The protocol of this study was approved by the Ethics Committee of the Children's Hospital of Soochow University (ethical approval no. 2023CS116) and was in line with the Declaration of Helsinki (as revised in 2013).

From January 1, 2016, to December 31, 2023, our surveillance study retrospectively enrolled a total of 113,625 patients with respiratory infections from three hospitals in Suzhou, China: the Children’s Hospital of Soochow University, Children’s Hospital of Wujiang District, and Affiliated Suzhou Hospital of Nanjing University Medical School. Among them, there were a total of 82,427 pediatric cases from the Children’s Hospital of Soochow University.

In this study, 200 children with M. pneumoniae from the Children’s Hospital of Soochow University were enrolled between June and December 2023. Notably, this testing was exclusive to this hospital and not extended to other participating hospitals. These patients, part of the respiratory infection surveillance from 2016 to 2023, underwent P1 gene typing, multi-site single nucleotide polymorphism (SNP) genotyping, and macrolide resistance testing on BALF samples. For each, we documented general information, clinical signs, BALF characteristics, and radiological findings. Inclusion criteria were as follows: The inclusion criteria were: (I) meeting the diagnostic criteria for CAP (Harris et al., 2011; Rose et al., 2020); (II) selection of pediatric patients under 18 years of age; (III) obtaining BALF by bronchoscopy in eligible patients; (IV) BALF testing positive for M. pneumoniae DNA. Exclusion criteria included: (I) presence of tuberculosis, interstitial lung disease, pulmonary eosinophilic infiltration, or pulmonary vasculitis; (II) incomplete clinical data.

The collected respiratory specimens were subjected to nucleic acid (DNA) testing for M. pneumoniae (Thermo Scientific™ KingFisher™ Flex Magnetic Particle Processors, Thermo Fisher). All atypical pathogens were tested with the use of the commercial real-time PCR based kits (Multiplex Combined Real-time PCR Detection Kit for Respiratory Viruses, Multiplex Combined Real-time PCR Detection Kit for Respiratory Bacteria, Jiangsu Uninovo Biological Technology Co. Ltd., China).

P1 gene sequences from M. pneumoniae–positive specimens were typed using nested PCR and restriction fragment length polymorphism (RFLP), as previously described. The RepMP2/3 and RepMP4 repetitive regions were amplified concurrently, and the PCR products from type 2 specimens were sequenced to identify type 2 variants (Hung et al., 2021; Pereyre et al., 2007).

A novel SNP genotyping method for M. pneumoniae was developed utilizing matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) (Zhao et al., 2021). The detection limit of this method for nucleic acids was 10²–10³ copies/reaction. Six SNP site-based genotypings were simultaneously detectable using a multiplex PCR and mass probe assay, as detailed in Supplementary Table S1.

Mutations associated with macrolide resistance, specifically at positions 2063, 2064, and 2,617 in the 23S rRNA gene of M. pneumoniae, were identified as previously reported (Lin et al., 2010).

Mycoplasma pneumoniae pneumonia exhibits varying degrees of severity and is categorized into General M. pneumoniae Pneumonia (GMPP) for milder cases and Severe M. pneumoniae Pneumonia (SMPP) for more severe presentations (Zhang et al., 2022). Diagnosis of MPP was conducted following the Chinese Medical Association guidelines for management of CAP in children: (1) fever (>37.3°C) or acute respiratory symptoms or both; (2) decreased breath sounds, dry or wet rales; (3) chest radiograph showed at least one of the following: spotted or patchy infiltration; interstitial changes; lobar parenchymal infiltration shadow; enlargement of hilar lymph nodes; (4) positive PCR results or ≥4-fold increase in MP antibody titer. SMPP was defined as MPP with one of the following: (1) poor general condition; (2) increased breathing rate; (3) cyanosis and dyspnea; (4) multilobed or ≥2/3 infiltration of the lung; (5) transcutaneous oxygen saturation ≤92% in room air; (6) extrapulmonary complications.

Based on local interventions in Suzhou, the COVID-19 pandemic was categorized into four phases. In Phase I (March to May 2020), extensive NPIs such as school closures and mobility restrictions were implemented. During Phase II (June to August 2020), schools initiated a gradual reopening, prioritizing graduating classes, and kindergartens also partially resumed operations. The III phase (September 2020 to November 2022) featured the “Dynamic Zero-COVID” strategy, with widespread SARS-CoV-2 testing and responsive NPIs following any detected local transmission. Schools and kindergartens remained fully operational throughout this period. Finally, during Phase IV (January to November 2023), all NPIs were lifted.

Continuous variables are presented as median (interquartile range, IQR), while categorical variables are presented as frequencies (percentages). Comparisons among different groups were made using the chi-square (χ²) test or Fisher’s exact test. Phylogenetic trees were constructed using R language and specific software packages. A two-sided p-value below 0.05 was considered statistically significant. Data were analyzed using SPSS 23.0 (released in 2015, SPSS Inc., Chicago, IL, United States).

In this study, the M. pneumoniae epidemic season was defined as consecutive months with a PCR positivity rate for M. pneumoniae exceeding 14.89%. Prior to the pandemic, M. pneumoniae epidemics typically began increasing in May and June, reaching a peak in August, and lasting about 7 months. In Phases I–III of the pandemic, the positive detection rate for M. pneumoniae was low. Nonetheless, in Phase IV, post-pandemic, the incidence of M. pneumoniae rebounded to endemic levels, featuring an epidemic season similar in timing but with a notably higher positive detection rate (Figures 1B–D).

This study identified 2019 and 2023 as the peak years for M. pneumoniae positive detection rates. The age composition ratio in 2023 revealed a significantly higher proportion of children aged 6 and older compared with 2019. In 2019, 42.3% of cases involved children aged 6 years and older; by 2023, this proportion had significantly increased to 55.0%, indicating a statistically significant shift in age distribution (χ² = 36.095, p < 0.001) (Figure 2).

Plastic bronchitis (PB) is a significant indicator of severe and refractory M. pneumoniae pneumonia. From 2018 to 2023, the annual counts of PB cases were 65, 74, 8, 33, 31, and 211, respectively, showing a marked increase in severity related to M. pneumoniae infection in 2023 (Figure 3A,B). PB on imaging often manifests as pulmonary consolidation and atelectasis (Figure 3C). The pathology of M. pneumoniae pneumonia frequently shows fibrinoid inflammatory changes (Figure 3D).

In this study, 200 M. pneumoniae-positive cases from 2023 were randomly selected for P1 and SNP genotype analysis. Of the 200 samples, 8 were either suspected to have a low M. pneumoniae pathogen load. The remaining 192 samples were classified into 11 types based on the reference typing scheme, with the most common being SNP-27 (36.5%), SNP-0 (21.4%), SNP-11 (18.8%), and SNP-34 (17.7%). Further details are provided in Supplementary Table S2.

Among the 156 qualified BALF samples, 81.4% were P1 genotype and 18.6% were P2 genotype. Patients with severe M. pneumoniae pneumonia who had the P1 genotype were significantly more common than those with the P2 genotype (p = 0.021). There was no significant difference in the SNP genotype distribution between patients with GMPP and SMPP (p > 0.05) (Table 1).

Of the 192 qualified BALF specimens analyzed for macrolide resistance mutations, the total mutation rate was 97.3%. The mutation rate for 23S rRNA at position 2063 was 96.17%, with A2063G being the most common mutation. Two samples had co-occurring mutations of A2063G and C2617T. The mutation rate at position 2064 in the 23S rRNA resistance gene was 1.6%. The mutation rate at the 2,617 site of the 23S rRNA was 1.0%.

For the construction of the phylogenetic tree with the 156 M. pneumoniae positive cases, they were classified into P1 and P2 types based on P1 genotyping. The results indicated that the P1 type was predominant, with a significantly higher positive rate of macrolide resistance genes (p = 0.021) (Figures 4A,C). Cluster analysis of the 192 M. pneumoniae positive cases revealed no significant difference in macrolide resistance among different SNP types (Figure 4B).

Mycoplasma pneumoniae, a ubiquitous pathogen with a global distribution, is responsible for 10–30% of CAP cases in children. Infections can lead to regional outbreaks every 3–5 years, lasting 1–2 years (Yamazaki and Kenri, 2016). Since the onset of the COVID-19 pandemic in Suzhou, the transmission trajectory of M. pneumoniae has been unusual, differing from previous epidemic patterns.

Our survey reveals that M. pneumoniae infections peaked in 2019 and again in 2023 within the period from 2016 to 2023. The incidence of M. pneumoniae pneumonia in 2023 significantly exceeds that of 2019, indicating a major outbreak of concern. Throughout Phases I–III of NPIs, M. pneumoniae positive cases were consistently below 200 per month, with a positivity rate lower than the average of 14.3%. The surge might be due to interventions such as mandatory mask-wearing and social distancing, or to antigenic shifts and increased population susceptibility (Chow et al., 2023; Liu B. et al., 2024). Herd immunity waned over the three-year NPIs period, due to inadequate exposure to the pathogen. Consequently, a weakened immune system struggles to combat M. pneumoniae infections following the relaxation of epidemic controls. This phenomenon is not unique to M. pneumoniae; it is also observed with other respiratory pathogens, such as influenza and respiratory syncytial virus (Jiang et al., 2024).

Mycoplasma pneumoniae infections show a clear seasonal pattern. This study found that before the epidemic, peak infection rates were in summer, followed by autumn, while the lowest rates were in spring and winter (Tian et al., 2017; Wang et al., 2022). This finding is corroborated by earlier studies. Post-epidemic, autumn experienced the highest infection rates, with summer in the second place. Studies have established that M. pneumoniae optimally grows at 36–37°C, with infection rates rising as minimum temperatures increase. The warmer conditions of summer and autumn facilitate M. pneumoniae replication (Meyer Sauteur et al., 2022a,b). During the NPIs, strict control measures consistently kept M. pneumoniae positivity rates low, masking seasonal patterns. However, as these measures were relaxed, the seasonality of M. pneumoniae infections became more apparent. Related to the immune gap, the peak season’s outbreak appears delayed. The delayed reemergence of M. pneumoniae is likely due to waning herd immunity, as indicated by epidemic patterns. This delay is similar to the pattern observed with respiratory syncytial virus (RSV) (Li M. et al., 2024).

The positive detection rate of M. pneumoniae varied among different age groups in this study, with the school-age group having the highest rate, while the preschool, infant, and young children groups had the lowest. This finding is consistent with previous studies (Jiang et al., 2023). The elevated M. pneumoniae detection rate among school-age children could stem from decreased social distancing and heightened cross-infection following the relaxation of post-epidemic (Phase IV) control measures.

Further analysis of the relationship between M. pneumoniae positive detection rates and age across various phases showed that susceptibility to M. pneumoniae was lower in all age groups during the NPIs phase. Conversely, there was a significant increase in the percentage of positive M. pneumoniae detections in the age group above 6 years post-outbreak, and this difference was statistically significant.

This consideration is relevant to the prolonged implementation of NPIs, spanning up to 3 years. During the NPIs implementation, preschool and school-age children faced chronic masking, home isolation, reduced pathogen exposure, and weakened herd immunity. The phenomenon of mass infections following an outbreak is known as “immunization debt.” Similar observations regarding RSV infections were reported by Chinese scholar Jiang, who noted a significant shift in the age distribution of RSV infections in the post-epidemic period (Jiang et al., 2024).

The PB is a relatively rare respiratory disease in children, characterized by the formation of gelatinous or firm bronchial casts within the airways. These casts can rapidly move along the bronchial tree, obstructing the airways and causing clinical symptoms such as shortness of breath and wheezing, and can be life-threatening in severe cases. PB is a critical complication of CAP, indicating severe disease. In our study, we quantified cases of PB diagnosed via bronchoscopy in patients testing positive for M. pneumoniae between 2018 and 2023.A notable upsurge in PB cases was observed in 2023.The pneumonia caused by M. pneumoniae infection in 2023 not only showed a significant rise in case numbers but also increased in severity, resulting in a considerable societal burden. The sudden increase in severe cases could be attributed to several factors. Initially, the risk of future epidemics increases as the proportion of susceptible individuals grows due to prolonged minimal exposure to pathogens. Notably, children born during the epidemic are especially impacted by the issue of immunization debt (Cohen et al., 2021). Our recent research indicates that prolonged minimal exposure to M. pneumoniae can weaken antibodies, highlighting an immune gap with attenuated humoral immunity as a key driver of increased M. pneumoniae susceptibility.

Secondly, macrolide antibiotics, commonly used to treat M. pneumoniae, have led to pathogen resistance due to overuse. Studies show that the resistance rate of M. pneumoniae to macrolides in China was between 70 and 90% before NPIs (Xue et al., 2018). This study revealed that the drug resistance rate of M. pneumoniae soared to 97.3% post-NPIs. The exceedingly high resistance rate to macrolides presents significant clinical challenges, underscoring the urgent need for alternative antibiotics to treat M. pneumoniae infections.

Mycoplasma pneumoniae is divided into two main genotypes, P1 type 1 and P1 type 2, which are differentiated by nucleotide variations in the repetitive elements RepMP2/3 and RepMP4 of the MPN141 gene, encoding the P1 adhesin protein (Cousin-Allery et al., 2000). This study found that the P1 type 1 genotype was predominant in 2023. P1 type 1 showed a higher prevalence of severe infections, positive drug-resistant genes, and pulmonary complications than P1 type 2. This finding is consistent with previous research (Rodman et al., 2021). Consider genotypes as contributing factors to the prevalence of M. pneumoniae infections.

The increased accessibility of markers through high-throughput whole genome sequencing (WGS) has demonstrated SNP analysis’s advantages in assessing drug resistance, evolution, and molecular epidemiology (Lipworth et al., 2019). However, few clinical studies have investigated the influence of the M. pneumoniae SNP genotype to date. For genotype I, M. pneumoniae strains were predominantly SNP27, SNP11, and SNP15 types; for genotype II, SNP 0 was predominant in this study (Supplementary Table S1). Compared to the data from 2017 to 2018, there was a significant change in the genotype; SNP 14 was predominant in Suzhou during that period (Zhao et al., 2019).

This study did not comprehensively evaluate all environmental and behavioral factors that might influence the epidemiological patterns of M. pneumoniae, and thus has some limitations. Future studies should consider incorporating a broader spectrum of environmental and behavioral factors and examine their relationship with NPIs to thoroughly assess the epidemiological patterns of M. pneumoniae.