New Resistant Mycoplasma pneumoniae Subclades Track With Severe Pediatric Disease in Taiwan
The post-pandemic return of Mycoplasma pneumoniae in Taiwan was not a routine seasonal rebound. It arrived with a larger clinical footprint, a heavier neurologic burden, and a genomic pattern that suggests the organism itself was changing as pediatric case counts rose.
Compared with the 2019 peak, 2024 data showed a doubling of incidence to 546 per 100,000 population, a 2.2-fold increase in intensive care unit admissions to 10.9 per 100,000 population, and a 13.7-fold surge in central nervous system complications to 6.1 per 100,000 population. Those three endpoints moved in the same direction, which makes the epidemic more than a testing artifact or a transient flare in respiratory disease. It was a more severe wave.
ICU Admission Rose Alongside Objective Markers of Systemic Inflammation
The study drew from a large hospital network spanning four Chang Gung Memorial Hospital branches, then paired that surveillance with a prospective cohort from 2023 to 2024. That combination matters because it links population incidence to individual clinical severity. In the broader surveillance dataset, monthly positivity rates ranged from 10% to 20% during nonpeak periods and rose to 40%–50% during June and July, with major outbreaks in both 2019 and 2024. Seasonal activity also showed a secondary rise in January through March in some years, which suggests that the pathogen remained active year-round rather than behaving as a short, self-limited summer event.
Children 5 to 9 years old carried the highest incidence, followed by those 0 to 4 years old and adolescents 10 to 17 years old. The age gradient is clinically familiar: school-age children have the contact patterns that sustain respiratory spread, but younger children may show more fragile reserve once infection becomes clinically apparent. The most consequential finding was not simply where cases appeared, but where they landed. ICU admission clustered with neutrophilia, thrombocytosis, pleural effusion, and central nervous system involvement, a pattern that points to a strong inflammatory phenotype rather than isolated airway infection.
One sentence captures the severity signal plainly: central nervous system involvement carried the strongest association with ICU admission.
In the prospective cohort of 226 children with laboratory-confirmed infection, neutrophils greater than 7,000/μL had an adjusted odds ratio of 6.2 for ICU admission, with a 95% CI of 1.3 to 29.9. Platelets greater than 400 × 10³/μL had an adjusted odds ratio of 12.5, with a 95% CI of 2.7 to 57.9. Pleural effusion had an adjusted odds ratio of 9.7, with a 95% CI of 1.6 to 58.1. CNS involvement produced an adjusted odds ratio of 27.9, with a 95% CI of 4.1 to 188.3. The confidence intervals are wide, which is common in severe pediatric infection cohorts, but the direction and magnitude of association are difficult to dismiss.
Neurologic Disease Became a More Visible Part of the Clinical Phenotype
The neurologic endpoint deserves special attention because it was defined carefully. CNS complications included isolated encephalopathy, encephalitis, brainstem or basal ganglia encephalitis, acute disseminated encephalomyelitis, cerebellitis, Guillain-Barré syndrome, and transverse myelitis, and each event required PCR confirmation of M. pneumoniae in respiratory or CSF samples without an alternative cause. That definition increases confidence that the neurologic burden reflected true disease linked to the organism rather than coincidental viral coinfection or nonspecific postinfectious illness.
The 13.7-fold surge in CNS complications is particularly important in pediatrics because these syndromes often present after respiratory symptoms are already established. A child with fever and cough may initially appear to have an ordinary atypical pneumonia, but evolving headache, lethargy, ataxia, abnormal movements, or altered mental status should prompt urgent reassessment. The clinical challenge is that M. pneumoniae does not confine itself to the lungs. It can trigger immune-mediated injury, and in some cases the neurologic syndrome becomes the dominant reason for hospitalization.
The study’s CSF analysis reinforces that point. Paired CSF and serum samples were collected on the same day in children with CNS disease, with IgG and albumin measured by commercial ELISA and the CSF IgG index calculated. Those measurements are consistent with an effort to characterize intrathecal immune activity, which fits the biology of postinfectious or para-infectious neurologic injury.
Four Emerging Subclades Replaced Earlier Lineages by Mid-2024
Genomic analysis provides the most compelling explanation for why the epidemic looked different. Whole-genome sequencing was performed on 113 isolates recovered from 234 specimens, and the analysis was expanded using 100 earlier Taiwanese genomes and 485 publicly available genomes spanning 1960 through 2024. Four emerging subclades replaced pre-pandemic lineages by mid-2024, and one subclade was associated with severe outcomes.
This is not a trivial phylogenetic detail. Replacement of older lineages by a small number of newly expanding subclades implies selective pressure, population susceptibility, or a fitness advantage in transmission, persistence, or tissue behavior. Because Taiwan already had a high prevalence of macrolide-resistant M. pneumoniae before the pandemic, with resistance rates reaching 70%, the clinical environment was already primed for treatment failure or delayed response when the organism returned in force.
The study also reminds clinicians that macrolide resistance does not explain everything. Severe disease has been reported even where resistance is low, and prior work has not consistently shown that resistance alone predicts ICU admission. The post-pandemic increase in severity likely reflects a convergence of factors: immunity gaps created by reduced circulation during COVID-19 mitigation, repeated exposure in school-age children, entrenched resistance, and the rise of novel subclades that may behave differently in the host.
Surveillance Scope Captured the Shift From Outpatient Illness to Critical Disease
The surveillance model covered outpatient clinics, emergency departments, general wards, and ICUs, and testing was clinician-driven on the basis of symptoms such as acute lower respiratory tract infection, persistent fever, and suspected CNS disease. Respiratory specimens were tested by real-time PCR, and starting in 2019, the BioFire FilmArray Respiratory Panel was also used. This matters because the data do not depend on one narrow setting. They reflect the full clinical spectrum, from early respiratory illness to ICU-level disease and neurologic complications.
The study excluded children with chronic lung disease, genetic or metabolic disorders, neurologic disease, immunodeficiency, or cancer from the prospective risk-factor cohort. As a result, the identified risk factors are more likely to represent signals from otherwise typical pediatric hosts rather than from children already predisposed to critical illness. For practicing clinicians, that makes the predictors more actionable: neutrophilia, thrombocytosis, pleural effusion, and especially neurologic involvement should raise concern for escalation even in children without major underlying disease.
A Moving Pathogen Demands Faster Recognition and Closer Follow-Up
The clinical message is straightforward. A post-pandemic rise in Mycoplasma pneumoniae in children can no longer be treated as a simple return to baseline activity. In Taiwan, the resurgence was larger than the 2019 peak, more likely to require ICU care, and more likely to involve the nervous system. The same period also saw replacement of older lineages by four emerging subclades, one linked to severe outcomes, which keeps the question of pathogen evolution firmly in view.
The immediate bedside implication is earlier recognition of children at risk for deterioration. Persistent fever, pleural effusion, rising neutrophils, marked thrombocytosis, or any neurologic symptom should lower the threshold for escalation, especially in regions with high macrolide resistance. What the study cannot yet define is the exact contribution of each subclade to virulence, transmission, or neurotropism. The association is strong enough to justify surveillance, but the mechanism remains unfinished, and future sequencing tied to clinical outcomes will need to determine whether these new lineages are simply replacing older ones or driving a genuinely more severe pediatric phenotype.