Abstract
Objective
To investigate the clinical characteristics of macrolide resistance - associated mutations in hospitalized children with Mycoplasma pneumoniae pneumonia (MPP) and to evaluate their associations with severe Mycoplasma pneumoniae pneumonia (SMPP) and prolonged cough.
Methods
A total of 1,094 hospitalized children with MPP were retrospectively enrolled. Among them, 1,069 underwent A2063G/A2064G mutation testing, whereas 25 were not tested. Patients were classified into mutation-positive, mutation-negative, and untested groups according to mutation testing results. Clinical features, laboratory parameters, treatment profiles, and outcomes were compared among groups. Logistic regression and receiver operating characteristic (ROC) curve analyses were performed to identify risk factors for SMPP and prolonged cough.
Results
The proportion of SMPP was higher in the mutation-positive group than in the mutation-negative group. In addition, patients in the mutation-positive group exhibited a more pronounced inflammatory response and greater immune imbalance, as reflected by increased neutrophil percentage and D - dimer levels, along with decreased CD4 and CD8 counts. Univariate analysis showed that A2063G/A2064G mutation status was associated with both SMPP and prolonged cough. However, multivariable analysis identified fibrinogen and AST as independent factors associated with SMPP, while prolonged cough was additionally associated with the duration of azithromycin and clarithromycin therapy, as well as CD8 levels.
Conclusion
Children with macrolide resistance-associated mutations exhibited a higher inflammatory burden and a higher unadjusted proportion of SMPP. However, after adjustment, A2063G/A2064G mutation status was not independently associated with SMPP or prolonged cough. Clinical management should integrate inflammatory markers, oxygenation status, and clinical presentation for individualized monitoring and treatment.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s13052-026-02310-w.
Keywords: Mycoplasma pneumoniae pneumonia, Azithromycin, Macrolide resistance-associated mutations, Severe Mycoplasma pneumoniae pneumonia, Children
Introduction
Mycoplasma pneumoniae (MP) is a major cause of community - acquired pneumonia (CAP) in children, particularly in preschool - and school - aged populations, accounting for approximately 10% − 40% of pediatric CAP cases [1–4]. Although MPP is usually self - limiting, some children may develop SMPP [5, 6], characterized by worsening symptoms, prolonged fever, progressive radiographic lesions, or extrapulmonary complications [7–10]. Even with conventional macrolide therapy, these cases often require further evaluation and comprehensive management. When SMPP is accompanied by cardiovascular, digestive, or neurological involvement, elevated white blood cell counts, neutrophil proportions, and C - reactive protein (CRP) levels are often observed, indicating more severe disease and possible impairment of pulmonary function and quality of life [8, 11].
Macrolides are the first - line treatment for pediatric MPP and inhibit MP protein synthesis by targeting domain V of 23 S rRNA [12]. However, macrolide resistance has increased rapidly in recent years [13, 14]. The A2063G and A2064G mutations are the main molecular basis of resistance [15–17], with detection rates reported to exceed 90% in pediatric MPP cases in China [18, 19]. Although resistance may prolong disease course and increase the risk of complications, in vitro resistance does not necessarily indicate clinical failure of azithromycin, and some children may still benefit from treatment [18, 20–23].
In addition to antimicrobial resistance, disease severity in MP infection is closely associated with host immune responses. Elevated CRP, LDH, D - dimer, and neutrophil proportions have been associated with prolonged disease [2, 18]. Excessive immune activation may aggravate lung injury through the Toll - like receptor 2 (TLR2) - myeloid differentiation primary response 88 (MyD88) - nuclear factor κB (NF - κB) pathway and inflammatory cytokine release [1, 19]. Moreover, long non - coding RNA metastasis - associated lung adenocarcinoma transcript 1 (MALAT1), age, immune status, and co - infection may also affect disease severity and prognosis [2, 19].
The treatment of macrolide-resistant MP infection in children remains challenging, as alternative agents such as tetracyclines and fluoroquinolones are limited by age and safety concerns [24]. Clinically, early identification of high - risk children and comprehensive evaluation of inflammatory and immune status are essential to optimize azithromycin use and guide adjunctive immunomodulatory therapy [18, 25]. However, although macrolide resistance-associated mutations are common in pediatric MPP, it remains unclear whether these mutations independently increase the risks of SMPP and prolonged cough after adjustment for these factors.
The novelty of this study lies in moving beyond simple comparisons of clinical phenotypes between mutation-positive and mutation-negative patients by further evaluating whether macrolide resistance-associated mutations are independent risk factors for adverse clinical outcomes. This may help determine whether detection of these mutations alone is sufficient to justify discontinuing azithromycin and switching to alternative antibiotics, thereby providing evidence for clinical management.
Methods
Study design and population
This retrospective observational cohort study was conducted in the Department of Pediatrics at the First Affiliated Hospital of Guangzhou Medical University. Hospitalized children diagnosed with MPP between January 2023 and December 2024 were retrospectively enrolled and stratified according to A2063G / A2064G mutation status. The inclusion criteria were as follows: (i) hospitalized children and (ii) diagnosis of pediatric MPP [6]. The exclusion criteria were: (i) resolution of fever and complete disappearance of cough before admission; (ii) incomplete clinical data that could not meet the requirements for analysis; (iii) presence of hematologic malignancies; (iv) chronic granulomatous disease; (v) purulent meningitis; (vi) immunodeficiency disorders; and (vii) no use of azithromycin during treatment. The study was approved by the Institutional Review Board (GZMC 2009-08-1336) and adhered to the Declaration of Helsinki as previously described. The requirement for informed consent was waived due to the retrospective design. No human or animal experiments were conducted. All procedures were performed in accordance with the Declaration of Helsinki (1964) and its subsequent amendments or comparable ethical standards.
Data collection and definitions
Clinical data were retrospectively collected, including demographic characteristics, hospitalization information, underlying diseases and comorbidities, laboratory findings, chest imaging results, treatment regimens during hospitalization, duration of symptoms, diagnostic results, and clinical outcomes. The diagnostic criteria for MPP and SMPP were based on the Chinese guidelines for the diagnosis and treatment of pediatric MPP [6]. MPP was defined as: (i) positive MP - DNA or RNA detection; and / or (ii) a single serum MP antibody titer ≥ 1:160 (particle agglutination method), or a ≥ 4 - fold increase or a decrease to one - quarter in paired serum antibody titers during the disease course. SMPP was defined as confirmed MPP meeting any of the following criteria: (i) persistent high fever, defined as a body temperature ≥ 39 °C for ≥ 5 days, or fever lasting ≥ 7 days with no decrease or a further increase in peak body temperature; (ii) the presence of any respiratory symptom, including wheezing, shortness of breath, dyspnea, chest pain, or hemoptysis; (iii) extrapulmonary organ involvement that did not meet the criteria for critical illness; (iv) peripheral oxygen saturation ≤ 0.93 at rest while breathing room air at sea level; (v) radiographic findings meeting any of the following conditions: (1) high-density consolidation involving ≥ 2/3 of a single lung lobe, or high-density consolidation involving two or more lung lobes regardless of the extent of involvement; (2) diffuse unilateral or bilateral bronchiolitis, with or without bronchitis, including bronchial wall thickening, peribronchial exudation, and mucus plug formation leading to atelectasis; (vi) progressive worsening of clinical symptoms accompanied by radiographic progression of lesions by more than 50% within 24–48 h; or (vii) marked elevation of at least one inflammatory or tissue injury marker, including CRP, LDH, or D-dimer. As there is no universally accepted pathogen-specific cutoff for prolonged cough in pediatric Mycoplasma pneumoniae infection, prolonged cough was defined as a total cough duration exceeding 27 days, corresponding to the median value of this cohort. This cutoff is close to 4 weeks and falls within the clinically recognized range of prolonged acute cough in children [26–28]. A structured follow - up protocol was implemented for all patients. A follow - up visit was scheduled at the respiratory outpatient clinic one week after discharge, during which a comprehensive evaluation based on symptoms and signs was performed. Chest computed tomography or radiography was conducted when persistent cough, wheezing, or abnormal oxygenation was present. Routine imaging follow - up was not performed in asymptomatic patients.
Detection of macrolide resistance - associated mutations
Among the 1,094 included patients, 1,069 underwent A2063G/A2064G mutation testing using bronchoalveolar lavage fluid (BALF) samples, whereas 25 patients were not tested. In patients who underwent BALF examination, A2063G/A2064G mutations were detected using nested polymerase chain reaction combined with capillary electrophoresis and single-strand conformation polymorphism analysis (nPCR-CE-SSCP), according to previously reported methods [29].
Clinical outcome assessment
The primary outcomes of this study were the occurrence of SMPP and prolonged cough.
Statistical analysis
All statistical analyses were performed using Python 3.11 and R Studio (version 2026.01.1 + 403). Continuous variables were tested for normality, and most were non - normally distributed; therefore, they were expressed as median and interquartile range (IQR). Comparisons between two groups were performed using the Mann - Whitney U test, while comparisons among three or more groups were conducted using the Kruskal - Wallis H test. Categorical variables were presented as counts and percentages and analyzed using Pearson’s χ² test or Fisher’s exact test, as appropriate. A two - sided P value < 0.05 was considered statistically significant.
Univariate logistic regression analysis was performed for all candidate variables. Variables with P < 0.05 in univariate analysis were included in multivariate logistic regression models to estimate adjusted odds ratios (ORs) and 95% confidence intervals (95% CIs) [18, 20]. Multicollinearity was assessed prior to model construction, and variables with significant collinearity were selected based on clinical relevance and statistical performance. Receiver operating characteristic (ROC) curve analysis was further used to evaluate the predictive performance of relevant indicators for SMPP and prolonged cough [30, 31].
Results
Study population
During the study period from January 2023 to December 2024, a total of 1,196 pediatric cases initially met the inclusion criteria. Of these, 29 patients were excluded, including those with resolution of fever and complete disappearance of cough before admission (n = 1), incomplete clinical data (n = 21), hematologic malignancies (n = 2), chronic granulomatous disease (n = 3), purulent meningitis (n = 1), and immunodeficiency disorders (n = 1). Among the remaining 1,167 patients, 73 who did not receive azithromycin during treatment were further excluded. Finally, 1,094 eligible patients were included in the analysis (Fig. 1).
Fig. 1.
Study flowchart
All included patients received azithromycin therapy. Among them, 958 patients were positive for A2063G/A2064G mutations (mutation-positive group), 111 were negative for A2063G/A2064G mutations (mutation-negative group), and 25 were not tested for these mutations (untested group).
Clinical characteristics of 1,094 children with MPP
A total of 1,094 children with MPP were included in the analysis. The median age was 6.00 years, and 571 patients (52.2%) were male. Bacterial and viral co-infections were identified in 321 (29.3%) and 244 (22.3%) patients, respectively. The median total duration of fever and cough was 7.00 and 27.00 days, respectively. Most pulmonary lesions were completely resolved among patients who completed follow-up imaging. Detailed demographic, laboratory, treatment, clinical course, and imaging findings are summarized in Table 1 and Supplementary Table 1.
Table 1.
Clinical characteristics of 1,094 children with MPP
| Category / Variable | Overall Results |
|---|---|
| Demographic Data | |
| Age (years), median (IQR) | 6.00 (4.00–8.00) |
| Male, n (%) | 571 (52.2%) |
| Infection-Related Indicators | |
| Bacterial co-infection, n (%) | 321 (29.3%) |
| Viral co-infection, n (%) | 244 (22.3%) |
| Laboratory Findings | |
| PaO₂, median (IQR) | 85.55 (75.41–96.20) |
| White blood cell count (×10⁹/L), median (IQR) | 7.54 (6.05–9.62) |
| Neutrophils (%), median (IQR) | 60.10 (50.00-68.53) |
| Lymphocytes (%), median (IQR) | 29.30 (22.30-38.32) |
| CRP (mg/L), median (IQR) | 1.30 (0.54–2.38) |
| LDH(U/L), median (IQR) | 314.00 (272.00-363.55) |
| Fibrinogen (g/L), median (IQR) | 4.29 (3.72–4.80) |
| D-dimer, median (IQR) | 733.50 (487.00-1088.50) |
| CD3, median (IQR) | 1356.00 (997.00-1941.00) |
| CD4, median (IQR) | 734.00 (526.00-1089.00) |
| CD8, median (IQR) | 459.00 (320.00-665.00) |
| Treatment | |
| Methylprednisolone use, n (%) | 317 (29.0%) |
| Duration of azithromycin use (days), median (IQR) | 6.00 (4.00–6.00) |
| Clinical Course | |
| Duration of fever before treatment (days), median (IQR) | 6.00 (4.00–7.00) |
| Total duration of fever (days), median (IQR) | 7.00 (5.00–9.00) |
| Total duration of cough (days), median (IQR) | 27.00 (22.00–33.00) |
| Imaging Findings (Initial Imaging) | |
| Atelectasis, n (%) | 46/801 (5.74%) |
IQR, interquartile range; PaO₂, arterial partial pressure of oxygen; CRP, C-reactive protein; LDH, lactate dehydrogenase; CD, cluster of differentiation; BALF, bronchoalveolar lavage fluid. Percentages for initial imaging findings were calculated using the number of patients who underwent initial imaging examination as the denominator. Percentages for follow-up imaging outcomes were calculated using the number of patients who completed follow-up imaging as the denominator. Percentages for other categorical variables were calculated using the total number of patients in each group as the denominator
Comparison of clinical characteristics among three groups
Compared with the mutation-negative group, the mutation-positive group had a higher proportion of SMPP and showed greater inflammatory burden and immune imbalance, as reflected by elevated inflammatory markers and lower lymphocyte subset counts. The mutation-positive group also had a longer fever course, indicated by a lower proportion of patients with fever duration ≤ 3 days, and more frequent use of clarithromycin, methylprednisolone, and doxycycline. The untested group did not show an obviously poorer short-term clinical course than the tested groups, but this finding should be interpreted cautiously because of the small sample size. Detailed between-group comparisons, including laboratory findings, treatment profiles, clinical course, and imaging outcomes, are presented in Table 2 and Supplementary Table 2.
Table 2.
Comparison of clinical characteristics and treatment outcomes among children with MPP Stratified by A2063G/A2064G mutation status
| Category / Variable | Mutation-Positive Group | Mutation-Negative Group | Untested Group | P value |
|---|---|---|---|---|
| Demographic Data | ||||
| Age (years), median (IQR) | 6.00 (4.00–8.00) | 5.00 (3.00–8.00) | 6.00 (3.00–9.00) | 0.1737 |
| Male, n (%) | 496 (51.77%) | 62 (55.86%) | 13 (52.00%) | 0.7173 |
| Infection-Related Indicators | ||||
| Bacterial co-infection, n (%) | 258 (26.93%) | 55 (49.55%) | 8 (32.00%) | < 0.0001*** |
| Viral co-infection, n (%) | 191 (19.94%) | 49 (44.14%) | 4 (16.00%) | < 0.0001*** |
| Laboratory Findings | ||||
| PaO₂, median (IQR) | 85.30 (75.00-95.78) | 88.20 (78.70–99.10) | 90.70 (85.00-94.60) | 0.3870 |
| White blood cell count (×10⁹/L), median (IQR) | 7.47 (6.07–9.51) | 8.55 (6.13-11.00) | 7.46 (5.84–8.76) | 0.0294* |
| Neutrophils (%), median (IQR) | 60.90 (51.70–69.00) | 48.60 (40.10–63.10) | 52.60 (47.30–63.70) | < 0.0001*** |
| Lymphocytes (%), median (IQR) | 28.25 (21.80-36.82) | 37.10 (27.15–48.10) | 37.00 (24.50–39.40) | < 0.0001*** |
| CRP (mg/L), median (IQR) | 1.35 (0.60–2.38) | 0.85 (0.11–2.19) | 1.05 (0.70–2.66) | 0.0035** |
| LDH (U/L), median (IQR) | 315.00 (275.00-364.50) | 293.00 (249.40-349.75) | 324.50 (287.05–347.00) | 0.0499* |
| Fibrinogen (g/L), median (IQR) | 4.31 (3.80–4.81) | 3.74 (3.01–4.63) | 4.16 (3.11–4.80) | < 0.0001*** |
| D-dimer, median (IQR) | 763.00 (528.50-1107.50) | 464.00 (275.50–711.00) | 691.50 (467.50-1168.50) | < 0.0001*** |
| CD3, median (IQR) | 1310.50 (985.25-1869.50) | 2005.00 (1224.75-2577.75) | 1580.00 (1037.00-1934.50) | < 0.0001*** |
| CD4, median (IQR) | 707.00 (513.00-1010.25) | 1098.50 (658.75-1427.25) | 764.00 (615.00-1090.50) | < 0.0001*** |
| CD8, median (IQR) | 434.50 (314.75–647.00) | 608.00 (432.50-863.50) | 483.00 (326.00-642.50) | < 0.0001*** |
| Neutrophils in BALF (%), median (IQR) | 93.50 (88.00-96.25) | 88.50 (71.62–94.56) | 82.38 (65.06–89.31) | < 0.0001*** |
| Macrophages in BALF (%), median (IQR) | 3.12 (1.25-8.00) | 7.25 (2.25–23.75) | 12.25 (6.31–29.81) | < 0.0001*** |
| Treatment | ||||
| Clarithromycin use, n (%) | 520 (54.28%) | 34 (30.63%) | 11 (44.00%) | < 0.0001*** |
| Doxycycline use, n (%) | 238 (24.84%) | 14 (12.61%) | 5 (20.00%) | 0.0146* |
| Methylprednisolone use, n (%) | 295 (30.79%) | 18 (16.22%) | 4 (16.00%) | 0.0021** |
| Duration of methylprednisolone use (days), median (IQR) | 2.00 (2.00–3.00) | 3.00 (2.00-5.50) | 1.00 (1.00–2.00) | 0.0381* |
| Clinical Course | ||||
| Total duration of fever (days), median (IQR) | 7.00 (5.00–9.00) | 5.00 (0.50-8.00) | 7.00 (5.00–7.00) | < 0.0001*** |
| Total duration of cough (days), median (IQR) | 27.00 (22.00–33.00) | 27.00 (21.00-35.25) | 24.00 (19.50–26.00) | 0.0390* |
| SMPP, n(%) | 679(70.9%) | 64(57.7%) | 13 (52.0%) | 0.0030** |
| Fever duration ≤ 3 days, n (%) | 160 (16.70%) | 48 (43.20%) | 5 (20.00%) | < 0.0001*** |
| Imaging Findings (Initial Imaging) | ||||
| Atelectasis, n (%) | 32/694 (4.61%) | 14/85 (16.47%) | 0/22 (0.00%) | 0.0003*** |
IQR, interquartile range; PaO₂, arterial partial pressure of oxygen; CRP, C-reactive protein; LDH, lactate dehydrogenase; CD, cluster of differentiation; BALF, bronchoalveolar lavage fluid; SMPP, severe mycoplasma pneumoniae pneumonia. Percentages for initial imaging findings were calculated using the number of patients who underwent initial imaging examination as the denominator. Percentages for follow-up imaging outcomes were calculated using the number of patients who completed follow-up imaging as the denominator. Percentages for other categorical variables were calculated using the total number of patients in each group as the denominator. * P < 0.05, ** P < 0.01, *** P < 0.001
Risk factor analysis for SMPP
In the analysis of SMPP, several inflammatory, immune-related, oxygenation, and BALF indicators were significant in univariate analysis. In the multivariable model, fibrinogen and AST remained independent risk factors, whereas PaO₂ and BALF neutrophil percentage were independently associated with a lower risk of SMPP. ROC analysis showed that PaO₂ had the best discriminatory performance among the evaluated indicators. Detailed regression and ROC results are shown in Tables 3 and 4; Fig. 2.
Table 3.
Logistic regression analysis of risk factors for SMPP in children
| Variable | Univariate Logistic Regression | Variables | Multivariable Logistic Regression | ||||
|---|---|---|---|---|---|---|---|
| OR | 95%CI | P value | OR | 95%CI | P value | ||
| PaO₂ | 0.9681 | 0.9553–0.9801 | < 0.001*** | Fibrinogen | 2.4386 | 1.3655–4.6150 | 0.0038** |
| Lymphocyte | 0.9764 | 0.9667–0.9860 | < 0.001*** | AST | 1.0913 | 1.0299–1.1671 | 0.0063** |
| Neutrophil | 1.0222 | 1.0129–1.0316 | < 0.001*** | PaO₂ | 0.9618 | 0.9333–0.9889 | 0.0079** |
| Fibrinogen | 1.3547 | 1.1688–1.5729 | < 0.001*** | BALF Neutrophil Percentage | 0.9198 | 0.8301–0.9888 | 0.0409* |
| Methylprednisolone | 1.6899 | 1.2573–2.2907 | 0.0006*** | ||||
| BALF Macrophage | 0.9809 | 0.9680–0.9934 | 0.0031** | ||||
| BALF Neutrophil Percentage | 1.0164 | 1.0046–1.0287 | 0.0064** | ||||
| CD3 | 0.9998 | 0.9996-1.0000 | 0.0105* | ||||
| AST | 1.0140 | 1.0038–1.0262 | 0.0150* | ||||
| ALT | 1.0136 | 1.0021–1.0282 | 0.0427* | ||||
| Mutation-positive group | 2.2465 | 0.9988–5.0126 | 0.0465* | ||||
IQR, interquartile range; CI, confidence interval; OR, odds ratio; PaO₂, arterial partial pressure of oxygen; BALF, bronchoalveolar lavage fluid; CD, cluster of differentiation; AST, aspartate transaminase; ALT, alanine transaminase. * P < 0.05, ** P < 0.01, *** P < 0.001
Table 4.
ROC curve analysis of key indicators for predicting SMPP
| Risk Factor | Cutoff Value | Sensitivity | Specificity | AUC (95% CI) | P value |
|---|---|---|---|---|---|
| Fibrinogen | 4.14 | 0.6237 | 0.5104 | 0.5779 (0.541–0.6148) | < 0.0001*** |
| AST | 34.65 | 0.4960 | 0.6339 | 0.5739 (0.538–0.6099) | < 0.0001*** |
| PaO₂ | 85.80 | 0.5881 | 0.7167 | 0.6898 (0.6363–0.7433) | < 0.0001*** |
| BALF Neutrophil Percentage | 90.38 | 0.6627 | 0.4925 | 0.5865 (0.5285–0.6445) | 0.0034** |
CI, confidence interval; AUC, area under the receiver operating characteristic curve; PaO₂, arterial partial pressure of oxygen; BALF, bronchoalveolar lavage fluid. * P < 0.05, ** P < 0.01, *** P < 0.001
Fig. 2.
Forest plots of SMPP and prolonged cough. The forest plots show the odds ratios (ORs) and 95% confidence intervals (CIs) for (a) SMPP and (b) prolonged cough, derived from multivariable logistic regression analysis.* P < 0.05, ** P < 0.01, *** P < 0.001
Risk factor analysis for prolonged cough
For prolonged cough, univariate analysis identified multiple inflammatory, immune-related, treatment-related, and clinical variables as significant factors. In the multivariable model, the duration of azithromycin therapy, duration of clarithromycin therapy, CD8 level, and piperacillin use were associated with prolonged cough, whereas CD3 level, neutrophil proportion, and fibrinogen were associated with a lower risk. ROC analysis indicated limited predictive performance of individual indicators. Detailed results are shown in Tables 5 and 6, Supplementary Tables 3, and Fig. 2.
Table 5.
Logistic regression analysis of risk factors for prolonged cough in children with MPP
| Variables | Univariate Logistic Regression | Variables | Multivariable Logistic Regression | ||||
|---|---|---|---|---|---|---|---|
| OR | 95%CI | P value | OR | 95%CI | P value | ||
| Neutrophils | 0.9911 | 0.9827–0.9995 | 0.0382* | Duration of azithromycin therapy | 1.5582 | 1.1742–2.1226 | 0.0037** |
| Untested group | 4.3033 | 1.5188–15.4782 | 0.0116* | Duration of clarithromycin therapy | 1.6619 | 1.1379–2.4561 | 0.0095** |
| Mutation-positive group | 4.2756 | 1.6130-14.7397 | 0.0082** | CD8 | 1.0021 | 1.0006–1.0039 | 0.0129* |
| CD3 | 1.0003 | 1.0001–1.0004 | 0.0012** | CD3 | 0.9992 | 0.9985–0.9999 | 0.0300* |
| Duration of azithromycin therapy | 1.3619 | 1.1627–1.6027 | 0.0002*** | Use of piperacillin | 1.8025 | 1.0556–3.0983 | 0.0315* |
| Duration of azithromycin use (days) | 1.1163 | 1.0563–1.1811 | 0.0001*** | Neutrophils | 0.9762 | 0.9543–0.9979 | 0.0336* |
| Use of piperacillin | 1.7246 | 1.3117–2.2707 | < 0.0001*** | Fibrinogen (g/L) | 0.7255 | 0.5342–0.978 | 0.0369* |
| CD8 | 1.0010 | 1.0005–1.0014 | < 0.0001*** | ||||
| Fibrinogen (g/L) | 0.7444 | 0.6469–0.8549 | < 0.0001*** | ||||
CI, confidence interval; OR, odds ratio; CRP, C-reactive protein; CD, cluster of differentiation. * P < 0.05, ** P < 0.01, *** P < 0.001
Table 6.
ROC curve analysis of key indicators for predicting prolonged cough
| Risk Factor | Cutoff Value | Sensitivity | Specificity | AUC (95% CI) | P value |
|---|---|---|---|---|---|
| Duration of azithromycin therapy | 2.5000 | 0.1313 | 0.8982 | 0.5097 (0.4744–0.5449) | 0.5974 |
| Duration of clarithromycin therapy | 1.5000 | 0.4004 | 0.7083 | 0.5555 (0.5119–0.5992) | 0.0176* |
| CD8 | 442.5000 | 0.5262 | 0.6465 | 0.5941 (0.5499–0.6382) | < 0.0001*** |
| CD3 | 1320.5000 | 0.5212 | 0.6414 | 0.5875 (0.5421–0.6329) | 0.0002*** |
| Piperacillin use | 0.5000 | 0.2619 | 0.7690 | 0.5154 (0.4864–0.5445) | 0.3078 |
| Neutrophils | 54.6500 | 0.6863 | 0.4727 | 0.5973 (0.5583–0.6363) | < 0.0001*** |
| Fibrinogen (g/L) | 3.8250 | 0.7528 | 0.4234 | 0.6009 (0.5615–0.6404) | < 0.0001*** |
CI, confidence interval; AUC, area under the receiver operating characteristic curve; CRP, C-reactive protein; CD, cluster of differentiation. * P < 0.05, ** P < 0.01, *** P < 0.001
Baseline characteristics and clinical outcomes of mutation-positive children according to doxycycline use
Among mutation-positive patients, those who received doxycycline after macrolide therapy had more severe baseline disease than those treated with macrolides alone, including older age, longer fever duration, higher inflammatory burden, and lower CD4 and CD8 levels. Because baseline characteristics were not balanced between groups, this analysis should be interpreted as a descriptive subgroup comparison rather than evidence of doxycycline inefficacy. Multivariable analysis showed that doxycycline use was not independently associated with SMPP, and no significant difference in the incidence of prolonged cough was observed between the two treatment groups. Detailed results are summarized in Table 7 and Supplementary Tables 4 and 5.
Table 7.
Baseline characteristics and clinical outcomes of mutation-positive children according to doxycycline use
| Category / Variable | Macrolide Only (n = 720) | Macrolide Followed by Doxycycline (n = 238) | P value |
|---|---|---|---|
| Baseline Characteristics | |||
| Age | 6 (3–7) | 8 (5.25-9) | < 0.0001*** |
| Duration of fever | 7 (5–9) | 8 (6–9) | 0.0018** |
| Disease Severity | |||
| PaO₂ | 85.2 (75-95.1) | 85.4 (75-98.5) | 0.8340 |
| Oxygen saturation | 93 (90.5–95) | 90 (89–92) | 0.0841 |
| Laboratory Findings | |||
| Neutrophils | 59.55 (50.48–68.03) | 64.65 (56.6–72) | < 0.0001*** |
| CRP (mg/L) | 1.26 (0.49–2.3) | 1.59 (0.82–2.52) | 0.0010** |
| D-dimer | 726 (491-1080.5) | 852 (626.5-1157.25) | 0.0011** |
| CD4 | 748 (552-1078.75) | 605.5 (460.5-871.5) | < 0.0001*** |
| CD8 | 460 (325–676) | 396 (289.25-566.25) | 0.0004*** |
| Infection-Related Indicators | |||
| Bacterial infection | 200 (27.8%) | 58 (24.4%) | 0.3134 |
| Viral infection | 150 (20.8%) | 41 (17.2%) | 0.2614 |
| Treatment | |||
| Duration of azithromycin therapy | 2 (1–2) | 1 (1–1) | < 0.0001*** |
| Use of methylprednisolone | 193 (26.8%) | 102 (42.9%) | < 0.0001*** |
IQR, interquartile range; PaO₂, arterial partial pressure of oxygen; CRP, C-reactive protein; CD, cluster of differentiation. * P < 0.05, ** P < 0.01, *** P < 0.001
Discussion
This retrospective cohort study of hospitalized children with MPP demonstrated that patients harboring A2063G/A2064G mutations exhibited a higher inflammatory burden and a greater unadjusted proportion of SMPP. However, after adjustment for inflammatory, immune, oxygenation, and treatment-related variables, mutation status was not independently associated with SMPP or prolonged cough. The main contribution of this study is that it shows that macrolide resistance-associated mutations should not be considered the sole determinant of adverse clinical outcomes, and that clinical interpretation of mutation status should be integrated with inflammatory burden, immune status, oxygenation, and treatment-related factors.
The mutation-positive group showed more pronounced abnormalities in inflammatory and immune parameters, consistent with previous reports [32] indicating that macrolide-resistant MPP is often accompanied by stronger inflammatory responses. However, mutation status was not retained as an independent risk factor in multivariable models, suggesting that the apparent association with adverse outcomes may be partially explained by enhanced host inflammatory responses and immune dysregulation rather than by the resistance mutation per se [6, 33–35]. Previous studies have also shown that a subset of children harboring A2063G/A2064G mutations can still achieve short-term clinical responses to azithromycin: one study reported that 36 of 155 patients (23.23%) defervesced within 3 days of treatment [18], while another study reported a mean time to defervescence of 4.6 days [4]. Consistently, in our cohort, 160 of 958 (16.70%) mutation-positive patients achieved defervescence within 3 days, indicating that the presence of resistance-associated mutations does not necessarily predict poor short-term clinical response to azithromycin.
Compared with mutation status, inflammatory markers, indicators of tissue or organ injury, oxygenation status, and immune parameters showed stronger associations with SMPP and prolonged cough. Fibrinogen and AST were independent risk factors for SMPP, whereas PaO₂ was a protective factor. For prolonged cough, treatment duration and immune indicators were retained in the multivariable regression model. These findings suggest that disease progression in pediatric MPP is driven by multiple pathophysiological pathways and cannot be explained solely by A2063G/A2064G mutation status [18, 36].
Subgroup analysis of resistance-associated mutations further supports an individualized management approach. Children who received doxycycline after macrolide therapy had more severe baseline disease, but doxycycline use was not independently associated with SMPP, and the incidence of prolonged cough did not differ significantly between groups. This does not indicate inefficacy of doxycycline but highlights that clinical decision-making should be guided by overall disease severity and individual patient risk. Mechanistically, our findings are biologically plausible. MP infection can activate the TLR2-MyD88-NF-κB innate immune pathway [1], and MP - derived community - acquired respiratory distress syndrome (CARDS) toxin may further trigger inflammasome activation and cytokine release [37], with excessive host inflammatory responses contributing to lung injury and prolonged symptoms [38]. These mechanisms also support the observation that inflammatory and immune status may be more closely associated with disease progression than the presence of resistance mutations.
Based on these findings, the study has important clinical implications: detection of A2063G/A2064G mutations alone should not be used as the sole criterion for discontinuing azithromycin [18, 20–22]. The broadly comparable short-term clinical course in the small untested group also suggests that routine mutation testing may not be required for every patient, particularly in selected low-risk cases. Treatment of macrolide-resistant MP infection in children remains challenging because alternative agents, such as tetracyclines and fluoroquinolones, are limited by age and safety considerations [24]. Clinical decision-making should therefore integrate overall disease severity, treatment response, patient age, drug safety, and inflammatory burden, with dynamic assessment of the potential need for adjunctive corticosteroids or intravenous immunoglobulin in children with more severe disease or at higher risk of progression [3, 39, 40].
This study has several limitations. First, it is a retrospective observational study in which treatment regimens were determined by attending clinicians according to clinical judgment rather than by randomization, which may introduce confounding by indication; consequently, the analysis of doxycycline use should be interpreted as an association rather than as evidence of treatment efficacy. Second, although multiple inflammatory, immune, and clinical variables were adjusted for, residual confounding cannot be fully excluded. Third, the study was conducted at a single center and included only hospitalized children, limiting the generalizability of the findings to milder outpatient cases. Fourth, not all patients had both baseline and follow-up imaging available, which may have introduced baseline assessment bias and follow-up bias. Fifth, 25 patients did not undergo mutation testing, and although they were included as an untested group, the small sample size limits the interpretability of comparisons; future studies with larger untested cohorts could further validate these findings. Sixth, because there is currently no universally accepted clinical threshold for defining prolonged cough in pediatric MP infection, the 27-day cutoff used in this study was data-driven and based on the median total cough duration in this cohort. If a standardized guideline-based cutoff becomes available in the future, the outcomes could be reanalyzed using that definition. Finally, prospective multicenter cohort studies with standardized resistance testing, predefined cough outcome thresholds, and more rigorous methods to control confounding, such as propensity score matching or randomized controlled trials, are warranted to validate these findings and establish optimal management strategies for children with resistance-associated MPP.
Conclusion
This study showed that A2063G/A2064G mutation status was not an independent risk factor for SMPP. In contrast, indicators of inflammation, immune status, oxygenation, and liver function were more strongly associated with SMPP and prolonged cough. These findings suggest that mutation testing alone should not guide treatment decisions, and that clinical management should be based on a comprehensive assessment of inflammatory status, oxygenation, and disease progression.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Not applicable.
Abbreviations
- MPP
Mycoplasma pneumoniae pneumonia
- SMPP
Severe Mycoplasma pneumoniae pneumonia
- ROC
Receiver operating characteristic
- CRP
C-reactive protein
- CAP
Community-acquired pneumonia
- TLR2
Toll-like receptor 2
- MyD88
Myeloid differentiation primary response 88
- NF-κB
Nuclear factor κB
- MALAT1
Metastasis-associated lung adenocarcinoma transcript 1
- IRB
Institutional Review Board
- BALF
Bronchoalveolar lavage fluid
- nPCR-CE-SSCP
nested polymerase chain reaction combined with capillary electrophoresis and single-strand conformation polymorphism analysis
- IQR
Interquartile range
- OR
Odds ratio
- CI
Confidence interval
- PaO₂
Arterial partial pressure of oxygen
- MUMPP
Macrolide-unresponsive Mycoplasma pneumoniae pneumonia
- CARDS
Community-acquired respiratory distress syndrome
- NLRP3
NLR family pyrin domain containing 3
- TNF-α
Tumor necrosis factor-α
- IL
Interleukin
Author contributions
Jili Yuan, Zili Zhang, and Chengyu Lu were responsible for study conception and experimental design. Zili Zhang, Yilan Huang, and Jili Yuan drafted the initial manuscript. Zili Zhang, Jili Yuan, Yilan Huang, Ming Liang, and Jiaxing Xu revised the manuscript. Yilan Huang and Jili Yuan completed data collection. Yilan Huang, Ming Liang, and Jiaxing Xu performed the data analysis. All authors have read and approved the final manuscript.
Funding
This work was supported by grants from the National Natural Science Foundation of China (82330002 and 82370046), Natural Science Foundation of Guangdong Province (2024A1515030257), State Key Laboratory of Respiratory Disease (SKLRD-Z-202406), the Educational Teaching Reform Project of Guangzhou Medical University (2025JXGG11), and the Curriculum Teaching and Research Office Project of Guangzhou Medical University (2025ZLGC03).
Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
This study was approved by the Institutional Review Board (GZMC 2009-08-1336, approved on August 13, 2009). The Ethics Committee waived the requirement for informed consent due to the retrospective nature of the study.
Consent for publication
Not applicable.
Competing interests
Yilan Huang and Zili Zhang are co-first authors for this study. The authors declare that they have no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yilan Huang and Zili Zhang contributed equally to this work.
Contributor Information
Chengyu Lu, Email: fishlu1972@qq.com.
Jili Yuan, Email: yuan_jili@163.com.
References
- 1.Chen M, Deng H, Zhao Y, et al. Toll-Like Receptor 2 Modulates Pulmonary Inflammation and TNF-α Release Mediated by Mycoplasma pneumoniae [J]. Front Cell Infect Microbiol. 2022;12:824027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zhao Q, Zhang T, Zhu B, et al. Increasing Age Affected Polymorphonuclear Neutrophils in Prognosis of Mycoplasma pneumoniae Pneumonia [J]. J Inflamm Res. 2021;14:3933–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wang YS, Zhou Y L, Bai G N, et al. Expert consensus on the diagnosis and treatment of macrolide-resistant Mycoplasma pneumoniae pneumonia in children [J]. World J Pediatr. 2024;20(9):901–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.CHOSIDOW A, MAAKAROUN-VERMESSE Z OKV, et al. Post-COVID-19 resurgence of Mycoplasma pneumoniae infections in French children (ORIGAMI): a retrospective and prospective multicentre cohort study [J]. Lancet Infect Dis. 2026;26(4):374–83. [DOI] [PubMed] [Google Scholar]
- 5.Dungu KHS, Holm M, Hartling U, et al. Mycoplasma pneumoniae incidence, phenotype, and severity in children and adolescents in Denmark before, during, and after the COVID-19 pandemic: a nationwide multicentre population-based cohort study [J]. Lancet Reg Health Eur. 2024;47:101103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zhang H, Yang J, Zhao W, et al. Clinical features and risk factors of plastic bronchitis caused by refractory Mycoplasma pneumoniae pneumonia in children: a practical nomogram prediction model [J]. Eur J Pediatr. 2023;182(3):1239–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Foy HM. Infections caused by Mycoplasma pneumoniae and possible carrier state in different populations of patients [J]. Clin Infect Dis. 1993;17(Suppl 1):S37–46. [DOI] [PubMed] [Google Scholar]
- 8.Zhao X, Lv J, Wu M, et al. Clinical characteristics and risk factors for Mycoplasma pneumoniae pneumonia in children [J]. Front Pediatr. 2024;12:1438631. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jeon HE, Kang HM, Yang EA, et al. Early confirmation of mycoplasma pneumoniae infection by two short-term serologic IgM examination [J]. Diagnostics (Basel). 2021;11(2). [DOI] [PMC free article] [PubMed]
- 10.Zhang J, Ma HK, Li BW, et al. Changes in urinary renal injury markers in children with Mycoplasma pneumoniae pneumonia and a prediction model for related early renal injury [J]. Ital J Pediatr. 2024;50(1):155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Yang S, Lu S, Guo Y, et al. A comparative study of general and severe mycoplasma pneumoniae pneumonia in children [J]. BMC Infect Dis. 2024;24(1):449. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Viz-Lasheras S, Gómez-Carballa A, Bello X, et al. A diagnostic host-specific transcriptome response for Mycoplasma pneumoniae pneumonia to guide pediatric patient treatment [J]. Nat Commun. 2025;16(1):673. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.YAN C, TONG S, WU Y, et al. Macrolide-resistant Mycoplasma pneumoniae resurgence in Chinese children in 2023: a longitudinal, cross-sectional, genomic epidemiology study [J]. Lancet Microbe. 2025;6(12):101200. [DOI] [PubMed] [Google Scholar]
- 14.Global spatiotemporal dynamics. of Mycoplasma pneumoniae re-emergence after COVID-19 pandemic restrictions: an epidemiological and transmission modelling study [J]. Lancet Microbe. 2025;6(4):101019. [DOI] [PubMed] [Google Scholar]
- 15.YAMAZAKI T. Epidemiology of Mycoplasma pneumoniae Infections in Japan and Therapeutic Strategies for Macrolide-Resistant M. pneumoniae [J]. Front Microbiol. 2016;7:693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kim K, Jung S, Kim M, et al. Global Trends in the Proportion of Macrolide-Resistant Mycoplasma pneumoniae Infections: A Systematic Review and Meta-analysis [J]. JAMA Netw Open. 2022;5(7):e2220949. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wang X, Li M, Luo M, et al. Mycoplasma pneumoniae triggers pneumonia epidemic in autumn and winter in Beijing: a multicentre, population-based epidemiological study between 2015 and 2020 [J]. Emerg Microbes Infect. 2022;11(1):1508–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cheng J, Liu Y, Zhang G, et al. Azithromycin Effectiveness in Children with Mutated Mycoplasma Pneumoniae Pneumonia [J]. Infect Drug Resist. 2024;17:2933–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Alishlash AS, Atkinson TP, Schlappi C, et al. Mycoplasma pneumoniae carriage with de novo macrolide-resistance and breakthrough pneumonia [J]. Pediatrics. 2019;144(4). [DOI] [PubMed]
- 20.Cheng J, Liu Y, Zhang G, et al. Atelectasis predicts poor prognosis in pediatric macrolides-unresponsive Mycoplasma pneumoniae pneumonia with A2063/2064G mutations treated with azithromycin [J]. Front Cell Infect Microbiol. 2025;15:1604102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tong L, Huang S, Zheng C, et al. Refractory Mycoplasma pneumoniae pneumonia in children: early recognition and management [J]. J Clin Med. 2022;11(10). [DOI] [PMC free article] [PubMed]
- 22.Cheong KN, Chiu SS, Chan BW, et al. Severe macrolide-resistant Mycoplasma pneumoniae pneumonia associated with macrolide failure [J]. J Microbiol Immunol Infect. 2016;49(1):127–30. [DOI] [PubMed] [Google Scholar]
- 23.Zuckerman JM. Macrolides and ketolides: azithromycin, clarithromycin, telithromycin. Infect Dis Clin North Am. 2004;18(3):621-649, xi. [DOI] [PubMed]
- 24.Kim Y, Park GW, Kim S, et al. Fluoroquinolone and no risk of Achilles-tendinopathy in childhood pneumonia under eight years of age-a nationwide retrospective cohort [J]. J Thorac Dis. 2021;13(6):3399–408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bi Y, Zhu Y, Ma X, et al. Development of a scale for early prediction of refractory Mycoplasma pneumoniae pneumonia in hospitalized children [J]. Sci Rep. 2021;11(1):6595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Leconte S, Paulus D. Prolonged cough in children: a summary of the Belgian primary care clinical guideline [J]. Prim Care Respir J. 2008;17(4):206–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Shields MD. The difficult coughing child: prolonged acute cough in children [J]. Cough. 2013;9(1):11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Morice AH, Millqvist E, Bieksiene K, et al. ERS guidelines on the diagnosis and treatment of chronic cough in adults and children [J]. Eur Respir J. 2020;55(1). [DOI] [PMC free article] [PubMed]
- 29.Lin C, Li S, Sun H, et al. Nested PCR-linked capillary electrophoresis and single-strand conformation polymorphisms for detection of macrolide-resistant Mycoplasma pneumoniae in Beijing, China [J]. J Clin Microbiol. 2010;48(12):4567–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Faraggi D. Estimation of the area under the ROC curve [J]. Stat Med. 2002;21(20):3093–106. [DOI] [PubMed] [Google Scholar]
- 31.Cook NR. Use and misuse of the receiver operating characteristic curve in risk prediction [J]. Circulation. 2007;115(7):928–35. [DOI] [PubMed] [Google Scholar]
- 32.Zhan XW, Deng LP, Wang ZY, et al. Correlation between Mycoplasma pneumoniae drug resistance and clinical characteristics in bronchoalveolar lavage fluid of children with refractory Mycoplasma pneumoniae pneumonia [J]. Ital J Pediatr. 2022;48(1):190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Guo ZQ, Gu SY, Tian ZH, et al. A comprehensive review of Mycoplasma pneumoniae infection in chronic lung diseases: recent advances in understanding asthma, COPD, and bronchiectasis [J]. Front Med (Lausanne). 2024;11:1437731. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Zimmermann P, Ziesenitz VC, Curtis N, et al. The Immunomodulatory Effects of Macrolides-A Systematic Review of the Underlying Mechanisms [J]. Front Immunol. 2018;9:302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Chen L, Li H, Zhong P, et al. Clinical features of MPP with or without viruses among hospitalized children in 2023, Wenzhou, Zhejiang, China [J]. Ital J Pediatr. 2025;51(1):286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Li YT, Zhang J, Wang MZ, et al. Changes in coagulation markers in children with Mycoplasma pneumoniae pneumonia and their predictive value for Mycoplasma severity [J]. Ital J Pediatr. 2023;49(1):143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Bose S, Segovia JA, Somarajan SR, et al. ADP-ribosylation of NLRP3 by Mycoplasma pneumoniae CARDS toxin regulates inflammasome activity [J]. mBio. 2014;5(6). [DOI] [PMC free article] [PubMed]
- 38.Smit JM, VAN der Zee PA, Stoof SCM, et al. Predicting benefit from adjuvant therapy with corticosteroids in community-acquired pneumonia: a data-driven analysis of randomised trials [J]. Lancet Respir Med. 2025;13(3):221–33. [DOI] [PubMed] [Google Scholar]
- 39.You S Y, Jwa HJ, Yang EA, et al. Effects of Methylprednisolone Pulse Therapy on Refractory Mycoplasma pneumoniae Pneumonia in Children [J]. Allergy Asthma Immunol Res. 2014;6(1):22–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Liu S, Shan L, Kang XYX, et al. Effects of methylprednisolone or immunoglobulin when added to standard treatment with intravenous azithromycin for refractory Mycoplasma pneumoniae pneumonia in children [J]. World J Pediatr. 2017;13(4):321–7. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.