Predominance of Macrolide-resistant Bordetella pertussis in a Japanese Pediatric Intensive Care Unit

To the Editors: Macrolide-resistant Bordetella pertussis (MRBP), first reported in 2011, became predominant in China and subsequently spread in East and Southeast Asia.1 In Japan, MRBP was first detected in 2018, with the emergence of severe cases reported in 2024.2 However, reports on the prevalence and clinical impact of MRBP remain scarce except Chinese cohorts.3 Trimethoprim-sulfamethoxazole (TMP-SMX), the recommended treatment for MRBP, is contraindicated in infants aged under 2 months.4 Thus, the restrictive consensus in treatment complicates the management of young infants infected with MRBP. We herein report a high prevalence of MRBP in our pediatric intensive care unit, where most patients required mechanical ventilation, 2 needed extracorporeal membrane oxygenation (ECMO), and delayed TMP-SMX administration was associated with prolonged ventilation and poorer outcomes. Between January and September 2025, 12 pertussis patients were hospitalized in our pediatric intensive care unit (Table, Supplemental Digital Content 1, https://links.lww.com/INF/G518). Ten (83.3%) were infants under 2 months old. All were unvaccinated, and 7 patients had symptomatic household members. The median duration from onset to intubation was 10 days, and the median duration of mechanical ventilation was 11 days. In 10 patients (83.3%), high-level macrolide resistance was confirmed by the A2047G mutation in the 23S rRNA gene. Antimicrobial resistance testing against clarithromycin and azithromycin of isolates from 7 of 10 patients supported this result. All of the MRBP manifested a highly virulent genotype of the promotor of pertussis toxin (ptxP3).5 TMP-SMX was initiated at 11 days on median after symptom onset following informed consent from the parents. There were no adverse effects such as hyperbilirubinemia, kidney injury or myelosuppression except for 2 ECMO patients. Early interventions with TMP-SMX were not correlated with either the duration of mechanical ventilation (P = 0.057) or the length of intensive care unit (ICU) stay (P = 0.12, Spearman correlation analysis; Fig. 1). However, the treatment appeared to be effective in 7 patients who required mechanical ventilation with endotracheal intubation and 3 patients who did not. Patients were retrospectively divided into 2 groups based on the initiation of TMP-SMX treatment: early (within 14 days) and delayed (after 14 days). The early treatment group showed a lower incidence of pulmonary hypertension, although the difference did not reach statistical significance (16.7% vs. 75.0%; P = 0.09, Fisher exact test). Only 2 patients in the delayed treatment group required ECMO, resulting in a single fatality.FIGURE 1.: Association of TMP-SMX initiation date with mechanical ventilation duration and length of ICU stay. Scatter plots of TMP-SMX initiation date versus mechanical ventilation duration (A) or ICU stay length (B). The open circles indicate patients who underwent endotracheal intubation for mechanical ventilation; the closed circles indicate those without intubation.In 2025, Japan experienced a major outbreak of Bordetella pertussis, lacking data of macrolide resistance. Within a short period, an unprecedented number of severe MRBP patients admitted to our hospital. This observation raises the immediate concern that MRBP may be extending beyond previously reported regions into countries such as Japan. In this study, the delayed TMP-SMX treatment was associated with prolonged mechanical ventilation and poorer outcomes. Both the known side effects and turnaround time for MRBP diagnosis negatively influenced the decision to use TMP-SMX. Our results suggest that early recognition of MRBP and timely administration of TMP-SMX may reduce the risk of clinical deterioration of MRBP. Follow-up studies are warranted to confirm these findings.