Comprehensive study of community acquired Mycoplasma pneumoniae pneumonia in children in Baoding, China, 2023

Abstract

BACKGROUND

Mycoplasma pneumoniae (M. pneumoniae) is considered to be one of the causative agents of community acquired pneumonia in children with general or severe course of disease. Severe M. pneumoniae pneumonia (SMPP) has emerged as a crucial global health concern due to high mortality rate in children under 5 years, potentially life-threatening complications, and growing challenges in pediatric treatment associated with rising macrolide resistance. Additionally, MPP can be complicated by other bacterial and/or viral pathogens, which may exacerbate disease severity. After the lifting of strict non-pharmaceutical interventions (NPIs) worldwide, the dramatic rise of incidence of MPP in Asia and Europe was observed.

AIM

To perform the comprehensive study of community acquired MPP cases registered in 2023 in Baoding Hospital, China.

METHODS

A total of 1160 children from 1 month to 15 years old with confirmed MPP diagnosis were enrolled in the study. The blood and respiratory samples were collected within the 24 hours after admission. The hematological parameters, biochemical markers, cytokine profiles were assessed. The respiratory samples were tested for the presence of M. pneumoniae and other 23 bacterial/viral pathogens by multiplex polymerase chain reaction (PCR). The macrolide resistance mutations (A2063G, A2064G in the 23S rRNA gene of M. pneumoniae) were determined by PCR.

RESULTS

Number of MPP cases has dramatically increased starting August with peak in November. SMPP and general MPP (GMPP) were identified in 264 and 896 of 1160 hospitalized children. The binary logistic regression analysis identified six [C-reactive protein (CRP), lactate dehydrogenase, procalcitonin, erythrocyte sedimentation rate, fibrin and fibrinogen degradation products (FDPs), D-dimer] and four (neutrophils, CRP, FDPs, prothrombin time) predictors of SMPP in age groups 2-5 years and 6-15 years, respectively. Children with SMPP showed significantly higher levels of cytokine interleukin (IL)-17F (2-5 years), and cytokines interferon-gamma, tumor necrosis factor-alpha, IL-10 (6-13 years). Concomitant viral/bacterial pathogens were determined in 24.3% and 28.0% cases of SMPP and GMPP. Among them, Streptococcus pneumoniae (S. pneumoniae) and Haemophilus influenzae (H. influenzae) were predominant. 93.2% cases of MPP were associated with macrolide resistant M. pneumoniae.

CONCLUSION

Specific MPP epidemiological pattern associated with lifting NPIs was revealed: Increase of hospitalized cases, prevalence of S. pneumoniae and H. influenzae among concomitant pathogens, 93.2% of macrolide resistant M. pneumonia.

Key Words: Community acquired Mycoplasma pneumoniae pneumonia; Children; Macrolide resistance; A2063G or A2064G mutations

Core Tip: After the lifting of strict non-pharmaceutical interventions (NPIs), the dramatic rise of Mycoplasma pneumoniae pneumonia (MPP) incidence in Europe and Asia was observed. Comprehensive study of community acquired cases of MPP in 2023 in Baoding Hospital, China, was performed. The dramatic increase in the incidence of MPP cases, that was delayed for nine months after lifting NPIs, and unprecedented 93.2% prevalence of macrolide resistant Mycoplasma pneumoniae causing pneumonia in children were revealed. We hypothesized that strict NPIs may have created the “bottleneck” selecting the most successful bacterial clone(s) that subsequently spread, and it may represent a “founder effect”.

INTRODUCTION

Mycoplasma pneumoniae (M. pneumonia) is considered to be one of the causative agents of community acquired pneumonia (CAP), especially in children[1,2]. Another pathogen, Streptococcus pneumoniae (S. pneumoniae), is also common cause of CAP in children resulting in both mild and severe manifestations[3]. Currently, due to the effectiveness of 13-valent vaccine against S. pneumoniae worldwide, M. pneumoniae is becoming the leading pathogen responsible for 10%-40% of CAP cases in children depending on the regions and seasoning[4,5]. M. pneumoniae can lead to epidemic outbreaks every 2-3 years in Europe and Asia[6-9]. In particular, about 37.5% of CAP cases in children in Northern China were caused by M. pneumoniae with periodic outbreaks of M. pneumoniae pneumonia (MPP) every 2-3 years[9].

Depending on clinical symptoms and severity of disease, the M. pneumoniae pneumonia is diagnosed as general MPP (GMPP) or severe MPP (SMPP). It is supposed that SMPP in children can be associated with bacterial [S. pneumoniae, Pseudomonas aeruginosa (P. aeruginosa), Staphylococcus aureus (S. aureus), etc.] and/or viral (adenovirus, coronavirus, influenza virus, etc.) co-infections[10-12]. Although the macrolides are the recommended first-line antibiotics for treatment of children with MPP, it is often not effective due to the increasing number of macrolide resistant M. pneumoniae isolates[13]. Mainly, macrolide resistance of M. pneumoniae is result of point mutations, A2063G and A2064G, in 23S rRNA gene, but recently in vitro selection and genome analysis identified potentially novel mechanisms of M. pneumoniae macrolide resistance, partly associated with new point mutation A2067C in 23S rRNA gene and few mutations in ribosomal protein L4[14].

Recently, numerous data on clinical manifestations, prevalence, diagnostic approaches, treatment of MPP in children in China have been published[7,15-20]. Importantly, during the coronavirus disease 2019 (COVID-19) pandemic, the number of MPP as well as different respiratory bacterial/viral infections decreased significantly due to the strict non-pharmaceutical interventions (NPIs) worldwide[21-24]. However, in the late of 2022, after the lifting these measures, the delayed rise of incidence of MPP in Europe and Asia was observed[8]. The goal of this study was to conduct a comprehensive analysis of community acquired M. pneumoniae pneumonia cases registered in 2023 in Baoding Hospital, China, after the lifting of the strict NPIs.

MATERIALS AND METHODS

Study population

This study was done at the Baoding Hospital of Beijing Children’s Hospital, Capital Medical University, China. In 2023, a total of 4256 children ranging in age from 1 month to 16 years were admitted respiratory disease department for 1–21 days, and CAP was diagnosed during hospitalization. For 2680 of 4256 cases, MPP was confirmed by polymerase chain reaction (PCR) analysis. After excluding cases with incomplete clinical data and with immunodeficiency diseases, a total of 1160 children (544 male and 616 female) ranging in age from 1 month to 15 years [median age 7 years, interquartile range (IQR) 6-9] with confirmed MPP diagnosis were included in this study. The MPP was established based on clinical symptoms, laboratory parameters, and PCR detection of M. pneumoniae in clinical samples. According to the National Health Commission’s “Guidelines for the Diagnosis and Treatment of M. pneumonia in Children” (China NHCotPsRo, 2019), the severity of MPP (severe or general cases) was defined as the presence of one or more of the following manifestations: (1) Radiography: Infiltration of 2/3 of one lung, multilobar infiltration, pleural effusion, pneumothorax, atelectasis, lung necrosis or lung abscesses; (2) Hypoxemia: Cyanosis; marked increase in respiratory rate; marked chest wall retractions, tracheal tugging or nasal flaring; O₂ saturation less than 92%; (3) Extrapulmonary complications; (4) Persistent high fever for more than 5 days; and (5) Reluctance or inability to feed[25].

Treatment

All the patients were treated according to the China NHCotPsRo, 2019[25].

Clinical samples

The blood, sputum, endotracheal aspirates, bronchoalveolar lavage (BAL) and oropharyngeal swabs were collected from the children within the 24 hours after admission. Sputum, endotracheal aspirates, and BAL were analyzed for M. pneumoniae detection, identification of other bacterial/viral pathogens, and macrolide resistance genotyping of M. pneumoniae. The blood samples were used for clinical and biochemical assays.

Laboratory testing

The hematological and biochemical parameters of the blood samples were analyzed in the Laboratory Department of Baoding Hospital. Hematological parameters including white blood cell (WBC) count, percentages of neutrophils, lymphocytes and eosinophils, platelet count, and C-reactive protein (CRP) levels were measured using a fully automated hematology analyzer (Mindray 7500CS, Shenzhen, China). Biochemical parameters including alanine aminotransferase, aspartate aminotransferase, creatine kinase (CK), lactate dehydrogenase (LDH), and CK-MB were measured using a fully automated biochemistry analyzer (Beckman AU5811, California, United States). The erythrocyte sedimentation rate (ESR) was measured using a fully automated hematology analyzer (GBO-SUNS SRS100; Beijing, China) based on Westergren's method. Procalcitonin (PCT) concentrations were determined by automated chemiluminescent immunoassay (AutoLumo A2000 PLUS, Zhengzhou, China). Coagulation parameters including prothrombin time (PT), international normalized ratio, activated partial thromboplastin time (aPTT), fibrinogen, fibrinogen degradation products (FDPs), D-dimer, and antithrombin III (AT III) were measured using a fully automated coagulation analyzer (ACL TOP 750, Werfen, Barcelona, Spain). Serum cytokines levels [interleukin (IL)-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL- 10, IL-12p70, IL-17A, IL-17F, IL-22, interferon (IFN)-gamma, tumor necrosis factor (TNF) α, and TNF-β] were quantified using a 14-cytokines immunoassay panel (ID 914002 Quantobio, Beijing, China) analyzed by flow cytometry (Mindray BriCyte E6, Shenzhen, China).

PCR based identification of M. pneumonia, other bacterial/viral pathogens, and 23S rRNA genotyping for macrolide resistance

Simultaneous detection of M. pneumoniae and macrolide resistance-associated mutation sites (23S rRNA, A2063G and A2064C) was performed using a commercial multiplex PCR assay (M. pneumoniae and Macrolides-Resistant Strain Nucleic Acid Test Kit, Mole Bioscience, Jiangsu, China) according to the manufacturer’s protocol. The Respiratory Pathogen Multiplex Kit (Health Gene Technologies, Ningbo, China) was used to detect respiratory viral pathogens according to the recommended protocol. Multiplex fluorescent PCR assays were performed to detect bacterial pathogens using the Respiratory Pathogens Nucleic Acid Test Kit in accordance with the manufacturer’s instructions (Sansure Biotech, Changsha, China). The minimum detection limit for each pathogen is given in brackets, if available in the manufacturer’s instructions. A total of 23 pathogens were tested: 11 viruses [respiratory syncytial virus (RSV) 0.4 TCID₅₀/mL], human rhinovirus (HRV, 0.15 TCID₅₀/mL), human bocavirus (HBoV, 5000 copies/mL), human metapneumovirus (HMPV, 0.35 TCID₅₀/mL), human adenovirus (ADV, 2000 copies/mL or 1 TCID₅₀/mL), human parainfluenza virus (HPIV, 0.4 TCID₅₀/mL), influenza viruses A (IFA, 0.098 TCID₅₀/mL), B (IFB, 2.0 TCID₅₀/mL), H3N2 (0.1 TCID₅₀/mL), and H1N1 (0.098 TCID₅₀/mL), human coronavirus (HCoV, 0.3 TCID₅₀/mL), and 12 bacteria [S. pneumoniae, Acinetobacter baumannii (A. baumannii), S. aureus, Enterobacter cloacae (E. cloacae), P. aeruginosa, methicillin resistant S. aureus (MRSA), Klebsiella pneumoniae (K. pneumoniae), Haemophilus influenzae (H. influenzae), Escherichia coli (E. coli), Pseudomonas maltophilia (P. maltophilia), Mycobacterium tuberculosis (M. tuberculosis) , and Legionella pneumophila (L. pneumophila)].

Data management and statistical analysis

All the data including clinical and laboratory results were collected from medical and laboratory records and saved in a standardized database. Statistical software SPSS27.0 was used for data processing and analysis. Continuous variables (age, days of hospitalization, hematological and biochemical parameters) are presented as median (IQR); Mann-Whitney U test was used for two-group comparisons. Categorial variables are presented as numbers (percentages); the χ² test was used for two-group comparisons or the Fisher’s exact test (two-tailed) was applied when expected frequences were less than 5. Binary logistic regression was employed to identify factors associated with disease severity, followed by receiver operating characteristic curve analysis of significant factors to develop a predictive model for SMMP. The data with a P value < 0.05 were considered to be statistically significant.

RESULTS

Incidence of severe MPP cases

Severe and GMPP were identified in 264 (22.8%, 130 male/134 female) and 896 (77.2%, 414 male/482 female) of 1160 cases, respectively. Seasonal analysis of MPP cases in 2023 revealed specific epidemiological pattern (Figure 1). Following the lifting of the strict NPIs in December of 2022, the number of MPP cases was low during January-July, and 3–31 hospitalized MPP cases were registered monthly in that period. Since August, the number of registered MPP cases dramatically increased, starting from 48 cases in August till 288 cases in December with a peak of 491 cases in November.

Figure 1 Incidence of severe Mycoplasma pneumoniae pneumonia cases registered in Baoding hospital in 2023. SMMP: Severe Mycoplasma pneumoniae pneumonia.

The number of SMPP cases was relatively low during January-September, and it demonstrated a significant increase starting October (Figure 1). Because of the low numbers of both MPP and SMPP in each month of January-July, the rates of severe cases cannot be considered reliable. Starting August, the rates of severe cases of MPP became stable and were accounted around 20% (16.7%-25.6%) (Figure 1).

All 1160 children were divided into three age groups: 0-1 year, 2-5 years, and 6-15 years. Majority of the children belonged to the 6-15 years age group (896/1160, 77.2%), while the 0-1 year and 2-5 years groups comprised 51 (4.4%) and 213 (18.4%) children, respectively. SMPP was diagnosed in 13/51 (25.5%), 49/213 (23.0%), and 202/896 (22.5%) children of the following age groups 0-1 year; 2-5 years; 6-15 years, respectively. No statistical differences in age or sex were estimated between SMPP and GMPP cases.

Among children aged 2-5 and 6-15 years, the significant differences in duration of hospitalization period were observed depending on MPP severity. Children with GMPP (both age groups) showed shorter median hospitalization (8 days, IQR: 7-9) than those with SMPP (2-5 years: 10 days, IQR: 8-13; 6-15 years: 9 days, IQR: 7-12; P < 0.001). However, children 0-1 year showed no statistically significant difference in median hospitalization duration between GMPP (9 days, IQR: 7-11) and SMPP (9 days, IQR: 7-17; P = 0.391).

Clinical laboratory parameters

Hematological and biochemical parameters of blood samples from children aged 2-5 years and 6-13 years were analyzed separately to take into account the age-dependent variations in total and differential WBC counts and immune system parameters that might affect MPP clinical outcomes. These two groups represented the majority (1097/1160; 94.6%) of children included in this study. Due to limited sample sizes, infants aged 1-12 months (n = 18) and 1 year-old (n = 33), and adolescents aged 14-15 years (n = 12) were excluded from clinical parameter analyses.

For all the children the moderate leukocytosis was observed; however, statistically significant differences between SMPP and GMPP cases were found only in the 6–13-year age group (Table 1). In this group, SMPP cases exhibited significantly higher leukocyte levels compared to GMPP cases (P = 0.022). Conversely, children aged 2–5 years showed an opposite trend in leukocyte dynamics, with lower levels in SMPP cases than in GMPP cases, although these differences were not statistically significant (P = 0.792). Meanwhile, the percentage of neutrophils was significantly higher in SMPP cases in both age groups, but the percentages of lymphocytes and eosinophils were lower in SMPP cases compared to GMPP cases, although these differences were statistically significant only in the 6-13 years age group.

Table 1 Routine hematological and biochemical parameters in blood samples from children (2–13 years) with Mycoplasma pneumoniae pneumonia, median (interquartile range).

Laboratory parameters	SMPP cases (n = 245)	GMPP cases (n = 852)	P value
Children 2–5 years old (SMPP 49 cases vs GMPP 164 cases)
Leukocytes (109/L)	7.83 (6.37, 9.26)	8.11 (6.24, 9.82)	0.792
Neutrophils (%)	62.1 (52.7, 67.7)	56.7 (48.3, 64.2)	0.039
Lymphocytes (%)	28.1 (24.2, 37.6)	32.8 (25.8, 40.7)	0.073
Eosinophils (%)	0.6 (0.3, 2.8)	1.2 (0.4, 2.8)	0.248
Platelet count (109/L)	279 (218, 353)	299 (238, 377)	0.197
C-reactive protein (mg/L)	11.77 (4.68, 26.47)	7.846 (2.63, 17.21)	0.041
ALT (U/L)	15 (10, 28)	12 (10, 21)	0.072
AST (U/L)	32 (26, 42)	29 (24, 36)	0.066
CK (U/L)	56 (36, 127)	65 (45, 89)	0.956
CKMB (ng/mL)	2.92 (2.05, 3.65)	2.60 (2.03, 3.19)	0.233
LDH (U/L)	354 (294, 477)	288(256, 343)	< 0.001
Procalcitonin (ng/mL)	0.23 (0.10, 0.51)	0.12 (0.08, 0.26)	< 0.001
ESR (mm/h)	26 (17, 45)	30 (20, 50)	0.262
Prothrombin time (s)	12.5 (11.7, 13.0)	12.2 (11.6, 13.0)	0.47
INR	1.12 (1.05, 1.16)	1.11 (1.04, 1.17)	0.984
aPTT (s)	31.5 (28.9, 34.5)	33.0(30.4, 35.5)	0.088
Fibrinogen (g/L)	3.40 (3.16, 3.86)	3.68 (3.28, 4.10)	0.075
FDPs (mg/L)	2.2 (1.3, 4.1)	1.4 (0.9, 2.3)	0.008
D-dimer (mg/L)	0.389 (0.211, 0.711)	0.215 (0.142, 0.389)	< 0.001
Plasma antithrombin III (%)	110 (96, 119)	113 (103, 120)	0.271
Children 6–13 years old (SMPP 196 cases vs GMPP 688 cases)
Leukocytes (109/L)	8.47 (6.54, 10.92)	7.96 (6.28, 9.88)	0.022
Neutrophils (%)	68.1 (60.1, 76.4)	61.8 (53.5, 69.0)	< 0.001
Lymphocytes (%)	21.4 (15.9, 29.2)	27.7 (21.3, 34.9)	< 0.001
Eosinophils (%)	0.8 (0.2, 2.3)	1.1 (0.4, 2.8)	0.034
Platelet count (109/L)	272 (218, 335)	281 (231, 359)	0.056
C-reactive protein (mg/L)	18.59 (9.40, 40.09)	9.19 (3.47, 19.32)	< 0.001
ALT (U/L)	19 (13, 31)	14 (11, 22)	< 0.001
AST (U/L)	30 (20, 47)	25 (20, 30)	< 0.001
CK (U/L)	70 (43, 138)	66 (46, 97)	0.180
CKMB (ng/mL)	2.38 (1.81, 3.13)	2.33 (1.85, 3.03)	0.830
LDH (U/L)	321 (255, 439)	265 (232, 309)	< 0.001
Procalcitonin (ng/mL)	0.15 (0.09, 0.33)	0.10 (0.06, 0.16)	< 0.001
ESR (mm/h)	32 (20, 50)	33 (20, 50)	0.910
Prothrombin time (s)	13.1 (12.3, 14.0)	12.6 (12.0, 13.4)	< 0.001
INR	1.18 (1.11, 1.26)	1.14 (1.09, 1.21)	< 0.001
aPTT (s)	31.8 (29.3, 34.1)	32.7 (30.7, 35.1)	< 0.001
Fibrinogen (g/L)	4.00 (3.45, 4.61)	3.93 (3.53, 4.35)	0.137
FDPs (mg/L)	3.2 (1.7, 6.4)	1.5 (1.0, 2.5)	< 0.001
D-dimer (mg/L)	0.512 (0.226, 1.114)	0.236 (0.150, 0.376)	< 0.001
Plasma antithrombin III (%)	103 (96, 111)	106 (98, 114)	0.022

ALT: Alanine aminotransferase; aPTT: Activated partial thromboplastin time; AST: Aspartate transaminase; CK: Creatine kinase; CKMB: Creatine kinase-MB; ESR: Erythrocyte sedimentation rate; GMPP: General Mycoplasma pneumoniae pneumonia; INR: International standardized ratio; LDH: Lactate dehydrogenase; SMPP: Severe Mycoplasma pneumoniae pneumonia; FDPs: Fibrin and fibrinogen degradation products.

Majority of clinical laboratory parameters were elevated in children with SMPP compared to GMPP in both groups, while key coagulation markers including platelet count, aPTT, and plasma AT III were lower in children with SMPP (Table 1). Among the parameters analyzed, only five parameters (CRP, LDH, PCT, FDPs and D-dimer) showed statistically significant increase in SMPP cases compared to GMPP cases in both age groups. Notably, the majority of the studied parameters were statistically significant only in the 6-13 years age group.

The binary logistic regression analysis identified six predictors for SMPP in children aged 2-5 years and four predictors for SMPP in children aged 6-13 years, with only two parameters, CRP and FDRs, being common for both age groups (Figure 2 and Table 2). Notably, all four predictors for SMPP in children aged 6-13 years including CRP demonstrated statistically significant differences between SMPP and GMPP cases (Table 2). In contrast, among the six parameters included in multivariate predictive model with highest area under the curve = 0.812 for SMPP in children aged 2-5 years, two of them, CRP and ESR, showed no significant univariate associations with SMPP (Table 2).

Figure 2 Receiver operator characteristic curves of the logistic regression model for severe Mycoplasma pneumonia pneumonia prediction in children. A: Aged 2-5 years; B: Aged 6 -13 years old. AUC: Area under the curve; CRP: C-reactive protein; ESR: Erythrocyte sedimentation rate; FDPs: Fibrin and fibrinogen degradation products; LDH: Lactate dehydrogenase; NEUT: Neutrophils; PCT: Procalcitonin.

Table 2 Predictive biomarkers for severe Mycoplasma pneumonia pneumonia identified by binary logistic regression analysis.

Independent factors	AUC (95%CI)	Cut-off value	Sensitivity (%)	Specificity (%)	P value
Children 2-5 years old
CRP	0.574 (0.462-0.685)	20.28	37.8	0.833	0.179
LDH	0.713 (0.617-0.809)	328.5	64.9	0.728	< 0.001
PCT	0.623 (0.513-0.732)	0.175	0.568	0.693	0.025
ESR	0.454 (0.351-0.557)	14.0	0.892	0.158	0.4
FDPs	0.611 (0.504-0.719)	2.19	0.541	0.719	0.042
D-Dimer	0.654 (0.556-0.752)	0.232	0.730	0.553	0.005
CRP + LDH + PCT + ESR + FDPs + D-Dimer	0.812 (0.740-0.885)	-	0.919	0.553	< 0.001
Children 6-13 years old
NEUT	0.661 (0.616-0.706)	67.95	0.551	0.722	< 0.001
CRP	0.680 (0.634-0.725)	13.315	0.672	0.634	< 0.001
FDPs	0.731 (0.687-0.775)	2.195	0.651	0.703	< 0.001
Prothrombin time	0.592 (0.544-0.640)	13.05	0.511	0.631	< 0.001
NEUT + CRP + FDPs + prothrombin time	0.749 (0.707-0.792)	-	0.618	0.785	< 0.001

CRP: C reactive protein; ESR: Erythrocyte sedimentation rate; FDPs: Fibrin and fibrinogen degradation products; LDH: Lactate dehydrogenase; NEUT: Neutrophils; PCT: Procalcitonin; AUC: Area under the curve.

Analysis of 14 cytokines revealed a statistically significant increase in only proinflammatory cytokine IL-17F in children aged 2-5 years with SMPP compared to those with GMPP (Table 3). In contrast, children aged 6-13 years with SMPP showed significantly higher levels of two proinflammatory cytokines (IFN-γ and TNF-α) and anti-inflammatory cytokine IL-10 compared to GMPP (Table 3). Level of IL-6 in this age group was approximately twice higher in SMPP cases compared to GMPP cases although not statistically significant (P = 0.056). However, none of these cytokines were included into the final predictive model for SMPP in both age groups based on binary logistic regression analysis.

Table 3 Serum cytokine levels in children (2–13 years) with Mycoplasma pneumoniae pneumonia, median (interquartile range).

Cytokines (pg/mL)	SMPP cases (n = 73)	GMPP cases (n = 103)	P value
Children 2-5 years old (SMPP 12 cases vs GMPP 23 cases)
IL-1β	1.18 (1.03, 1.27)	1.16 (0.81, 1.55)	0.897
IL-2	0.43 (0.12, 0.66)	0.77 (0.05, 1.04)	0.340
IL-4	0.92 (0.78, 1.37)	1.05 (0.81, 1.31)	0.715
IL-5	1.31 (0.51, 2.01)	1.58 (1.21, 2.09)	0.175
IL-6	52.62 (6.32, 69.84)	36.75(16.98, 64.12)	0.889
IL-8	11.25 (2.08, 76.07)	15.76 (9.37, 27.94)	0.424
IL-10	5.45 (3.65, 11.83)	2.53 (1.60, 4.55)	0.056
IL-12p70	0.09 (0.00, 0.30)	0.15 (0.07, 0.26)	0.207
IL-17A	1.17 (0.91, 1.28)	1.06 (0.91, 1.39)	0.626
IL-17F	1.31 (1.01, 2.20)	0.95 (0.91, 1.09)	0.003
IL-22	1.64 (1.34, 2.94)	1.48 (1.15, 2.11)	0.520
IFN-γ	9.56 (4.19, 15.16)	5.81 (3.38, 22.22)	0.677
TNF-α	1.14 (0.25, 1.93)	0.57 (0.00, 1.65)	0.575
TNF-β	0.17(0.00, 0.71)	0.00 (0.00, 0.30)	0.140
Children 6-13 years old (SMPP 61 cases vs GMPP 80 cases)
IL-1β	1.26 (0.87, 2.10)	1.20 (0.99, 1.51)	0.496
IL-2	0.77 (0.18, 1.47)	0.82 (0.27, 1.30)	0.784
IL-4	0.99 (0.74, 1.61)	1.12 (0.90, 1.32)	0.474
IL-5	1.45 (1.08, 2.59)	1.73 (1.26, 2.32)	0.500
IL-6	50.00 (17.60, 174.22)	24.56 (11.09, 65.74)	0.056
IL-8	17.09 (8.81, 39.10)	12.20 (7.63, 81.44)	0.752
IL-10	3.69 (2.00, 8.62)	2.24 (1.56, 3.66)	0.003
IL-12p70	0.27 (0.12, 0.53)	0.26 (0.05, 0.40)	0.217
IL-17A	1.02 (0.81, 1.49)	1.08 (0.85, 1.41)	0.585
IL-17F	1.05 (0.92, 2.11)	1.12 (0.93, 1.56)	0.932
IL-22	1.80 (1.12, 3.52)	1.79 (1.24, 3.24)	0.917
IFN-γ	8.73 (3.11, 16.77)	4.69 (3.49, 9.78)	0.046
TNF-α	1.92 (0.64, 4.59)	0.98 (0.07, 1.89)	0.007
TNF-β	0.15 (0.00, 1.04)	0.06 (0.00, 0.74)	0.797

IFN: Interferon; IL: Interleukin; GMPP: General Mycoplasma pneumoniae pneumonia; SMPP: Severe Mycoplasma pneumoniae pneumonia; TNF: Tumor necrosis factor.

Concomitant viral and bacterial pathogens in children with M. pneumoniae pneumonia

In addition to M. pneumoniae, other viral/bacterial pathogens were identified in 27.3% (316/1160) children with MPP. These pathogens accompanied M. pneumoniae in 24.3% (64/264) and 28.0% (252/896) cases of SMPP and GMPP, respectively (Table 4). No statistically significant differences between the presence of concomitant pathogens in SMPP and GMPP cases were observed with the exception of S. pneumoniae/viral co-infection (P = 0.027).

Table 4 Concomitant bacterial and viral pathogens in children with Mycoplasma pneumoniae pneumonia, n (%).

Pathogens	Total (1160 cases)	SMPP (264 cases)	GMPP (896 cases)	P value
Mycoplasma pneumoniae mono-infection	844 (72.7)	200 (75.7)	644 (72.0)	0.213
Mycoplasma pneumoniae with other pathogens	316 (27.3)	64 (24.3)	252 (28.0)	0.213
Bacterial pathogens
Streptococcus pneumoniae	155 (13.3)	30 (11.4)	125 (14.1)	0.278
Haemophilus influenzae	72 (6.2)	10 (3.8)	62 (6.9)	0.064
Streptococcus pneumoniae + Haemophilus influenzae	22 (1.9)	3 (1.1)	19 (2.1)	0.442
Others	15 (1.2)	4 (1.5)	11 (1.1)	0.957
Total	264 (22.6)	47 (17.8)	217 (24.2)	0.092
Viral pathogens
RSV	7 (0.6)	0	7 (0.9)	0.361
RSV + ADV	3 (0.2)	1 (0.4)	2 (0.2)	0.540
HRV	8 (0.7)	3 (1.1)	5 (0.6)	0.565
HRV + ADV	1 (0.1)	0	1 (0.1)	1.0
HRV + HMPV	1 (0.1)	1 (0.4)	0	0.228
ADV	5 (0.5)	2 (0.7)	3 (0.3)	0.320
Others	12 (1.0)	4 (1.5)	8 (0.8)	0.486
Total	37 (3.2)	11 (4.1)	26 (2.9)	0.304
Any mix bacterial/viral pathogens
Streptococcus pneumoniae + virus	6 (0.6)	4 (1.6)	2 (0.2)	0.027
Haemophilus influenzae + virus(es)	6 (0.6)	1 (0.4)	5 (0.5)	1.0
Streptococcus pneumoniae + Haemophilus influenzae + virus(es)	3 (0.3)	1 (0.4)	2 (0.2)	0.540
Total	15 (1.5)	6 (2.4)	9 (0.9)	0.277

ADV: Human adenovirus; GMPP: General Mycoplasma pneumoniae pneumonia; HMPV: Human metapneumovirus; HRV: Human rhinovirus; SMPP: Severe Mycoplasma pneumoniae pneumonia; RSV: Respiratory syncytial virus.

A total of 6 bacterial pathogens, S. pneumoniae, A. baumannii, S. aureus, K. pneumoniae, H. influenzae, and E. cloacae, were detected, while other bacterial pathogens tested, P. aeruginosa, P. maltophilia, E. coli, M. tuberculosis, L. pneumophila, and MRSA-not found. In both groups, SMPP and GMPP, S. pneumoniae (14.4%, 38/264 vs 16.5%, 148/896, P = 0.397) and H. influenzae (5.7%, 15/264 vs 9.8%, 88/896, P = 0.061) were found to be predominant. Sporadically, S. pneumoniae and H. influenzae were simultaneously detected with viral pathogens, ADV, IFA, IFB, HPIV, RSV, HMPV (Supplementary Table 1). Other bacterial pathogens, A. baumannii, S. aureus, K. pneumoniae, E. cloacae, were rarely detected (Supplementary Table 1).

Among the viral pathogens tested, the RSV, HRV, ADV, HMPV, IFA, IFB, HCoV, HBoV and HPIV were identified, sometimes with bacterial pathogens simultaneously (Supplementary Table 1). However, the prevalence of viral pathogens was very low, and they were mostly presented sporadically. Totally, the highest rate was found for ADV (1.9%, 5/264) in SMPP cases, and for RSV (1.6%, 14/896) in GMPP cases (Supplementary Table 1). The prevalence of ADV in SMPP cases was 3.0-fold and 7.6-fold less that for H. influenzae and S. pneumoniae, respectively. The prevalence of RSV in GMPP cases was 6.3-fold and 10.6-fold less that for H. influenzae and S. pneumoniae, respectively.

In 4.9% (13/264) SMPP cases and 3.7% (33/896) GMPP cases (P = 0.108), two or three bacterial/viral pathogens were simultaneously detected with M. pneumoniae (Supplementary Table 1).

Occurrence of bacterial and viral pathogens among the children of different age groups with M. pneumoniae pneumonia

The highest occurrence of concomitant bacterial/viral pathogens was observed in the age group 0-1 year, less–in group 2-5 years, and even less–in group 6-15 years for both SMPP and GMPP cases (Table 5). S. pneumoniae and H. influenzae were predominant co-pathogens in all age groups (Table 6). The occurrence of S. pneumoniae was higher in SMPP cases than in GMPP cases among children aged 0-1 year (30.8% vs 15.8%, P = 0.442), but H. influenzae demonstrated opposite trend in this age group (7.7% vs 13.2%, P = 0,977) (Table 6). Meanwhile, among children aged 2-5 years and 6-15 years, both S. pneumoniae (2-5 years: 23.8% vs 22.4%, P = 0.847; 6-15 years: 14.8% vs 11.4%, P = 0.214) and H. influenzae (2-5 years: 15.9% vs 12.2%, P = 0.535; 6-15 years: 8.2% vs 4.0%, P = 0.040) were more predominant in GMPP cases compared to SMPP cases.

Table 5 Concomitant bacterial and viral pathogens in children of different age groups with Mycoplasma pneumoniae pneumonia, n (%).

Groups	Age group, years	Number of cases, n	Mycoplasma pneumoniae mono-infection	Mycoplasma pneumoniae with other pathogens
				Total	Viral	Bacterial	Bacterial and viral
SMPP	0-1	13	7 (53.8)	6 (46.2)	0	6 (46.2)	0
	2-5	49	35 (71.4)	14 (28.6)	1 (2.0)	12 (24.6)	1 (2.0)
	6-15	202	158 (78.2)	44 (21.8)	10 (4.9)	29 (14.4)	5 (2.5)
	Total	264	200 (75.7)	64 (24.3)	11 (4.2)	47 (17.8)	6 (2.3)
GMPP	0-1	38	23 (60.5)	15 (39.5)	3 (7.9)	10 (26.3)	2 (5.3)
	2-5	164	108 (65.9)	56 (34.1)	4 (2.4)	48 (29.3)	4 (2.4)
	6-15	694	513 (73.9)	181 (26.1)	19 (2.8)	159 (22.9)	3 (0.4)
	Total	896	644 (71.9)	252 (28.1)	26 (2.9)	217 (24.2)	9 (1.0)

GMPP: General Mycoplasma pneumoniae pneumonia; SMPP: Severe Mycoplasma pneumoniae pneumonia.

Table 6 Occurrence of Streptococcus pneumoniae and Haemophilus influenzae in children of different age groups with Mycoplasma pneumoniae pneumonia, n (%).

Groups	Age group, years	Number of cases, n	Mycoplasma pneumoniae with other pathogens
			Total	Streptococcus pneumoniae	Haemophilus influenzae
SMPP	0-1	13	6 (46.2)	4 (30.8)	1 (7.7)
	2-5	49	14 (28.6)	11 (22.4)	6 (12.2)
	6-15	202	44 (21.8)	23 (11.4)	8 (4.0)
	Total	264	64 (24.3)	38 (14.4)	15 (5.7)
GMPP	0-1	38	15 (39.5)	6 (15.8)	5 (13.2)
	2-5	164	56 (34.1)	39 (23.8)	26 (15.9)
	6-15	694	181 (26.1)	103 (14.8)	57 (8.2)
	Total	896	252 (28.1)	148 (16.5)	88 (9.8)

GMPP: General Mycoplasma pneumoniae pneumonia; SMPP: Severe Mycoplasma pneumoniae pneumonia.

Antibiotic treatment

All the patients were treated with macrolides according to the China NHCotPsRo, 2019[26]. If the macrolide resistance of M. pneumoniae was confirmed, doxycycline or minocycline could be used as an alternative. In cases of suspected or determined co-infections with other bacteria, the cefuroxime, ceftriaxone or others were additionally prescribed. Together, children were receiving the combined antibiotic therapy more often than macrolide monotherapy in both SMPP and GMPP cases in all age groups (P < 0.001) with the exception of children aged 0-1 year with SMPP (P = 0.117) (Table 7). Moreover, the children of 6-15 years old were receiving the combined antibiotic therapy more frequently in cases of SMPP compared to cases of GMPP (P = 0.029).

Table 7 Antibiotic treatment of the children of different age groups with Mycoplasma pneumoniae pneumonia, n (%).

Age group (years)	SMPP (n = 264)				GMPP (n = 896)				P value
	Total, n	Mono macrolides	Combined antibiotic therapy	P value	Total, n	Mono macrolides	Combined antibiotic therapy	P value
0-1	13	4 (30.8)	9 (69.2)	0.117	38	9 (23.7)	29 (76.3)	< 0.001	0.891a, 0.891b
2-5	49	10 (20.4)	39 (79.6)	< 0.001	164	51 (31.1)	113 (68.9)	< 0.001	0.146a, 0.146b
6-15	202	28 (13.9)	174 (86.1)	< 0.001	694	144 (20.7)	550 (79.3)	< 0.001	0.029a, 0.029b
Total	264	42 (15.9)	222 (84.1)	< 0.001	896	204 (22.8)	692 (77.2)	< 0.001	0.017a, 0.017b

^aP values indicate statistical significance of mono-macrolide therapy between severe Mycoplasma pneumoniae pneumonia and general Mycoplasma pneumoniae pneumonia groups.

^bP values indicate the statistical significance of combined antibiotic therapy between severe Mycoplasma pneumoniae pneumonia and general Mycoplasma pneumoniae pneumonia groups.

GMPP: General Mycoplasma pneumoniae pneumonia; SMPP: Severe Mycoplasma pneumoniae pneumonia.

After the therapy with macrolides alone or in combination with other antibiotics, all the symptoms including fever and cough were disappeared, and the physical signs were normalized including the disappearance of rales in the lungs. Chest X-ray or CT scan showed obvious absorption of pulmonary inflammation. All drug resistant cases have been eventually cured.

Antibiotic resistance of M. pneumoniae to macrolides

A total of 2498/2680 (93.2%) cases of confirmed MPP were associated with macrolide resistant M. pneumoniae strains carrying A2063G or A2064G mutations in the 23S rRNA gene. The similar value (93.7%) was observed for 1160 MPP cases included in the study. Among them, a total of 246/264 (93.2%) of M. pneumoniae associated with SMPP, and 841/896 (93.9%) of M. pneumoniae associated with GMPP, were found to be resistant to macrolides.

DISCUSSION

CAP is caused by different bacterial and viral pathogens including M. pneumoniae, S. pneumoniae, K. pneumoniae, etc., and certain correlation was found between the causative agents, region, age group, season, and CAP severity[1,3]. This study reported a dramatic increase in hospitalized MPP cases, accounting for 63% of all hospitalized CAP cases in children in Baoding area in 2023, with an unprecedented 93.2% prevalence of macrolide resistant M. pneumoniae with A2063G or A2064G mutations in the 23S rRNA gene.

The specific epidemiological pattern, associated with lifting NPIs in December of 2022, was revealed that affected the seasonal incidence of hospitalized MPP cases in Baoding hospital in 2023, including the low number of MPP cases during January-September followed by dramatic increase with a peak in November that was accompanied by the highest incidence of severe cases of MPP in autumn and winter. It is similar with epidemiological trend observed in 2023 in different regions of China: In Shanghai, where the active spreading of MPP started in the late of summer with a peak in October[26], in Shandong province-with a peak between October and November, especially among school children[27], in Shihezi, Xinjiang region-with a peak between September and November involving preschool and school children[3], and in other Asian and European countries[8]. Also, in 2022 in Shijiazhuang, the SMPP cases were mainly observed during winter season[17], and in 2023 in Shihezi, the severe CAP cases in children often happened during both autumn (44.3%) and winter (28.8%)[3].

Importantly, during the last decade the reports of M. pneumoniae pneumonia in children under 5 years old have been increased[15]. In this study, 22.7% of MPP cases occurred in children under 5 years of age, with 23.5% of SMPP cases among them. The similar trend was reported in Shanghai in 2023[28]. These findings indicate that MPP has re-emerged post-COVID-19, widely involving children of all age groups in the epidemic process. Our data also show that the children aged 0-1 year have responded to therapy slower than other age groups and have demonstrated longest hospitalization duration even in GMPP cases.

Identification of biomarkers that allow to predict the severity of disease plays a crucial role in personalized therapy, improving outcomes and reducing side effects. Numerous studies analyzed different biomarkers that could predict severity of MPP[17,29,30]. In general, leucocyte counts, neutrophil percentage, concentration of CRP, PCT, LDH, D-dimer were higher, but lymphocyte percentage – lower, in patients with SMPP, while some other parameters, such as cytokines showed different results[17,29]. For example, concentrations of IFN-γ, IL-2, IL-5, IL-6, IL-8, IL-10 were remarkably greater in the SMPP group, while TNFα did not statistically differ between SMPP and GMPP groups[17]. In another study, levels of IFN-γ, Il-6, IL-10, TNF-α were elevated in the SMPP group, but the levels of IL-2 and IL-4 did not show statistically significant differences between SMPP and GMPP groups[29]. Meanwhile, multiplex analysis revealed that the levels of 13 cytokines (IL-2, IL-10, IL-11, IL-12, IL-20, IL-28A, IL-32, IL-35, IFN-a2, IFN-g, IFN-b, B-cell activating factor, and thymic stromal lymphopoietin) were higher in SMPP group compared to GMPP group[30]. The observed variations in expression of cytokines can be explained by different study designs. Indeed, the cohort groups significantly varied, i.e., 417 children of 0.2-15 years old[17], 203 children aged ≥ 1 year[29], 36 children of 3-9 years old[30].

In our study, we analyzed the immunological parameters taking into account the age-dependent variations of children’s immunity that might affect MPP clinical outcomes. As known, several dynamic changes in the relative proportions of differential WBC counts during infancy, especially during the first year, and at approximately 5 years, occur in children[31]. The innate and adaptive immunity of children under 5 years old pass development phase, while the children about 5 years old possess more matured immunity[32,33]. Indeed, in the age group 2-5 years, only IL-17F demonstrated significant increase in SMPP cases compared to GMPP, while in the group 6-13 years old, the levels of IFN-γ, TNF-α and IL-10 were higher in SMPP cases compared to GMPP. It is in agreement with recent suggestion that the relationship between ILs and severity of MPP is complex, and it can be influenced by individual differences, infectious agents, and host immune status[17]. Furthermore, our finding may reflect the certain differences of the “pathogen-host” interaction in children of different age groups. Notably, the binary logistic regression analysis revealed two different age-specific models for SMPP prediction. For the children aged 2-5 years, six predictors (CRP, LDH, PCT, ESR, FDPs, D-dimer) of SMPP were included in the model, while for the children aged 6-13 years – four predictors (neutrophils, CRP, FDPs, PT). These findings suggest that SMPP is associated with a more pronounced systemic inflammatory response and coagulation dysfunction although the predictive markers were different in age groups.

Severity of MPP can be often associated with other bacterial/viral pathogens. In particular, the necrotizing pneumonia as complication of MPP can be the result of co-infection[11]. Severe cases of disease can be also associated with co-infection of M. pneumoniae and adenovirus[10] as well as M. pneumoniae and rhinovirus[28]. In present study, co-infecting pathogens were determined in 316/1160 (27.3%) cases of MPP. They were detected in children with both severe and general MPP. However, no significant differences between severity of disease and the presence of co-infecting pathogens (SMPP–24.3%, GMPP–28.0%) was observed. These data indicate that severity of M. pneumoniae pneumonia is multifactorial and may depends not only on species/genus of infectious agents, also their genotypes, immune status and the age of patients, etc. In this study, independently of severity, the majority of children with bacterial/viral co-infections belonged to the youngest age group (0-1 year) suggesting that this age group was more prone to mixed infections that others.

The presence of co-pathogens can also reflect regional and timing variations even in the same country. Indeed, in North East of China, Dandong, Liaoning province, the viral co-infection rate of 38.75% was observed in 748 MPP cases collected in 2021-2023[18]. In South of China, Guangzhou, Guandong province, the viral co-infection rate of 27.27% was detected among 308 patients with MPP registered in October/2023-January/2024[20]. In this study, the viral agents were identified in 4.5% cases of MPP cases registered in January-December/2023 in North of China, Baoding, Hebei province. It also should be taken into account that in addition to regional and timing variations, some other factors can affect the detection rate of pathogens, e.g., specimen type[34], sensitivity of test-systems, also the methods used[10].

During the last decade, the increase of macrolide resistance of M. pneumoniae was observed worldwide including Europe, Western Pacific regions, regions of Americas[13]. However, in Europe, United States, and some other countries it is still not high compared to Asian counties. Indeed, in 2023, in Denmark, macrolide resistance was detected in less than 2% of samples tested[35], in New Zealand it was as low as 9%[36], in the United States-approximately 10%[37]. In contrast, the burden of macrolide resistance of M. pneumoniae is extremely high in Asian countries. In East Asian countries an incidence of drug-resistant MPP exceeds 70%[38]. The emergence of drug resistant M. pneumoniae was also described for Korea, North China, Japan[2,16,39].

In China, the macrolide resistant rate of M. pneumoniae associated with CAP and collected in 2013 varied from 20% (Xiahe, 1/5 cases) to 71.4%-84.2% (Wuhan, 5/7 cases, Changchun, 14/19 cases, Beijing, 235/310 cases, Changsha, 15/18 cases, Shenyang, 85/101 cases) and 100% (Shanghai, 22/22 cases)[40]. In Nanjing, resistant rate of M. pneumoniae was 92.39% during 2014-2016; a total of 276 cases[41]. In 2016, the macrolide resistance reached 66.7% in Harbin, 74.3% in Kunming, 80% in Urumqi, 81.8% in Shanghai, 86.7%-in Beijing; a total of 835 cases[42]. Since recently the high ratio of drug resistant M. pneumoniae has become specific for North part of China, including Beijing region[39]. The stable growth of resistant rate of M. pneumoniae in Beijing has been registered starting from 75.8% in 2013 to 97.4% in 2019[40]. In 2019, M. pneumoniae epidemic in Qingdao was characterized by 100% macrolide resistant rate; 46 cases tested[43]. In 2019, macrolide resistant rate of M. pneumoniae was 98.78% in Weihai; a total of 82 cases[44]. The high incidence of resistance, 85.7%, was observed in Taiwan in 2020; a total of 21 cases[45], and 96% in Wuhan in 2020-2022; a total of 50 cases[46]. In our study, the number of registered and analyzed MPP cases in Baoding hospital in 2023 (2680 cases) was much higher than that in any other regions described above; 93.2% cases of them were associated with macrolide resistant M. pneumoniae strains, and two mutations, A2063G and A2064G, in the 23S rRNA gene were responsible for macrolide resistance that corresponded to M. pneumoniae resistance worldwide[13].

Together, these data suggest an appearance and successful spread of resistant M. pneumoniae clone(s), possibly in response to extensive use of macrolides for treatment of MPP in China[25]. However, until now the impact of macrolide resistance in developing of severe course of MPP has not been proven. On one hand, MPP caused by macrolide resistant M. pneumoniae strains lead to longer duration of fever and hospitalization compared to macrolide sensitive strains[47]. On the other hand, in Europe/North America, where resistance rates are lower (< 15%), severe MPP is less commonly linked to resistance. Obviously, infectious agents, individual differences and host immune response (e.g., cytokine levels) may play a bigger role in severity than resistance alone[1,3,17] that is supported by the results of this study. Currently, the pathogenesis of M. pneumoniae is still not completely clear that requires further studies[48].

Interestingly, since 2023, after the lifting of strict NPIs worldwide, the rise of incidence of MPP in Europe and Asia was observed[8]. In particular, in our study, 2680 hospitalized MPP cases were registered in 2023 in Baoding hospital, Hebei province, with 93.2% of M. pneumoniae macrolide resistance. Also, a total of 15051 cases of MPP were registered in 2023 in Children's Hospital Affiliated to Zhengzhou University, Henan province, and the macrolide resistant M. pneumonia strains were detected in 91% of them[49].

We believe that macrolide resistance provided by well-known mutations A2063G and A2064G is not only the reason for recent effective spread of M. pneumoniae and suggest the existence of M. pneumoniae clone(s) with certain biological advantages, potentially highly contagious, in particular, in Baoding and Zhengzhou. These clone(s) could be accumulated and selected during COVID-19 pandemic when the strict NPIs were used until the end of December of 2022. Subsequently, the most successful clone(s) spread through the “bottleneck” that may represent a “founder effect”. However, this intriguing hypothesis requires further genetic studies.

Given the spread of macrolide resistant M. pneumoniae, the treatment of MPP in children may require alternative medicals rather than macrolides. Finally, a safe and effective vaccine against M. pneumoniae needs to be developed to provide protective immunity and reduce the incidence of MPP in children as it was effectively done for S. pneumoniae worldwide[50].

The study has several limitations. The first, identification rate of pathogens including M. pneumoniae and S. pneumoniae in clinical samples can vary depending on the specimen type. Second, only the patients hospitalized with MPP were included in the study, and it could not represent the general status of MPP in Baoding area.

CONCLUSION

The current data show the dramatic increase in the incidence of hospitalized MPP cases in Baoding area in 2023, that was delayed for nine months after lifting NPIs, characterized by unprecedented 93.2% of macrolide resistant M. pneumoniae, and the prevalence of S. pneumoniae and H. influenzae among concomitant pathogens. On one hand, it indicates prolonged suppression of M. pneumoniae transmission during NPIs and suggests potential alterations in herd immunity patterns. On the other hand, stringent NPIs may have created an “bottleneck” selecting the most successful clone(s) that subsequently spread, and it may represent a “founder effect” that requires further studies.

ACKNOWLEDGEMENTS

We are thankful to Kai-Xuan Cao and Dan-Yang Liu for technical support.