High Mycoplasma pneumoniae Loads and Persistent Long-term Mycoplasma pneumoniae Infection in Lower Airway Associated with Severity of Pediatric Mycoplasma pneumoniae Pneumonia

An increased number of refractory mycoplasma pneumoniae (MP) pneumonia (MPP) cases have been reported. However the duration of MP infection in lower airway and the course of anti-MP treatment remains unclear. We retrospectively reviewed the medical records of 94 MPP children. Patients were classified into two groups. The long-term group (Group LT) was defined as bronchoalveolar lavage fluid (BALF) remained MP-positive by PCR after 30 days of the disease course. The non-long-term group (Group NLT) was defined as BALF became MP-negative by PCR within 30 days of disease and patients who only needed one bronchoscopy lavage therapy. MP loads, clinical outcomes were analyzed along with other clinical measurements. airway. High MP bacterial loads and persistent long-term MP infection in lower airway were associated with severity of pediatric MPP.


Introduction
An increased number of refractory, severe, and even fatal Mycoplasma pneumoniae (MP) pneumonia (MPP) cases have been reported in recent years [1][2][3] . Although the patient's body temperature can return to normal after treatment, airway and pulmonary lesions are still severe in some refractory MPP (RMPP) patients. Furthermore, some RMPP patients have sequelae, mainly atelectasis, due to airway obliterans (AO) [4] .
Cell-mediated immunity and cytokine responses against the pathogens, MP resistance to macrolides and airway hypersecretion have been reported in MPP [5,6] . However some patients received timely treatment with sensitive antibiotics such as moxifloxacin, corticosteroid, and bronchoscopy lavage therapy (BLT), AO was still unavoidable. We observed airway hypersecretion and subsequent airway damage was severe in some children with RMPP, who required several treatments with BLT. Therefore, many factors of RMPP remain unclear, such as why airway damage persists, how long it takes to clear MP, and the course of anti-MP treatment. Elucidating these issues is important to help guide treatment strategies and further understand the underlying mechanisms of pediatric RMPP.
Fluorescence quantitative PCR (FQ-PCR) for MP gene detection was applied in all patients.
Our studies suggested that MP could persist in the lower airway for up to 4 months, which is consistent with the severity of airway and pulmonary damage. In addition, MP culture was performed in 15 bronchoalveolar lavage fluid (BALF) samples after 30 days of the disease course; 12 of 15 samples were positive, suggesting active MP infection. Moreover, 4 MP was detected by FQ-PCR in the peripheral blood of one patient with severe MPP, which suggested a rare case of MP-associated bloodstream infection. To  Patients were classified into two groups. The long-term group (Group LT) was defined as BALF remained MP-positive by FQ-PCR after 30 days of the disease course. The non-longterm group (Group NLT) was defined as BALF became MP-negative by PCR within 30 days of disease and patients who only needed one BLT assessed by two clinicians.
The clinical and demographic data were collected and recorded for each patient. Variables included age, gender, disease course before first admission to our department, and inflammatory markers [the maximal white blood cells (WBC), C reactive protein (CRP, normal range<8mg/L), lactate dehydrogenase (LDH)] within 15 days of disease course.
Additionally the MP loads, findings of bronchoscopy and chest imaging, and treatment with anti-MP antibiotics and corticosteroid were recorded.
Definitions RMPP was diagnosed according to the following criteria: (1) patients with MPP had persistent fever and deterioration of clinical and radiological findings after 7 days of disease course; (2) patients were treated with macrolides for 7 days or more e [6,7] {}.
The diagnosis of AO was based on the obliteration of the lumen of bronchi, bronchial branches, bronchial segments, or bronchial subsegments under bronchoscopy (Fig. 1A) [6] .
The imaging findings of AO were atelectasis.

Fiberoptic Bronchoscopy and BALF Collection
Fiberoptic bronchoscopy and BALF Collection were performed as previously reported [6] .

FQ PCR
FQ PCR was performed as previously reported [8] . The standard curves for each MP Quantitative was determined by analyzing serial 10-fold dilutions of stand red MP DNA (M129 strain, 1 copy/μl to 10 8 copies/μl). Each DNA of specimens and stand red was assayed in triplicate.

Study Population
A total of 94 MPP patients (age range: 3 years 4 months-13 years 7 months) were enrolled in this study. The cohort was 50% male (n=47). All patients in Group LT and 40 patients in 6 Group NLT were finally diagnosed with RMPP. There was no significant difference in gender and age between Groups LT and NLT (P>0.05) (Table1).

Clinical characteristics
The average duration of disease before hospitalization was 17.40±13.33 days and 9.27±3.79 days in Groups LT and NLT, respectively (P<0.001).
In Group LT, most patients resided from all over the country and 12 patients had type I respiratory failure, which suggested patients in this group were severe and difficult to treat. Mean WBC count, CRP level, and LDH concentration were 11.85±4.61×10 9 /L, 104.96±54.20 mg/L, and 681.00±298.24 IU/L, respectively.

A clear statistical difference was observed in inflammatory markers (WBC, CRP, LDH)
between the two groups (P<0.001) ( Table 1).

Bronchoscopy Findings
In Group LT, bronchoscopy carried out at the early stage of disease revealed mucus plug

Chest Imaging Findings
In Group LT, chest imaging at the early stage of disease revealed consolidation with high 7 density (2/3-1 pulmonary lobe in 4 patients; ≥1 lobe in 46 patients) in 50 patients and pleural effusion in 34 patients. Additionally, chest imaging revealed pulmonary embolism in 8 patients between 11 and 29 days of disease and necrotizing pneumonia (NP) in 28 patients between 14 and 60 days of disease (Fig. 3).
In Group NLT, chest imaging revealed consolidation with high density (<1/2 pulmonary lobe in 4 patients, 1/2-2/3 lobe in 24 patients, and 2/3-1 lobe in 11 patients) in 39 patients and bronchiolitis in 5 patients. Only two patients also had a small amount of pleural effusion.
A clear statistical difference was observed in lung damage between the two groups (P<0.05) ( Table 1).

MP Loads by PCR MP Loads in Paired Throat Swab and BALF Collected at the Same Time Point
In Group LT, throat swabs were performed in 20 patients 22-45 days after disease; 15 patients were MP-positive (including one patient with 10 5 copies/mL in 45 days) and 5 patients were MP-negative. Additionally, MP loads in throat swabs were lower than those in BALF. All BALF samples were MP-positive.
In Group NLT, throat swabs were performed in 10 RMPP patients 18-22 days after disease; 3 patients were MP-positive and 7 patients were MP-negative. Five BALF samples were MPpositive.
These results suggest that patients in Group LT had long-term persistent upper airway infection and MP was cleared from the upper airway before the lower airway (Table 2).

Treatment and Clinical Outcomes
All patients were treated with macrolides and BLT and followed up for at least 3 months.
Four non-RMPP patients in Group NLT were not treated by glucocorticoid.
In Group LT, moxifloxacin and doxycycline were also added to the treatment course in 14 patients and 2 patients, respectively. Azithromycin, which has anti-inflammatory effects, was administered for 3-6 months in 23 patients. Twenty-one patients whose CRP levels were >100 mg/L received high-dose methylprednisolone therapy (10-30 mg/kg/d for 3 days); after 3 months, chest imaging findings were normal in only 7 patients. These seven patients had started corticosteroid treatment 7-9 days into the disease course. In Group LT, chest imaging after 3 months of treatment revealed incomplete absorption of pulmonary lesions in 33 patients, including 13 AO patients.
In Group NLT, chest imaging after 3 months of treatment revealed incomplete absorption of pulmonary consolidation in seven patients, including two AO patients.
No obvious side effects were observed in any patients. Furthermore, there were no deaths in our study.
Taken together, these results indicate MP infection was difficult to clear, and absorption of lung lesions was very slow, especially in Group LT.

Discussion
Pediatric MPP is a significant public health problem. The clinical severity of MP infection is associated with MP loads in oropharyngeal secretions [9,10] . Nilsson et al. observed that the majority of patients with MP-positive throat swabs by PCR had persistent, sometimes long-term, upper airway infections (median time of carriage of MP DNA was 7 weeks after disease onset) [11] . However, to our knowledge, there is no report on the duration of MP infection in the lower airway of human. BALF is the most reliable specimen for diagnosing and assessing lower respiratory tract infection. In our study, all patients received BLT and some patients received BLT several times. We found MP was cleared in the upper airway earlier than the lower airway, and MP loads in BALF in the early stage of disease were associated with the subsequent duration of MP infection in the lower airway. Additionally, inflammatory markers (WBC, CRP, and LDH), airway and lung injury, and pulmonary complications were highly associated with MP loads and persistent MP infection in BALF.
Strikingly, in Group LT, all patients had persistent infection even for up to 120 days, and some patients still had high MP loads 50 days after disease onset. Given that PCR cannot distinguish live bacteria from dead bacteria, we also performed MP culture for 15 patients whose MP loads were >10 5 copies/mL; 12 of these patients were MP-positive, which suggested persistent long-term MP infection in the lower respiratory tract.
In our study, high MP loads and prolonged infection were mainly accompanied by airway damage, such as epithelial desquamation, necrosis, and AO, and pulmonary damage, such as consolidation and necrosis (sometimes including recurrent low or moderate fever in few patients). These findings suggested persistent MP activation and chronic respiratory infection rather than a state of chronic MP colonization. A previous study reported that adequate antibiotic treatment does not shorten the period of persistence [11] , which was confirmed in the present study.

It has been reported that MP clearance was difficult in animal experiments or in vitro.
Airway hypersecretion, epithelial desquamation, and remodeling cause AO, which compromises further clearance in vitro [12] , which was consistent with our present study.
Other studies have shown that MP is resistant to killing by neutrophils, suggesting that the bacterium might circumvent bactericidal activity of neutrophil extracellular traps [13] . B cells and MP-specific antibodies are also crucial for its clearance in the lungs of mice [14] .
Interleukin (IL)-1 plays an important role in recruiting and activating lung innate immune cells that are critical for MP clearance [15] . MP respiratory tract infection in IL-12 knockout mice results in improved bacterial clearance and reduced pulmonary inflammation such as neutrophilic alveolar infiltrate and airway obstruction [16] . Intranasal IL-12 therapy inhibits MP clearance and sustains airway obstruction in murine pneumonia [17] . In addition, host genetic background was also found to be important for MP disease severity [18,19] .
However the underlying mechanisms of persistence remain unclear.
Macrolide-resistant (MR)-MP strains are emerging and becoming increasingly common, especially in Eastern Asia, and have been reported at a rate of >90% in Beijing from 2008 to 2012 [20] . Macrolides are still used as first-line agents at high frequencies in countries with a high burden of MR-MP strains, such as China and Japan, mainly because of their low minimum inhibitory concentrations against the bacteria, anti-inflammatory effect and low toxicity in young children. Yang et al. reported that timely azithromycin treatment may be ineffective in treating MR-MP pneumonia in vivo, but we are unable to withhold antibiotic treatment to confirm this hypothesis due to ethical issues. However, the early use of azithromycin aids in reducing extrapulmonary complications in MPP [21] . Yoon et al. found macrolide treatment for macrolide-susceptible MPP did not contribute to significant clinical improvement compared to no antimicrobial treatment [22] . Kutty et al. reported that among MP PCR-positive children, length of stay was similar between children who did and did not receive an antibiotic with MP activity [23] .
In countries in which the prevalence of MR-MP strains is high, the replacement of macrolides as the first-line treatment with tetracyclines or fluoroquinolones was considered. To date, no tetracycline or fluoroquinolone resistance has been reported in MP clinical isolates. Both minocycline and tosufloxacin have shown good in vitro activities against MR-MP. However, the duration of fever following commencement of treatment with tosufloxacin in patients infected with MR-MP was not significantly different from the duration of fever following commencement of treatment with macrolides [24] . Compared to tosufloxacin, minocycline has a relatively high blood concentration and a long half-life [25] and efficiently penetrates into lung tissues and bronchial mucus [26] [27] . None of our patients received minocycline therapy. Although some of our patients received timely treatment with azithromycin, moxifloxacin, or doxycycline singly or in combination even for several courses, MP remained difficult to clear. This might suggest that antibiotics with potent intracellular activity are required for resolution of MP infection.
A previous study showed that 3-day methylprednisolone pulse therapy which was initiated 7-15 days after disease onset could be applied for RMPP treatment despite appropriate antibiotic therapy and appeared to be efficacious in 12 children [28] . However methylprednisolone pulse therapy was given 9 days after disease onset in our 14 patients in Group LT, chest imaging findings remained abnormal after 3 months of treatment.
Therefore, the underlying pathogenesis of RMPP is still unclear.
Compared with C57BL/6 mice, BALB/c mice exhibited host-dependent MP infection-related airway obstruction associated with chemokine [19] , which suggested some pediatric RMPP patients were highly reactive to MP. Community-acquired respiratory distress syndrome toxin induces exfoliation of mucosal cells and ciliostasis [29] . Airway epithelial remodeling during persistent MP infection may involve epithelial-to-mesenchymal transition in a 13 human airway epithelium model [12] . Therefore, high MP loads and epithelial desquamation, subsequent persistent MP infection, epithelial remodeling, and AO can lead to a vicious cycle of MP disease.
We found that MP may survive for a long time, especially in RMPP patients who suffer from MP associated NP, AO, or pulmonary embolism. Furthermore, although MP-associated bloodstream infection is rare, it can occur.
Our study has several limitations. The numbers of blood samples and throat swabs were small. The reason for the small number of blood samples is that the MP-negative rate is very high, in RMPP, thus clinicians do not often request blood samples for MP detection.
The number of children treated with moxifloxacin and doxycycline is small. Additionally, in some RMPP patients of Group LT, we did not store the last collected BALF sample.
Therefore, the actual duration of persistence is likely underestimated.
Taken together, our study confirms persistent long-term MP infection and difficult MP clearance in the lower airway of RMPP patients. Additionally, patients with MP loads ≥10 7 copies/mL in BALF are perhaps at high risk of RMPP and may be more likely to develop sequelae. Lastly, patients who have high levels of inflammatory markers, such as CRP and LDH, and high-density consolidation (especially more than one pulmonary lobe) on imaging may also have higher MP loads and subsequent persistent MP airway infection that is accompanied by NP, AO, or pulmonary embolism.  Tables   Table 1 Demographic and clinical features, inflammatory markers, and the findings of bronchoscopy and chest imaging in all patients  Fiberoptic bronchoscopy revealed mucous plug in RMPP patients.

Figure 3
A: Enhanced pulmonary CT revealed bilateral pulmonary embolism. B: Chest CT revealed low density lesions and multiple cavities within high density consolidation