Introduction

Bronchiolitis is an inflammatory condition that affects terminal and respiratory bronchioles and has various causes1. Pathologically, it is characterized by inflammatory cell infiltration, secretion exudation, and potential local granulation tissue proliferation, leading to centripetal narrowing of bronchioles, ultimately resulting in bronchiolitis obliterans (BO)2,3. Acute infectious bronchiolitis is more common in bronchiolitis, and various pathogens can cause it, such as bacteria, viruses, Mycoplasma pneumoniae (MP), and fungi. Due to its low spatial resolution, bronchiolitis may manifest as nodular opacities on chest X-ray images. However, on high-resolution computed tomography (HRCT), it may present with the tree-in-bud sign due to bronchiolar secretions, centrilobular nodules, bronchiolar narrowing, and increased radiolucency due to gas trapping, creating mosaic attenuation compared to the surrounding normal lung tissue (mosaic sign)4.

MP is the most prevalent community-acquired pathogen among school-aged children in Asia, especially in China5. After the cancellation of COVID-19 epidemic control measures in 2023, there was a significant surge in cases of Mycoplasma pneumoniae pneumonia(MPP) in China6, accompanied by a gradual increase in the drug resistance rate of MPP7. Cases of pediatric bronchiolitis caused by MP have been reported successively8, with some children experiencing residual BO, prompting attention from pediatricians. Currently, the diagnosis of bronchiolitis MPP primarily relies on HRCT, as no well-established predictive model is available. The purpose of this study was to collect clinical data from children with MPP who underwent HRCT at our hospital and to analyze the predictors of bronchiolitis MPP. The purpose of this study was to provide evidence for the early diagnosis and treatment of bronchiolitis MPP and to reduce the use of HRCT and the occurrence of sequelae.

Methods

Patients

This was a retrospective analysis of clinical data from pediatric patients who were diagnosed with MPP and underwent chest HRCT in the First Affiliated Hospital of Xiamen University between June 2023 and December 2023. Our study protocol was approved by the Human Research Ethics Committee of the First Affiliated Hospital of Xiamen University. Guardians of all participants signed written informed consent at the time of hospital admission.

The inclusion criteria for patients were as follows: (1) met the diagnostic criteria for Mycoplasma pneumoniae: ① Single serum MP antibody titer ≥ 1:160 (PA method); during the disease, a fourfold or greater increase in antibody titers of two serum MP samples; ② Positive MP RNA results on RT‒PCR of pharyngeal swabs. (2) Chest HRCT was performed on admission, and the patients were divided into the following groups: ① Bronchiolitis MPP group: characterized by the presence of centrilobular nodules, tree-in-bud sign, and bronchial wall thickening, with or without patchy opacities, linear opacities, and areas of consolidation. ② Non-bronchiolitis MPP group: absence of bronchiolitis features. Disputed imaging features necessitated assessment by two or more pediatric radiologists.

Exclusion Criteria (1) Previously treated at other hospitals, with significant missing information regarding diagnosis and treatment. (2) Patients with underlying lung lesions that may complicate patient grouping, such as bronchiolitis obliterans, bronchiectasis, and isolated lung lesions. (3) Patients with serious underlying diseases or who had undergone hematopoietic stem cell transplantation. (4) Patients who were previously diagnosed with cured MPP but still tested positive for etiological agents, leading to hospitalization due to other pathogens. Among the 574 patients admitted with MPP who underwent HRCT, 42 were excluded. Finally, 532 children were included in this study and were randomly divided into a training cohort and a validation cohort at a 7:3 ratio (Fig. 1). The training cohort (372 in total, containing 155 patients with bronchiolitis MPP) was used to analyze the influencing factors of bronchiolitis MPP and construct a predictive model. The validation cohort (160 in total, containing 61 patients with bronchiolitis MPP) was used to verify the accuracy of the prediction model.

Fig. 1
figure 1

Flowchart of enrollment and exclusion of subjects.

Full size image

Data collection

The sex, age, clinical symptoms at admission, laboratory examination within 24 h of admission, etiological test, and imaging data during the acute phase were obtained from the case information system of the First Affiliated Hospital of Xiamen University. The laboratory parameters included the white blood cell (WBC) count, neutrophil percentage (N%), percentage of eosinophils (E%), lactic dehydrogenase (LDH), blood platelet (PLT) count, C-reactive protein (CRP), D-dimer ferritin (D-D), fibrinogen degradation products (FDP), and ferritin (Fer).

Definition

Hypoxemia: At the time of admission, before using β2 agonists and while breathing ambient air, hypoxemia is defined as SpO2 ≤ 0.90 in children under 1 year of age, and as SpO2 ≤ 0.93 in children aged 1 year and above. Tachypnea: a respiratory rate greater than 60 breaths per minute for infants under 2 months of age; greater than 50 breaths per minute for infants aged 2 to 12 months; greater than 40 breaths per minute for children aged 1 to 5 years; and greater than 30 breaths per minute for children older than 5 years. Coagulation disorders: D-D>1.5 ug/mL.

Statistical analysis

The data were organized and analyzed using R software version 4.3.3. Normally distributed continuous variables were presented as mean ± standard deviation (\(\bar x\)± s), and group comparisons were performed using t-tests. Non-normally distributed continuous variables were expressed as median (interquartile range) [P50 (P25-P75)], and group comparisons were conducted using the Mann‒Whitney U test. Categorical variables were reported as frequencies and percentages (%), with group differences assessed using the chi-square test and Fisher’s exact test. Variables exhibiting significance in univariate logistic regression analysis (P < 0.05) were included in multivariable logistic regression analysis. Nomogram models were constructed using R software based on regression coefficients. Internal validation was conducted using the Bootstrap method with 1,000 resamples. Model calibration was evaluated using the Hosmer-Lemeshow test, where P > 0.05 indicated satisfactory calibration. The discriminative ability of the model was assessed using receiver operating characteristic (ROC) curves and the area under the curve (AUC), with an AUC > 0.7 indicative of good discrimination. Calibration curves were utilized to assess the accuracy of the predictive models. Decision curve analysis (DCA) was employed to evaluate the clinical utility of the model.

Results

Demographic features of the enrolled patients

A total of 532 patients were enrolled in this study. Among them, 216 patients met the diagnostic criteria for bronchiolitis MPP, with a median age of 5 (3–7) years, while the remaining 316 patients were classified as non-bronchiolitis MPP with a median age of 7 (6–9) years. The significance of demographic factors and clinical manifestations is presented in Table 1. In addition to age, factors such as atopy, family history of atopy, wheezing, tachypnea, hypoxemia, depression of suprasternal fossae during breathing, and rare or decreased breath sound differed between children in the bronchiolitis MPP group and those in the non-bronchiolitis MPP group (P < 0.05).

Table 1 Characteristics of the subjects after admission. The data are expressed as the median (quartile distance) M (P25, P75) or the number of patients (percentage) n (%).WBC White blood cell count, N% neutrophil percentage, E% eosinophil percentage, PLT blood platelet counts, CRP C-reactive protein, LDH lactic dehydrogenase, Fer ferritin, D-D D-dimer ferritin, FDP Fibrinogen degradation products.
Full size table

Among the examined laboratory parameters, WBC, PLT, and LDH levels were greater in the bronchiolitis MPP group compared to the non-bronchiolitis MPP group (P < 0.05). Conversely, the levels of Fer and CRP were lower in the bronchiolitis MPP group compared to the non-bronchiolitis MPP group (P < 0.05). The rates of pleural effusion and coagulation disorders were lower in the bronchiolitis MMP group compared to the non-bronchiolitis MPP group (P < 0.05).

Regarding co-infections, there were no differences between the two groups. (P > 0.05) (Table 2 ).

Table 2 Co-infection of Bronchiolitis MMP group and non-bronchiolitis MMP group. Other viruses include human pneumoviruses, bocaviruses, coronaviruses, and the novel SARS-CoV-2. Other bacteria include Streptococcus agalactiae, Escherichia coli, Streptococcus pyogenes, Acinetobacter baumannii, and Streptococcus anginosus.
Full size table

Patients were randomly divided into a training cohort and a validation cohort at a 7:3 ratio. There were no differences in general characteristics, clinical features, or laboratory examination results, between the two cohorts (P > 0.05).

Univariate and multivariate logistic regression analysis of the predictive factors for bronchiolitis MMP

Univariate analysis of the training cohort revealed that age, atopy, wheezing, tachypnea, hypoxemia, depression of suprasternal fossae when breathing, rale, decreased breath sound, WBC, PLT, CRP, Fer, and pleural effusion were significantly different between the bronchiolitis MPP group and the non-bronchiolitis MPP group. Significant differences (P < 0.05). Logistic stepwise regression analysis revealed that age (OR = 0.86, 95% CI: 0.76–0.96), atopic constitution (OR = 2.28, 95% CI: 1.24–4.17), wheezing(OR = 27.81, 95% CI: 6.77–114.20), hypoxemia (OR = 4.02, 95% CI: 1.06–15.22), and pleural effusion (OR = 0.09, 95% CI: 0.02–0.40) were found to be independent predictors of bronchiolitis MPP (Table 3).

Table 3 Univariate and multivariate logistic regression analyses of predicted factors for the development of bronchiolitis MMP. WBC White blood cell count, N% neutrophil percentage, E% eosinophil percentage, PLT blood platelet count, CRP C-reactive protein, LDH lactic dehydrogenase, Fer ferritin, D-D D-dimer ferritin, FDP Fibrinogen degradation products.
Full size table

Establishment and validation of the nomogram model

A nomogram model for bronchiolitis MMP was constructed based on the logistic regression results. As shown in Fig. 2, a total of 5 independent variables, including age, atopy, wheezing, hypoxemia, and pleural effusion, were applied for the construction of the model. The scores of each predictive factor were calculated individually and the total scores were calculated. The corresponding value of the total score was the predicted probability of bronchiolitis MMP.

Fig. 2
figure 2

A nomogram model for predicting bronchiolitis MMP.

Full size image

Model evaluation

An evaluation of the nomogram model was then performed. In the training cohort, the area under the ROC curve of the nomogram model was 0.83 (95% CI: 0.79–0.88; Fig. 3), with a sensitivity of 0.751 and a specificity of 0.768. In the validation cohort, the area under the ROC curve (AUC) was 0.87 (95% CI: 0.76–0.90; Fig. 3), with a sensitivity of 0.838 and a specificity of 0.689.

Fig. 3
figure 3

ROC curve of the nomogram model.

Full size image

Moreover, as depicted in Fig. 4, the calibrated curve of the nomogram model for predicting prognosis in children with bronchiolitis MPP closely approximated the ideal curve in both the training and validation cohorts. Additionally, excellent consistency in predicting the risk of death among children with bronchiolitis MPP was confirmed by the Hosmer–Lemeshow goodness-of-fit test (χ2 = 5.612, P = 0.691). The decision curve analysis (DCA) for the model is presented in Fig. 5, suggesting that our nomogram benefits across a threshold probability range of 9–93%.

Fig. 4
figure 4

Calibration curve of the nomogram model. (a) Training cohort. (b) The validation cohort.

Full size image
Fig. 5
figure 5

The decision curve analysis plots of the nomogram. (a) The training cohort. (b) The validation cohort.

Full size image

Discussion

The pathogenesis of MPP remains incompletely understood. Most researchers propose that respiratory symptoms caused by MP result from direct injury, such as adhesion, nutrient deprivation, invasion, toxin production, cytokine-induced inflammatory injury, and immune evasion. Extrapulmonary manifestations are attributed to direct injury mediated by invasion and inflammatory factors and indirect injury such as vascular occlusion triggered by the host immune response9. Severe Mycoplasma pneumoniae pneumonia (SMPP) in children may lead to complications such as necrotizing pneumonia, bronchiolitis obliterans, and bronchiectasis10.

In current studies on MPP, WBC counts are generally within the normal range, but there is an imbalance in Th1/Th211. The WBC counts in both groups of MPP also fall within the normal range. Tanaka13 found that the upregulation of host cell-mediated immunity primarily caused centrilobular nodules. Xuefeng Xu et al.14. also found that CD4 + T cell counts were significantly associated with different radiological manifestations of MPP. In this study, the WBC counts in the bronchiolitis MPP group were slightly higher than that in the non-bronchiolitis group, which may suggest an imbalance in different immune cells, although further research is needed.

Our study revealed that younger age increases the likelihood of bronchiolitis MPP. Radiographic findings are closely associated with age. A study by Yeon Jin Cho et al.12 suggested that older children were more likely to exhibit lobar or segmental consolidation, whereas infants under 2 years of age were more prone to peribronchial infiltrates. The lung anatomy of infants differs from that of older children. The small airways in infants significantly impact total airway resistance, making them more susceptible to obstruction from the same degree of small airway inflammation compared to older children16. Additionally, the lack of effective collateral circulation in young children may further contribute to atelectasis. From late gestation to childhood, airway smooth muscle grows, while airway responsiveness to contractile stimuli is higher in young children17. Therefore, airway obstruction in young infants may mainly be due to mucosal swelling and secretion blockage. The bronchiolitis MPP group consisted of younger children, which may also explain the lower CRP and Ferritin levels, as younger children tend to have a milder systemic inflammatory response compared to older children15.

Children with atopy are more predisposed to developing bronchiolitis MPP. Those with atopy demonstrate heightened airway responsiveness18, potentially resulting in more severe smooth muscle contraction and increased susceptibility to bronchiolar secretion blockage. Furthermore, airway hyperresponsiveness and compromised lung function foster airway remodeling in children, triggering Th2 responses that precipitate asthma attacks. Simultaneously, MP infection can elicit Th2 reactions in asthmatic children postinfection, thereby establishing a bidirectional effect19. Children with atopy usually have epithelial cell dysfunction and decreased ciliary motion, and atopy may be a risk factor for refractory Mycoplasma pneumoniae pneumonia (RMPP)20.

After MP infection, wheezing and hypoxemia are more likely to occur in children with bronchiolitis. When inflammation occurs in the cartilage-free structures of the small bronchi, narrowing during inhalation becomes more likely1. The smooth muscles surrounding the small bronchi develop into a ring shape and are prone to spasms under various stimuli, leading to respiratory obstruction, air trapping, and wheezing. Wheezing can be categorized into early transient wheeze, early-onset persistent wheeze, and late-onset wheeze/asthma. Early-onset persistent wheeze is primarily associated with respiratory tract infections, while children with atopy are more predisposed to developing delayed wheeze/asthma21. A meta-analysis involving nearly 300,000 children22 suggested that recurrent wheezing or asthma may ensue after recovery from bronchiolitis, suggesting that bronchiolitis is a significant risk factor for asthma. Additionally, MP infection heightens the risk of recurrent wheezing in children with inhaled allergen-positive bronchiolitis23. This heightened risk may stem from shared antigens between the MP and the human body. After MP infection, the body can produce autoantibodies to form immune complexes, cause cross immune responses, and trigger cross-reactions, leading to lung and other tissue damage24. Previous studies have found that MPP patients with allergic conditions have higher serum IgE levels and IL-4/INF-γ ratio2025, which may maintain and amplify MP-mediated airway injury and bronchial hyperresponsiveness and be associated with the delayed occurrence of recurrent wheezing.

The results of this study suggest that the absence of pleural effusion indicates bronchiolitis MPP. According to univariate regression, the patients in the bronchiolitis MPP group exhibited lower CRP and ferritin levels and a lower rate of coagulation disorders. Other studies have indicated that inflammatory markers vary among children with different imaging findings, and elevated levels of inflammatory cytokines26 are associated with the extent of consolidation. The incidence of pleural effusion is directly proportional23, indicating the potential for severe, refractory Mycoplasma pneumoniae pneumonia27,28. The lower incidence of pleural effusions and decreased inflammatory markers in patients with bronchiolitis MPP suggest a milder inflammatory response and a lower probability of progression to RMPP than other imaging findings. Dyspnea may be better relieved by timely and effective treatment.

In clinical practice, not only MP can lead to bronchiolitis; other common pathogens such as rhinovirus, respiratory syncytial virus, and adenovirus can also cause bronchiolitis. To reflect clinical reality more closely, this study did not exclude these co-infections. In our comparison of co-infection, there were no significant differences between the two groups. Therefore, we thought that these pathogens have minimal interference in this study, while more comprehensive research is needed in the future. Limitations of this study include the following: (1) it was a retrospective cohort study with some limitations in data collection and a lack of more specific laboratory tests as predictors, and (2) the present study was only based on our single institutional clinical data. The model lacks other units as an external test cohort, and it needs validation in a large prospective study.

In summary, age, atopy, wheezing, hypoxemia, and pleural effusion can be combined to establish a nomogram. The accuracy, differentiation, and clinical applicability of the model are good. It effectively identifies children with bronchiolitis MPP, providing clinicians with an accurate and efficient tool for diagnosing bronchiolitis MPP and reducing the use of HRCT. This facilitates early individualized intervention by doctors.