Association between normalized Mycoplasma pneumoniae sequence reads and severity of pediatric pneumonia: a cross-sectional study

Introduction

Mycoplasma pneumoniae (MP) is a predominant pathogen responsible for community-acquired pneumonia (CAP) in children, particularly affecting school-aged populations. Epidemiological data indicate that MP accounts for approximately 10% to 40% of hospitalized CAP cases in children, with even higher prevalence in certain Asian countries (1). In recent years, the widespread use of antibiotics and emergence of macrolide-resistant strains—such as the A2063G mutant—have altered the epidemiological characteristics of MP pneumonia (MPP), leading to an increasing proportion of severe cases (2). Early identification of severe cases is therefore critical for timely intervention and improved outcomes; however, effective biomarkers for predicting disease progression are currently lacking. Conventional biomarkers of severe MPP (SMPP) [C-reactive protein (CRP), D-dimer (DDi), lactate dehydrogenase (LDH)] lack specificity for MP infection and exhibit delayed kinetics, limiting early prediction (3). The neutrophil-to-lymphocyte ratio (NLR) shows inconsistent performance in pediatric populations (4).

Traditional diagnostic methods for MPP, including serological assays and polymerase chain reaction (PCR), offer moderate sensitivity but have limitations in early detection, specificity, and the risk of false positives, especially in cases with low bacterial loads or delayed immune response (5). Recent advances in next-generation sequencing (NGS) have revolutionized the detection of microorganisms in clinical samples. Targeted NGS (tNGS) (6) utilizes specific probes or primers to capture target DNA sequences, which are then sequenced using high-throughput methods. Compared with whole-genome sequencing (WGS) or metagenomic NGS (mNGS), tNGS minimizes irrelevant sequencing, thereby increasing coverage and depth of target regions. This method not only identifies pathogens but also differentiates subtypes and detects resistance and virulence genes, all while reducing costs. Numerous studies have confirmed the advantages of tNGS (7-9), leading to its adoption as a first-line diagnostic tool for MPP (10).

In tNGS, normalized sequence reads, defined as the number of sequencing reads mapped to a specific microorganism per 100,000 total sequencing reads, are used to mitigate variations in sequencing depth among samples, reflecting the relative abundance of the microorganism. One study involving 770 patients who underwent both quantitative PCR and tNGS demonstrated a positive correlation between bacterial load and normalized sequence reads (r=0.3774, P<0.001) (11). Thus, tNGS can quantify microbial load, which may correlate with disease severity.

Despite prior studies linking MP load to disease severity (12-14), the clinical significance of normalized sequence reads from oropharyngeal swabs remains unclear. Clarifying this relationship could provide a quantitative marker for early prediction of SMPP. This study aims to evaluate the relationship between normalized MP sequence reads derived from tNGS and clinical severity in children with MPP. By exploring the association between normalized MP sequence reads and clinical outcomes, we seek to identify potential early biomarkers for SMPP and provide evidence to support precision management strategies in pediatric MPP. We present this article in accordance with the STROBE reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-1-804/rc).

Methods

Study participants

This retrospective, single-center clinical study was conducted at Fujian Children’s Hospital. We collected data on consecutive pediatric patients hospitalized with MPP from January 2023 to December 2023. Patients without chest X-ray or computed tomography (CT) imaging evidence (n=41), and those lacking complete oropharyngeal swab tNGS testing (n=39) were excluded. Ultimately, a total of 262 children with MPP were included in the final analysis (Figure 1). Sample size was estimated via G*Power 3.1 (effect size =0.2, α =0.05, power =80%), requiring 238 patients; 262 were enrolled to account for missing data.

Figure 1 Flowchart of patient selection. The final study cohort consisted of 262 patients equally divided into GMPP and SMPP groups based on severity criteria defined at admission. CT, computed tomography; GMPP, general Mycoplasma pneumoniae pneumonia; MP, Mycoplasma pneumoniae; MPP, Mycoplasma pneumoniae pneumonia; ROC, receiver operating characteristic; SMPP, severe Mycoplasma pneumoniae pneumonia; tNGS, targeted next-generation sequencing.

Data extracted from medical records included demographic information, clinical manifestations, laboratory test results, and imaging findings. The inclusion criteria were as follows: (I) patients diagnosed with MPP with no age restriction; (II) chest CT imaging and clinical indicators obtained at admission; and (III) at least one of the following laboratory criteria: (i) a fourfold or greater increase in serum MP antibody titers between the acute and convalescent phases; or (ii) a positive MP culture or detection of MP DNA/RNA. Based on the Chinese Guidelines for the Management of CAP in Children (2019 revised edition) (15), patients were classified into two groups at the time of hospital admission, prior to the initiation of any major therapeutic interventions. SMPP was defined as MPP meeting any of the following criteria: (I) poor general condition; (II) disturbance of consciousness; (III) hypoxemia, manifested as any of the following: cyanosis; respiratory rate (RR) ≥70/min (infants) or ≥50/min (children >1 year old); accessory muscle use (grunting, nasal flaring, chest retractions); intermittent apnea; or oxygen saturation <92%; (IV) high-grade fever, specifically ultra-high fever or persistent high fever >5 days; (V) signs of dehydration or refusal to feed; (VI) severe radiographic findings, including pulmonary infiltration involving ≥2/3 of a single lung, multi-lobar infiltration, pleural effusion, pneumothorax, atelectasis, pulmonary necrosis, or lung abscess; (VII) extrapulmonary complications. General MPP (GMPP) was defined as pediatric CAP patients who fulfilled the diagnostic criteria for MPP but did not meet any of the severity criteria for SMPP listed above. The exclusion criteria included congenital heart disease, inherited metabolic disorders, neurological diseases, congenital bronchopulmonary malformations, immunodeficiency, long-term use of immunosuppressive agents, pulmonary tuberculosis, asthma, and bronchial foreign bodies.

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments, and approved by the Medical Ethics Committee of Fujian Children’s Hospital (approval No. 2025ETKLRK01005). Written informed consent was obtained from all participants’ guardians.

Data collection

We collected pediatric patient records, including general demographic information (age and sex), clinical data (duration from symptom onset to admission, length of stay, duration of fever, use of oxygen therapy, corticosteroids, second-line antibiotics, and fiberoptic bronchoscopy), laboratory parameters (CRP, LDH, serum ferritin, and DDi), imaging findings (pulmonary consolidation, atelectasis, and pleural effusion), and tNGS sequencing results. LDH was measured using a Roche Cobas c702 analyzer (Roche Diagnostics, Germany) with reagents from Jiujiang Strong Biotech Co., Ltd. (Beijing, China). DDi was measured using a Stago STA-R Evolution coagulometer (Stago, France) with proprietary reagents. CRP was measured using an immunoturbidimetric assay with reagents from Suzhou Dewo Biotechnology Co., Ltd. (Suzhou, China). Radiological assessments were conducted by two independent radiologists blinded to clinical outcomes. Pulmonary consolidation was quantified as single-lobar or multi-lobar involvement based on chest CT. Inter-observer agreement was excellent (Cohen’s κ =0.89, indicating almost perfect agreement). All examinations were completed within 24 hours of hospital admission. Patients with incomplete data were excluded from the analysis.

Specimen collection

Oropharyngeal swabs were collected according to a standardized protocol within 24 hours of hospital admission and before initiation of antibiotic therapy. Patients were instructed to rinse their mouths with saline; if unable to do so, a small amount of water was provided for gargling. Subsequently, patients were asked to open their mouths and vocalize an “ah” sound to expose the oropharynx. When necessary, a tongue depressor was used to gently retract the tongue. The swab was then inserted beyond the base of the tongue and used to sample the posterior pharyngeal wall, the region behind the uvula, and both tonsillar pillars with moderate pressure and rotational motion to ensure adequate contact, while avoiding contact with the tongue. The swab was immediately placed into viral transport medium or an equivalent buffer solution.

Process of tNGS

All samples were processed using a standardized nucleic acid extraction protocol, library preparation workflow, sequencing platform, and bioinformatics pipeline to minimize technical variability across samples and ensure consistency of sequencing quality. Collected samples were sent to Fuzhou Kingdem Diagnostics Laboratory Co., Ltd. for tNGS. The pathogen panel initially targeted 198 common respiratory pathogens, including 79 viruses, 80 bacteria, 32 fungi, and 7 species of chlamydia/mycoplasma. It has since been upgraded to cover 225 targets (81 viruses, 93 bacteria, 43 fungi, and 8 chlamydia/mycoplasma species), as detailed in table available at https://cdn.amegroups.cn/static/public/tp-2025-1-804-1.xls.

Nucleic acid extraction was performed using a nucleic acid extraction and purification kit (Meiji Biotechnology Co., Ltd., Guangzhou, China) on an automated extraction instrument (Jiangsu Moluo Biotechnology Co., Ltd., Taizhou, China). Nuclease-free water (Invitrogen, California, USA) was used as a blank control to monitor for potential environmental or reagent contamination during nucleic acid extraction. The extracted nucleic acids were subsequently amplified and prepared for sequencing using the Respiratory Pathogen Multiplex Detection Kit RP100 (Guangzhou Jinqi Ruisheng Biotechnology Co., Ltd., Guangzhou, China). Library concentration was measured using the Invitrogen Qubit 4.0 Fluorometer with the Qubit™ dsDNA HS Assay Kit (Thermo Fisher Scientific, California, USA), with a quality control threshold of ≥0.5 ng/µL.

Library fragment size was assessed using the Qsep100 system (Hangzhou Houze Biotechnology Co., Ltd., Hangzhou, China); fragments ranging from 250 to 350 bp were required to comprise at least 20% of the total library. Sequencing was performed using the MR100 Sequencing Reaction Kit (Guangzhou Jinqi Ruisheng Biotechnology Co., Ltd., Guangzhou, China) on the KM MiniSeq Dx-CN next-generation sequencer (Guangzhou Jinqi Ruisheng Biotechnology Co., Ltd., Guangzhou, China) (7).

Bioinformatics analysis

Sequencing data were processed using the following workflow: Raw reads were subjected to quality control using fastp v0.20.1 with default settings to trim adapters and remove low-quality sequences (16). The cleaned reads were aligned to a clinical-grade, tNGS-specific pathogen database using Bowtie2 v2.4.1 in “very-sensitive” mode (17). Based on the alignment results, taxonomic classification was performed to determine the microbial composition, and the normalized sequence reads [reads per hundred thousand (RPhK)], defined as the number of sequencing reads mapped to a specific pathogen per 100,000 total sequencing reads in each sample, were calculated. Because tNGS data may contain a substantial proportion of host-derived sequences, normalization using RPhK was applied to partially account for variation in sequencing depth and the relative proportion of host DNA content across samples. Similar normalization strategies have been widely used in clinical metagenomic sequencing studies to enable cross-sample comparison of microbial abundance (18,19). For this retrospective study, positivity thresholds were defined as follows: RPhK ≥7 for viruses, ≥15 for bacteria, and ≥11 for fungi (7).

Covariates

To assess the influence of potential confounding factors, several key covariates were selected based on previous literature (2,20,21), including age, sex, pre-admission disease duration, fever duration (in days), and coinfection status, and macrolide resistance (A2063G mutation). Missing data occurred in 2 (0.7%) patients for DDi, 5 (1.9%) for LDH, and 6 (2.2%) for ferritin. Missing values were imputed using mean substitution.

Statistical analysis

Categorical variables were reported as counts and percentages. Continuous variables were assessed for normality via the Shapiro-Wilk test and Q-Q plots, with results presented as mean ± standard deviation (normal distribution) or median (interquartile range) (non-normal distribution). Comparisons were performed using the chi-square test for categorical variables, one-way ANOVA for normally distributed continuous variables, and the Kruskal-Wallis test for non-normally distributed continuous variables.

Patients were stratified into four quartiles based on normalized MP sequence reads obtained from oropharyngeal swab tNGS. Quartiles of MP reads: Q1 (114–17,947), Q2 (17,948–31,268), Q3 (31,269–43,753), and Q4 (43,754–65,091). To explore potential nonlinear associations between normalized MP sequence reads and severe MPP, restricted cubic spline (RCS) analysis was performed within the logistic regression framework. Covariates were selected via clinical expertise; adjusted model included age, sex, pre-admission duration, fever duration, coinfection, and A2063G mutation. Model discrimination was evaluated using receiver operating characteristic (ROC) curves and the area under the curve (AUC), with comparisons performed using the DeLong test. Model calibration was assessed using the Hosmer-Lemeshow goodness-of-fit test and calibration plots. Incremental predictive value was evaluated using integrated discrimination improvement (IDI) and continuous net reclassification improvement (NRI), while clinical utility was assessed using decision curve analysis (DCA).

Results were reported as odds ratios (ORs) with 95% confidence intervals (CIs). All statistical analyses were conducted using R software (version 4.2.2; http://www.R-project.org, The R Foundation for Statistical Computing, Shanghai, China) and Free Statistics software (version 2.1.1). Two-tailed tests were used, and a P value <0.05 was considered statistically significant.

Results

Demographic and clinical characteristics of MPP patients stratified by severity (general vs. severe cases)

A total of 262 children diagnosed with MPP were enrolled, consisting of 131 cases of GMPP and 131 cases of SMPP. As summarized in Table 1, no significant differences were observed between the two groups in sex distribution, age, or pre-admission disease duration (all P>0.05).

Compared with the GMPP group, the SMPP group had significantly longer length of stay (7.4±2.0 vs. 6.2±1.6 days, P<0.001) and fever duration [7.0 (6.0–8.0) vs. 5.0 (4.0–7.0) days, P<0.001]. The SMPP group also had higher rates of oxygen therapy (6.1% vs. 0%, P=0.007), glucocorticoid administration (85.5% vs. 63.4%, P<0.001), second-line antibiotic use (48.1% vs. 32.1%, P=0.008), and fiberoptic bronchoscopy (59.5% vs. 15.3%, P<0.001). Inflammatory markers, including CRP, DDi, and serum ferritin were significantly elevated in the SMPP group (all P<0.05), while LDH levels did not differ significantly (P=0.28). Notably, normalized MP sequence reads detected from oropharyngeal swabs by tNGS were significantly higher in the SMPP group [35,296 (20,004.5–47,737.0) vs. 28,602.0 (15,539.5–39,840.5), P=0.01]. Moreover, pulmonary consolidation was more frequently observed in the SMPP group compared with GMPP (58.8% vs. 33.6%, P<0.001). The prevalence of the macrolide resistance A2063G mutation was also significantly higher in SMPP patients (25.2% vs. 6.1%, P<0.001).

Demographic and clinical characteristics of MPP patients stratified by normalized MP sequence reads

When stratified by quartiles of normalized MP sequence reads obtained via tNGS (Q1–Q4), the Q4 group (highest normalized reads) exhibited significantly elevated levels of CRP [25.3 (13.3, 33.0) mg/L, P<0.001] and DDi [0.6 (0.4, 1.0) ng/mL, P=0.03], suggesting greater disease severity. Consistently, the Q4 group also showed higher frequencies of glucocorticoid administration (86.4%, P=0.04) and fiberoptic bronchoscopy (56.1%, P=0.004), further indicating a more severe clinical presentation compared with the lower quartiles (Table 2). Notably, the proportion of SMPP cases increased progressively across quartiles, with the highest prevalence in Q4 (P<0.05).

Higher normalized MP sequence reads were significantly associated with increased disease severity

Multivariable logistic regression analysis was performed to evaluate the association between normalized MP sequence reads and the risk of SMPP. In the unadjusted model, higher normalized MP sequence reads were significantly associated with increased odds of SMPP (OR =1.22, 95% CI: 1.05–1.43, P=0.01). After adjusting for potential confounders including age, sex, duration from symptom onset to admission, duration of fever, coinfection status, and A2063G mutation, the association remained statistically significant (adjusted OR =1.19, 95% CI: 1.01–1.41, P=0.04), suggesting that MP normalized sequence reads may be an independent risk factor for SMPP. RCS analysis demonstrated a linear association between normalized MP sequence reads and the risk of SMPP (P for non-linearity =0.801, Figure 2), with no evidence of a threshold effect, thereby justifying the use of a linear model.

Figure 2 Association between the normalized MP normalized sequence reads and the severity of MPP. Solid and dashed lines represent the predicted value and 95% confidence intervals. They were adjusted for age, sex, duration from symptom onset to admission, fever duration, coinfection, and A2063G mutation. MP, Mycoplasma pneumoniae; Ref, reference; SMPP, severe Mycoplasma pneumoniae pneumonia.

As shown in Table 3, when stratified by quartiles of normalized sequence reads, patients in the highest quartile (Q4) had a significantly increased risk of SMPP compared to Q1 (lowest quartile), with an adjusted OR of 2.69 (95% CI: 1.26–5.75, P=0.01). These findings suggested that higher normalized MP sequence reads may be associated with disease severity in MPP.

Association between normalized MP sequence reads and clinical parameters

Spearman correlation analysis (Figure 3) demonstrated that normalized MP sequence reads were positively correlated with CRP (r=0.27, P<0.001), DDi levels (r=0.17, P<0.01), LDH (r=0.12, P<0.05) and ferritin (r=0.12, P<0.05), indicating that higher pathogen loads are associated with enhanced inflammatory responses. Significant inter-correlations were also observed among inflammatory markers, particularly between CRP and DDi (r=0.35, P<0.001).

Figure 3 Association between normalized MP normalized sequence reads and clinical parameters. X-axis: clinical parameters and normalized MP sequence reads; Y-axis: correlation coefficient. *, P<0.05; **, P<0.01; ***, P<0.001. Pre-admission, duration from symptom onset to admission. CRP, C-reactive protein; DDi, D-dimer; LDH, lactate dehydrogenase; MP, Mycoplasma pneumoniae.

Model performance and incremental value

As a standalone predictor, normalized MP sequence reads demonstrated modest discriminative ability with an AUC of 0.592 (Figure 4A). The baseline clinical model (incorporating age, sex, fever duration, and A2063G mutation) achieved an AUC of 0.710 (95% CI: 0.648–0.771). Upon the addition of normalized MP sequence reads, the full model’s AUC increased to 0.723 (95% CI: 0.662–0.784). Although DeLong’s test indicated no statistically significant difference in AUC between the two models (Z =−0.857, P=0.40), the inclusion of normalized MP sequence reads yielded a significant IDI (IDI =0.015, 95% CI: 0.0002–0.0567, P=0.01), suggesting an enhanced overall predictive capability. However, the continuous NRI did not reach statistical significance (NRI =0.183, 95% CI: −0.0448 to 0.4656, P=0.11).

Figure 4 Evaluation of the predictive performance and clinical utility of the models. (A) ROC curves. The red curve represents the predictive performance of normalized MP sequence reads alone (AUC =0.592). The blue dashed curve represents the baseline clinical model (incorporating age, sex, fever duration, and A2063G mutation; AUC =0.710). The green curve represents the integrated model (clinical model + normalized MP sequence reads), which achieved an improved AUC of 0.723. (B) Calibration curve of the integrated model. The x-axis shows the predicted probability of SMPP, and the y-axis shows the observed frequency. The diagonal dashed line represents a perfect prediction (ideal), and the solid line represents the model’s performance based on 1,000 bootstrap resamples, indicating excellent agreement. (C) DCA. The net clinical benefit of the clinical model, MP reads model, and combined model is shown across different threshold probabilities. The combined model (full model) yields the highest net benefit. AUC, area under the curve; DCA, decision curve analysis; MP, Mycoplasma pneumoniae; ROC, receiver operating characteristic; SMPP, severe Mycoplasma pneumoniae pneumonia.

Internal validation using 1,000 bootstrap resamples yielded an optimism-corrected AUC of 0.723 (95% CI: 0.662–0.796), confirming the model’s robustness. The full model exhibited excellent calibration, as evidenced by both the Hosmer-Lemeshow test (χ² =6.67, P=0.57) and visual inspection of the calibration curve (Figure 4B), which showed minimal deviation from the ideal 45° reference line. Furthermore, DCA (Figure 4C) demonstrated greater net clinical benefit for the combined model (full model) across a wide range of threshold probabilities.

Discussion

This retrospective study demonstrated that children with SMPP exhibited significantly higher normalized MP sequence reads compared to those with GMPP. Even after adjusting for potential confounders, including macrolide resistance, elevated normalized reads remained independently associated with increased disease severity. Moreover, the positive correlations between pathogen load and systemic inflammatory biomarkers—such as DDi, LDH, and ferritin—support the hypothesis that excessive bacterial burden amplifies the host inflammatory cascade, contributing to SMPP pathogenesis (22-24). These findings highlight the potential role of pathogen load as both a mechanistic contributor and a prognostic factor in pediatric MPP. Interestingly, our study found no significant difference in LDH levels between the GMPP and SMPP groups, which contrasts with some prior reports. This discrepancy may be attributed to the relatively early admission of our cohort (median 6.9 days from onset) and the absence of patients with acute respiratory distress syndrome (ARDS), as LDH typically serves as a hallmark of late-stage severe lung injury and extensive tissue destruction.

Our findings are consistent with previous quantitative PCR (qPCR)-based studies demonstrating that higher MP DNA copy numbers in respiratory samples are associated with prolonged fever, extensive pulmonary consolidation, and extended hospitalization (12,13,25). Beyond confirming this trend, our data provide novel evidence that tNGS-derived normalized reads can serve as a semi-quantitative indicator of microbial burden. This builds upon our earlier finding that NGS-based platforms provide rapid and effective pathogen diagnosis in high-risk pediatric populations (26). Unlike qPCR, tNGS allows simultaneous detection of multiple pathogens while minimizing variability introduced by sequencing depth and library size (7,27). The observed moderate correlation between tNGS reads and qPCR copy numbers (r=0.38, P<0.001) underscores its reliability for pathogen quantification, although absolute quantification remains technically demanding (11).

Mechanistically, the association between higher normalized MP reads and disease severity may be mediated by the Community-Acquired Respiratory Distress Syndrome (CARDS) toxin. MP burden is directly proportional to CARDS toxin production, which possesses adenosine diphosphate (ADP)-ribosylating activity that triggers pro-inflammatory cytokine cascades and induces significant vacuolation and ciliostasis in airway epithelial cells. Our findings of a positive correlation between pathogen load and systemic inflammatory markers (CRP, DDi, and ferritin) support the hypothesis that a high bacterial burden serves as a potent stimulus for the host’s innate immune response, potentially leading to a “cytokine storm” that characterizes SMPP (22,27,28).

Clinically, early recognition of children at risk for SMPP is essential, particularly given the rising prevalence of macrolide-resistant strains (2). In this study, normalized MP reads alone provided modest discriminative ability (AUC =0.592), indicating limited utility when used alone to identify high-risk patients. This relatively modest performance may reflect the multifactorial pathogenesis of SMPP, in which host immune responses, pathogen virulence, and treatment timing all contribute to disease progression. When integrated with clinical covariates and macrolide resistance status, predictive accuracy substantially improved (AUC =0.723). Importantly, beyond the modest increase in AUC, the addition of normalized MP reads yielded a significant IDI (IDI =0.015, P=0.01), suggesting that pathogen load provides substantial incremental value in refining risk stratification. This clinical utility was further corroborated by DCA, which demonstrated that the full model achieved a higher net clinical benefit than the clinical model and MP reads model across a wide range of threshold probabilities. These findings suggest that normalized MP reads may serve as a valuable adjunctive biomarker within a multidimensional assessment framework combining imaging, inflammatory indices, and clinical course parameters, thereby facilitating timely therapeutic escalation.

There are several limitations in this study. First, its single-center and retrospective design may limit generalizability and introduce selection bias. Second, only oropharyngeal swabs collected at admission were analyzed, which may not fully represent lower respiratory tract pathogen burden or temporal dynamics. Nevertheless, oropharyngeal swabs represent a non-invasive and clinically practical sampling approach in pediatric practice. Third, although tNGS offers high sensitivity and broad-spectrum detection, it cannot distinguish viable from non-viable organisms, and quantitative validation against qPCR was not systematically conducted due to sample constraints. Finally, as tNGS provides semi-quantitative rather than absolute measurements, normalized reads should be interpreted as relative indicators of microbial load rather than precise counts. In addition, variations in host DNA content may influence the relative abundance of microbial reads in sequencing datasets, which represents an inherent limitation of semi-quantitative metagenomic approaches. Future studies incorporating longitudinal sampling and paired molecular assays are warranted to address these limitations.

Future investigations should validate these findings in large-scale, prospective, multicenter cohorts and explore integration of normalized MP reads with host immune signatures, radiological features, and clinical trajectories. Integration of advanced machine learning algorithms or multi-omics approaches could further refine predictive algorithms and enable precision management strategies for pediatric MPP. Ultimately, incorporating pathogen load metrics into routine clinical workflows may improve early risk stratification and guide individualized therapeutic interventions.

Conclusions

In conclusion, this study demonstrates a significant association between normalized MP sequence reads—quantified via tNGS—and both clinical severity and systemic inflammation in pediatric MPP. Although its standalone predictive value is modest, pathogen load offers significant incremental value to clinical models. These findings support normalized MP sequence reads as a potentially valuable adjunctive biomarker for risk stratification and clinical management of severe pediatric pneumonia.

Acknowledgments

We thank the staff of Fujian Children’s Hospital for their assistance with data collection.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-1-804/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-1-804/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2025-1-804/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2025-1-804/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Medical Ethics Committee of Fujian Children’s Hospital (No. 2025ETKLRK01005). Written informed consent was obtained from all participants’ guardians.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Kutty PK, Jain S, Taylor TH, et al. Mycoplasma pneumoniae Among Children Hospitalized With Community-acquired Pneumonia. Clin Infect Dis 2019;68:5-12. [Crossref] [PubMed]
2. Dekyi Xiao Y. Predominance of A2063G mutant strains in the Mycoplasma pneumoniae epidemic in children: A clinical and epidemiological study in 2023 in Wuhan, China. Int J Infect Dis 2024;145:107074. [Crossref] [PubMed]
3. Chen Q, Hu T, Wu L, et al. Clinical Features and Biomarkers for Early Prediction of Refractory Mycoplasma Pneumoniae Pneumonia in Children. Emerg Med Int 2024;2024:9328177. [Crossref] [PubMed]
4. Jiang Y, Liu J. Neutrophil-to-lymphocyte ratio for predicting disease severity and outcomes in children with Mycoplasma pneumoniae pneumonia: a systematic review and meta-analysis. BMC Pulm Med 2026;26:50. [Crossref] [PubMed]
5. Xu M, Li Y, Shi Y, et al. Molecular epidemiology of Mycoplasma pneumoniae pneumonia in children, Wuhan, 2020-2022. BMC Microbiol 2024;24:23. [Crossref] [PubMed]
6. Society of Clinical Microbiology of China International Exchange and Promotion Association for Medical and Healthcare. Expert consensus on the application and practice of targeted next-generation sequencing in infectious diseases. Zhonghua Yi Xue Za Zhi 2024;104:4375-3.
7. Yin Y, Zhu P, Guo Y, et al. Enhancing lower respiratory tract infection diagnosis: implementation and clinical assessment of multiplex PCR-based and hybrid capture-based targeted next-generation sequencing. EBioMedicine 2024;107:105307. [Crossref] [PubMed]
8. Deng Z, Li C, Wang Y, et al. Targeted next-generation sequencing for pulmonary infection diagnosis in patients unsuitable for bronchoalveolar lavage. Front Med (Lausanne) 2023;10:1321515. [Crossref] [PubMed]
9. Zhang P, Liu B, Zhang S, et al. Clinical application of targeted next-generation sequencing in severe pneumonia: a retrospective review. Crit Care 2024;28:225. [Crossref] [PubMed]
10. He M, Xie J, Rui P, et al. Clinical efficacy of macrolide antibiotics in mycoplasma pneumoniae pneumonia carrying a macrolide-resistant mutation in the 23 S rRNA gene in pediatric patients. BMC Infect Dis 2024;24:758. [Crossref] [PubMed]
11. Xing FF, Chiu KHY, Deng CW, et al. Post-COVID-19 Pandemic Rebound of Macrolide-Resistant Mycoplasma pneumoniae Infection: A Descriptive Study. Antibiotics (Basel) 2024;13:262. [Crossref] [PubMed]
12. Wang W, Wang L, Yin Z, et al. Correlation of DNA load, genotyping, and clinical phenotype of Mycoplasma pneumoniae infection in children. Front Pediatr 2024;12:1369431. [Crossref] [PubMed]
13. Liu J, Zhao F, Lu J, et al. High Mycoplasma pneumoniae loads and persistent long-term Mycoplasma pneumoniae DNA in lower airway associated with severity of pediatric Mycoplasma pneumoniae pneumonia. BMC Infect Dis 2019;19:1045. [Crossref] [PubMed]
14. Fang KN, Wang J, Ni JW. Correlation between Mycoplasma pneumoniae DNA replication level and disease severity in children with severe Mycoplasma pneumoniae pneumonia. Zhongguo Dang Dai Er Ke Za Zhi. 2019;21:876-0.
15. National Health Commission of the People's Republic of China, National Administration of Traditional Chinese Medicine. Guidelines for diagnosis and treatment of community-acquired pneumonia in children (2019 edition). Chin J Clin Infect Dis 2019;12:6-13.
16. Chen S, Zhou Y, Chen Y, et al. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 2018;34:i884-90. [Crossref] [PubMed]
17. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods 2012;9:357-9. [Crossref] [PubMed]
18. Wilson MR, Sample HA, Zorn KC, et al. Clinical Metagenomic Sequencing for Diagnosis of Meningitis and Encephalitis. N Engl J Med 2019;380:2327-40. [Crossref] [PubMed]
19. Gu W, Miller S, Chiu CY. Clinical Metagenomic Next-Generation Sequencing for Pathogen Detection. Annu Rev Pathol 2019;14:319-38. [Crossref] [PubMed]
20. Li L, Guo R, Zou Y, et al. Construction and Validation of a Nomogram Model to Predict the Severity of Mycoplasma pneumoniae Pneumonia in Children. J Inflamm Res 2024;17:1183-91. [Crossref] [PubMed]
21. Shen W, Sun X. Construction of a nomogram for early diagnosis of refractory Mycoplasma pneumoniae pneumonia in children. Transl Pediatr 2024;13:1119-29. [Crossref] [PubMed]
22. Zhang X, Sun R, Jia W, et al. A new dynamic nomogram for predicting the risk of severe Mycoplasma pneumoniae pneumonia in children. Sci Rep 2024;14:8260. [Crossref] [PubMed]
23. Liu F, Chen L, Wang MY, et al. Exploring high-risk factors for the prediction of severe mycoplasma pneumonia in children. Transl Pediatr 2024;13:2003-11. [Crossref] [PubMed]
24. Wang H, Xu WH, Liu JR, et al. Clinical phenotyping of severe Mycoplasma pneumoniae pneumonia in children. Zhonghua Er Ke Za Zhi 2024;62:669-5.
25. Deng F, Cao H, Liang X, et al. Analysis of cytokine levels, cytological findings, and MP-DNA level in bronchoalveolar lavage fluid of children with Mycoplasma pneumoniae pneumonia. Immun Inflamm Dis 2023;11:e849. [Crossref] [PubMed]
26. Wu J, Song W, Yan H, et al. Metagenomic next-generation sequencing in detecting pathogens in pediatric oncology patients with suspected bloodstream infections. Pediatr Res 2024;95:843-51. [Crossref] [PubMed]
27. Nie J, Zhou L, Tian W, et al. Deep insight into cytokine storm: from pathogenesis to treatment. Signal Transduct Target Ther 2025;10:112. [Crossref] [PubMed]
28. Zhang X, Sun R, Jia W, et al. Clinical Characteristics of Lung Consolidation with Mycoplasma pneumoniae Pneumonia and Risk Factors for Mycoplasma pneumoniae Necrotizing Pneumonia in Children. Infect Dis Ther 2024;13:329-43. [Crossref] [PubMed]