Exploratory Analysis of Factors Associated with Poor Prognosis in Children with Severe Mycoplasma pneumoniae Pneumonia

Summary

This retrospective study identified immune dysfunction and extrapulmonary complications as independent predictors of poor prognosis in children with severe Mycoplasma pneumoniae pneumonia. Serum IgA, IgG, IgM, and C-reactive protein (CRP) showed exploratory prognostic value, although prospective studies are needed to validate their clinical applicability.

Abstract

Severe Mycoplasma pneumoniae pneumonia (MPP) in children is associated with immune dysregulation and unfavorable outcomes; however, the clinical significance of routinely measured immune and inflammatory markers remains uncertain. This retrospective exploratory study evaluated factors associated with poor prognosis and the individual discriminatory performance of immunoglobulin A (IgA), immunoglobulin G (IgG), immunoglobulin M (IgM), and C-reactive protein (CRP). A total of 172 children with MPP treated between April 2018 and April 2021, and 40 healthy controls were included. Biomarker levels were compared across disease severity and prognostic groups. Individual receiver operating characteristic (ROC) curve analyses and exploratory logistic regression were performed. Serum IgA, IgG, and IgM levels were significantly lower, whereas CRP levels were significantly higher, in children with severe disease and in those with poor prognosis than in the corresponding comparison groups (P < 0.05). The areas under the ROC curve (AUCs) for IgA, IgG, IgM, and CRP were 0.788, 0.707, 0.696, and 0.796, respectively. Fever duration, lesion type, immune dysfunction, elevated CRP, and extrapulmonary complications differed significantly between the prognostic groups. Exploratory multivariable analysis identified immune dysfunction and extrapulmonary complications as independent factors associated with poor prognosis. These findings support further investigation of immune and inflammatory biomarkers for risk stratification in children with severe MPP; however, prospective studies are needed to validate their clinical utility, and the present results should not be interpreted as a validated clinical prediction model.

Introduction

Severe pneumonia remains a major cause of morbidity in children and results from complex interactions among the infecting pathogen, host immune responses, and the inflammatory cascade^1,2,3. Mycoplasma pneumoniae (MP) is a common cause of community-acquired pneumonia in children, and M. pneumoniae pneumonia (MPP) accounts for a substantial proportion of pediatric pneumonia cases⁴. Although most children recover with appropriate treatment, some develop severe or refractory disease characterized by persistent fever, progressive pulmonary lesions, extrapulmonary complications, and other unfavorable outcomes^5,6,7,8. Early identification of children at increased risk of poor outcomes is therefore of considerable clinical importance.

Previous studies have investigated inflammatory biomarkers, imaging findings, and clinical characteristics associated with MPP severity and prognosis; however, the reported prognostic value of immune indices has been inconsistent⁹. Rather than developing a clinical prediction model, the present study provides a stratified comparison of routinely measured immunoglobulin A (IgA), immunoglobulin G (IgG), immunoglobulin M (IgM), and C-reactive protein (CRP) levels across healthy controls, children with mild-to-moderate MPP, children with severe MPP, and prognostic groups. In addition, an exploratory analysis of factors associated with poor prognosis in children with severe MPP was performed. The primary objective was to determine whether these routinely measured biomarkers and clinical characteristics were associated with poor prognosis. It was hypothesized that lower immunoglobulin levels, higher CRP, immune dysfunction, and extrapulmonary complications would be associated with unfavorable outcomes.

Protocol

This retrospective study was conducted in accordance with institutional requirements and the Declaration of Helsinki and was approved by the Ethics Committee of Nanjing Tongren Hospital Affiliated to Southeast University Medical School (approval no. 2024-03-039-K001). The requirement for individual informed consent was waived due to the retrospective design and the use of anonymized data.

Clinical data

A total of 172 children with MPP treated between April 2018 and April 2021 were included as the case group, together with 40 healthy children as controls. The case group comprised 103 boys and 69 girls, with a mean age of 7.15 ± 2.97 years; the control group comprised 24 boys and 16 girls, with a mean age of 7.02 ± 2.31 years. Baseline age and sex did not differ significantly between the groups (P > 0.05). According to the criteria described¹⁰, MPP was classified as severe when at least two of the following were present: evident dyspnea, heart rate >120 beats/min, fever lasting >10 days, multilobar involvement, or large-area pulmonary infiltration. The remaining cases were classified as mild-to-moderate MPP.

Inclusion criteria

All children with MPP met the clinical diagnostic criteria for MPP¹¹. Diagnosis was supported by an MP antibody titer >1:160, a positive MP DNA test, or a fourfold or greater increase in serum MP-specific IgG between the acute and recovery phases. Clinical manifestations included cough, shortness of breath, tachycardia, cyanosis, or dyspnea, and chest imaging showed unilateral or bilateral pulmonary lesions. Eligible participants were 8 months to 10 years of age.

Children were excluded if they had confirmed viral, bacterial, or fungal pneumonia; severe dysfunction of major organs, including the heart, brain, liver, or kidneys; another pulmonary disease; a malignant tumor; congenital immune dysfunction; or insufficient clinical information for the required analyses.

Children in the control group were 8 months to 10 years of age, were in good health, and had normal findings on physical examination. The control group was used only for biomarker comparisons and did not contribute to the prognostic or regression analyses.

Data collection

The hospital information system was reviewed for age, sex, fever duration, lesion type and location, immune function status, CRP level, extrapulmonary complications, treatment information, and follow-up outcomes. Immune dysfunction was determined from the contemporaneous medical record prior to outcome assessment, based on immunoglobulin values that fell below the hospital laboratory's age-specific reference intervals. Data were initially extracted by Boling Han and cross-checked against the original medical records by Rui Wang. Any discrepancies were resolved through discussion between the two investigators. Data completeness was verified before analysis. The complete variable definitions, coding rules, and de-identified cohort dataset are provided in Supplementary File 1.

Measurement of immune function indicators and inflammatory markers

Within 24 hours after enrollment, 4 mL of fasting venous blood was collected from each participant. Samples were centrifuged at 1,500 × g for 10 min at 4°C, and the resulting serum was stored in accordance with the institutional laboratory protocol until analysis. Serum IgA, IgG, and IgM concentrations were measured by immunoturbidimetry, and CRP was measured by enzyme-linked immunosorbent assay. Specifically, serum samples were stored at –80 °C for no longer than 3 months before analysis, and repeated freeze-thaw cycles were avoided.

Treatment protocol

Treatment was individualized according to clinical severity and the applicable institutional practice. Conventional management included anti-infective therapy and supportive care, including oxygen therapy, nebulized inhalation, fluid administration and maintenance of water and electrolyte balance, antipyretic treatment, and other symptomatic measures as clinically indicated. Children with severe MPP were treated with azithromycin, and methylprednisolone was used as adjunctive anti-inflammatory treatment when clinically indicated. Although all patients were managed according to the same institutional principles, the specific supportive measures were individualized rather than strictly standardized. Because of the retrospective design, exact treatment timing and regimen details were not uniformly available for all participants.

Prognosis follow-up

Children with severe MPP were followed monthly through outpatient visits and telephone contact until April 2022. Prognosis was classified as good when body temperature normalized, respiratory symptoms improved, and pulmonary lesions were absorbed or resolved. Poor prognosis was defined as a composite outcome comprising persistent fever, worsening cough or dyspnea, absent radiographic improvement or radiographic deterioration, transfer to another hospital, or death¹². The components could overlap and were not analyzed as mutually exclusive outcomes. All 114 children with severe MPP completed follow-up, and no participants were lost to follow-up.

Statistical analysis

Analyses were performed using IBM SPSS Statistics version 22.0. Continuous variables were summarized as mean ± standard deviation. Two-group comparisons used the independent-samples t test, and comparisons among more than two groups used one-way analysis of variance. Categorical variables were summarized as n (%) and compared using the chi-square test. ROC curves were used to evaluate IgA, IgG, IgM, and CRP individually, with AUCs, 95% confidence intervals, cutoff values, sensitivity, and specificity reported. The four biomarkers were not combined into a multivariable prediction model. Variables associated with prognosis in univariable analyses were entered into an exploratory stepwise binary logistic regression. Regression results were reported as beta coefficients, standard errors, Wald statistics, odds ratios (ORs), 95% confidence intervals, and P values. Because of the limited number of poor-prognosis events, the regression was interpreted as an exploratory association analysis rather than a validated prediction model. A two-sided P < 0.05 was considered statistically significant. The normality of continuous variables was assessed using the Shapiro-Wilk test, and homogeneity of variance was evaluated using Levene's test. There were no missing values in the variables included in the analyses; therefore, no imputation was performed, and complete-case analyses were used.

Results

Comparison of immune and inflammatory markers among groups

IgA, IgG, and IgM levels were lower in the severe group than in the mild-to-moderate group and were lower in the mild-to-moderate group than in the control group. CRP showed the opposite pattern. Pairwise differences were statistically significant (P < 0.05; Table 1).

Comparison of immune and inflammatory markers between prognostic groups

Discussion

The present retrospective study examined routinely available immune and inflammatory markers, as well as clinical factors, associated with poor prognosis in children with severe MPP. The principal contribution is not a new validated prediction tool, but a stratified description of IgA, IgG, IgM, and CRP across severity and prognostic groups, together with an exploratory assessment of clinical factors associated with unfavorable outcomes. This positioning is important because similar biomarkers have been investigated prev...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Clinical chemistry/immunoturbidimetric analyzer	Siemens Healthineers	BN II	Instrument used for IgA, IgG, and IgM testing
C-reactive protein ELISA kit	Thermo Fisher Scientific / Invitrogen	BMS288INST	Routine clinical laboratory measurement of serum CRP
IBM SPSS Statistics	IBM Corp.	Version 22.0	Statistical analysis
IgA immunoturbidimetric assay/reagent	Siemens Healthineers	NA	Routine clinical laboratory measurement of serum IgA
IgG immunoturbidimetric assay/reagent	Siemens Healthineers	NA	Routine clinical laboratory measurement of serum IgG
IgM immunoturbidimetric assay/reagent	Siemens Healthineers	NA	Routine clinical laboratory measurement of serum IgM
Microplate reader	Thermo Fisher Scientific	Multiskan FC	Instrument used for ELISA measurement
Microsoft Excel	Microsoft Corporation	Version 2024	Data organization and table preparation
Microsoft Word	Microsoft Corporation	Version 2024	Manuscript preparation