevation of liver enzymes (aOR 3.055, 95% CI 2.257-4.137; aOR 2.635, 95% CI 1.767-3.930, respectively).

Risk factors associated with proteinuria and other manifestations
The result of the univariable and multivariable analysis between proteinuria and various risk factors are shown in Table 5  In this study, age and gender of the patients were the factors that may be associated with the development of EPM in MP. Many reports have con rmed that EPM owing to MP frequently occur in children [17]. A report showed that the incidence of EPM in adult was only 2.2% [18]. However, even among children, the incidence of EPM may be different and schoolchildren are more prone to occur EPM than preschool children are. In addition, increasing MRMP is an emerging issue and a previous study in China has shown that EPM are risk factors for refractory Mycoplasma pneumonia and macrolide resistance [11,19]. EPM rates were signi cantly related to the extent of macrolide resistance, with the MRMP population is prone to have a higher risk of all complications. In our study, MRMP was also associated with the increased development of each EPM in univariable analysis. Higher levels of LDH have been reported to be related to MRMP [20] and could be another risk factor for EPM. Clinically and radiologically severe patients, such as those with pleural effusion, the requirement of oxygen supply or ventilator use, moderate to severe dyspnea and the need for ICU care, were at risk of developing EPM.
Therefore, severe and macrolide non-responsive patients should be evaluated and considered for concomitant EPM in MP.
According to the ndings of this study, each EPM is associated with different factors and requires further investigation to understand the mechanism that causes a particular EPM in MP. Mucocutaneous manifestations are more prominent in children with higher ALT levels and a history of atopic sensitization, but less frequently when respiratory viruses are co-infected. Abnormal liver function was associated with higher levels of LDH and pleural effusion, whereas proteinuria developed more in children with respiratory virus co-infection.
The strength of this study is that it was a multicenter study (n=23) with a large number of study participants (n=9,190). Previous studies were limited due to the involvement of less number of hospitals (one or two) and, hence could provide less patient data. Additionally, pediatric pulmonary specialists, participating in the Pneumonia and Respiratory Disease Study Group of Korean Academy of Pediatric Allergy and Respiratory Disease, have joined this study to evaluate and treat MP. However, this study was a retrospective study with chart review, and some missing data that may cause bias, suggesting the need for further prospective studies. Since the data collected in this study were from hospitalized in tertiary hospitals, more severe patients could have been analyzed.
In conclusion, we found that 23.9% of children with pediatric MP patients had EPM in Korea. Each EPM is associated with different factors. Mucocutaneous manifestations were increased with elevated ALT, and atopic sensitization and elevation of liver enzymes were associated with elevated LDH and pleural effusion. Respiratory virus co-infection is a risk factor for proteinuria. Further studies are needed to reveal the risk factors and pathophysiology of EPM of mycoplasma infection to prevent and effectively treat MP.