We enrolled 203 children who were hospitalized with MP at the Department of Pediatrics, Kangbuk Samsung Medical Center from June to November 2011. In all cases, MP was diagnosed on the basis of radiologic findings and confirmed by serologic testing. Pneumonia was diagnosed as the presence of consolidation on a chest x-ray determined by two pediatric radiologists at the time of admission. We classified pneumonia as lobar or bronchopneumonia, with or without pleural effusion. Infection with M. pneumoniae was confirmed by a 4-fold increase in the serologic titer of antimycoplasma antibody (AMA) during the convalescent phase compared to AMA titer drawn during the acute phase, or by an initial AMA titer greater than 1:1280. The AMA titer was measured using an indirect microparticle agglutination method (Serodia-Myco II, Fujirebio, Tokyo, Japan). Other viral studies were not performed during this period.

We analyzed the medical records for patient characteristics including age, sex, duration of hospitalization and fever, family history of allergic diseases, and history of doctor's diagnosis of atopic dermatitis, allergic rhinitis, or asthma in his/her lifetime. The severity of respiratory illness was evaluated based on the extent of pulmonary consolidation, pleural effusion, and supplemental oxygen requirements as defined by oxygen saturation below 90% in room air, which was measured by pulse oxymetry.

All children who were diagnosed with MP were treated with macrolide antibiotics as well as broad spectrum antibiotics, because coinfection of other bacteria could not be ruled out. Refractory MP was defined as a case with persistent fever over 38.3℃ and a progressive pulmonary consolidation or pleural effusion despite appropriate macrolide antibiotics for 5 days or longer after admission. For the refractory cases, oral prednisolone (1 mg/kg/day) or intravenous methylprednisolone (1 mg/kg/day) was administered in addition to a macrolide antibiotic on the fifth day after admission.

We compared clinical features, laboratory tests, history of allergic diseases, and atopic sensitization rate between children with MP responding to macrolide and children with refractory MP.

This study was approved by The Institutional Review Board of the Kangbuk Samsung Medical Center.

Blood samples were analyzed for white blood cell count, lymphocyte, hemoglobin, platelet, lactate dehydrogenase (LDH), C-reactive protein, erythrocyte sedimentation rate, total eosinophil counts, titer of AMA and cold agglutinin, total immunoglobulin E (IgE) and allergen-specific IgE antibody. Serum total IgE and specific IgE levels to eight common aeroallergens (Dermatophagoides farinae, Dermatophagoides pteronyssinus, cat, dog, Al-ternaria, cockroach, tree pollen mixture, and weed pollen mixture) were assayed using ImmunoCAP (Pharmacia Diagnostics, Uppsala, Sweden). Atopic sensitization was defined as ≥0.35 IU/mL of at least one specific IgE level10,¹¹⁾.

Statistical analysis was performed using STATA release 11.1 (StataCorp LP, College Station, Texas, USA). Differences between two groups were analyzed using Student t test or chi-square test according to the characteristics of variables. Quantitative variables are expressed as means±standard deviation or median values with interquartile range (IQR). The analysis was conducted by transforming the data to a logarithmic scale when data had skewed distributions. Logistic regression was used to determine risk factors associated with refractory MP. Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were derived after adjusting for age and sex. A value of P<0.05 was considered statistically significant.

The clinical and laboratory characteristics of children with MP are shown in Table 1. A total of 203 children were diagnosed with MP. The median age of patients was 5 years (IQR, 3.0-7.0). Twenty-six children (12.8%) were treated with steroids. Mean duration of steroid therapy was 5.4 days and all the patients with refractory MP became afebrile within 24 hours after steroid therapy was started except for 1 case. The female patient with severe refractory MP showed massive left lower lobar consolidation with pleural effusion (Fig. 1A). On the 2nd admission day, she became dyspneic with oxygen saturation of 88%, so we started oxygen supplement and inserted chest tube. Fever and productive cough continued for the 5 days after admission. On the fifth day, we initiated methylprednisolone therapy (1 mg/kg/day). After two days after the initiation of steroid therapy, body temperature started to be reduced below 38℃ and dyspnea was resolved. Chest radiograph demonstrated decreased consolidation and pleural effusion on the left lower lobe (Fig. 1B). She was discharged on the 26th day of admission without sequelae (Fig. 1C). There were no adverse events observed with steroid therapy. Fourteen children were administered oral prednisolone and intravenous methylprednisolone was given to twelve of them. Seven children (3.4%) were hypoxic at initial presentation and required supplemental oxygen. However, none of the patients were required transfer to an intensive care unit or needed mechanical ventilation. Sixty-eight children (33.5%) had lobar pneumonia, and 135 children (66.5%) showed bronchopneumonia. Pleural effusion was present in 21 children (10.3%) and five of them required chest tube insertion and drainage. Patients demonstrated mild elevated C-reactive protein level and erythrocyte sedimentation rate. The prevalence of previous history of asthma or allergic rhinitis was 7.8% and 36.4%, respectively. The overall atopic sensitization rate was 21.6%.

We compared the clinical and laboratory findings between children with refractory MP and those with macrolide-responsive MP (Table 2). Children with refractory MP showed higher serum LDH levels and longer duration of hospitalization compared to children who with macrolide-responsive infections. The 26 children who received steroid therapy were hospitalized for a mean duration of 8.5 days, with the longest being 26 days, which would represent more severe pneumonia in children with refractory MP. Blood platelet count was lower in children with refractory MP than in children without. However, the mean age, sex, white blood cell count, lymphocyte count, hemoglobin level, C-reactive protein, erythrocyte sedimentation rate, total IgE concentration, and blood total eosinophil count did not differ significantly between the two groups.

We compared the respiratory severity, type of pneumonia, atopic sensitization, and previous history of allergic diseases between children with refractory MP and those with macrolide-responsive MP. The respiratory severity was investigated on the basis of oxygen requirement, pleural effusion, the number of involved lobes on chest x-ray, and chest tube insertion (Table 3). Children with refractory MP had a higher prevalence of supplemental oxygen therapy and pleural effusion (11.5% vs. 2.3%, P=0.015; 34.6% vs. 6.8%, P<0.001, respectively). There were no significant differences in type of pneumonia between children with and without refractory MP. However, in children with lobar pneumonia, those with more than 2 lobes involved were more likely to require steroid therapy compared to children with 1 lobe involved (P=0.002). The prevalence of atopic sensitization and history of asthma was higher in children with refractory MP than in children without (42.3% vs. 18.6%, P=0.006; 23.1% vs. 5.7%, P=0.002, respectively). After controlling for age and sex, children with atopic sensitization and history of asthma also showed a higher association with refractory MP requiring steroid therapy (aOR, 3.83; 95% CI, 1.55-9.47; and aOR, 5.40; 95% CI, 1.67-17.45, respectively). However, there were no associations discovered between refractory MP and allergic rhinitis or atopic dermatitis.