Disclaimer: Early release articles are not considered as final versions. Any changes will be reflected in the online version in the month the article is officially released.
Author affiliation: Japanese Red Cross Kumamoto Hospital, Kumamoto, Japan (T. Obara, K. Kano, N. Yogo, Y. Muto, K. Hirai); Okayama University, Okayama, Japan (T. Obara, T. Yorifuji, K. Tsukahara, T. Yumoto, H. Naito, A. Nakao)
Since the relaxation of COVID-19–related public health measures, health officials have observed increased pertussis notifications in some regions of Japan. By April 2025, reported cases had already exceeded the previous year’s total. Alerts from the Japan Pediatric Society and the National Institute of Infectious Diseases have raised concerns about the rising incidence and spread of macrolide-resistant Bordetella pertussis (MRBP) (1,2), which has increased in East Asia, particularly in China, since around 2008 and more recently in Japan in 2024 (3,4). Nonetheless, clinical reports describing localized increases in pertussis incidence and the impact of MRBP on critical illness in infants remain scarce.
During late March through early May 2025, we identified an outbreak in Japan involving 10 unvaccinated infants < 2 months of age with PCR-confirmed pertussis. All infants were born at term and without perinatal complications. Nine required admission to the pediatric intensive care unit (PICU) for respiratory failure; 6 of those infants had confirmed MRBP infection. We conducted a retrospective descriptive analysis to characterize clinical features and examine potential association with regional pertussis trends.
Kumamoto Prefecture, in central Kyushu in western Japan, spans ≈7,400 km2 and had a population of around 1,690,000 in fiscal year 2024, including 210,000 children (<15 years of age). The region has single PICU, an 8-bed general unit at the Japanese Red Cross Kumamoto Hospital, which provides emergency critical care for children. Since January 1, 2018, pertussis has been classified in Japan as a category V infectious disease, requiring physician reporting of confirmed cases. We diagnosed pertussis at admission using the BIOFIRE FILMARRAY Respiratory Panel (bioMérieux, https://www.biomerieux.com). We performed antimicrobial susceptibility test or 23S rRNA gene sequencing to estimate macrolide susceptibility of infected strains. We extracted data from medical records and conducted analysis with Stata 19 (StataCorp LLC, https://www.stata.com).
Figure
Figure. Cumulative reported pertussis cases (all ages) in Kumamoto Prefecture and infant PICU admissions at Kumamoto Red Cross Hospital, from study of a localized outbreak of macrolide-resistant pertussis in infants, Japan,…
We investigated annual cumulative pertussis cases in Kumamoto Prefecture (all ages, since 2018) together with infant PICU admissions at Japanese Red Cross Kumamoto Hospital (Figure) (5). During the 2025 outbreak, both the number and proportion of PICU admissions were higher than in previous years. All 9 infants had identifiable sick contacts within their household, most frequently school-aged siblings (Table). Six patients received azithromycin (5-day oral treatment) before PICU admission. After the fourth PICU admission, we tested nasopharyngeal swab specimens from the subsequent 6 infants by using bacterial culture or genetic analysis. One sample tested positive for B. pertussis on culture, showing erythromycin resistance. The remaining 5 samples underwent direct DNA sequencing of the B. pertussis 23S rRNA gene, revealing the A2047G mutation in all cases. The remaining infant, not in PICU, had a macrolide-sensitive strain.
All PICU patients required intubation a median interval of 1 day after admission. The most frequent indications were paroxysmal coughing with desaturation and bradycardia. Most infants had persistent coughing and elevated airway resistance after intubation; 8 of 9 required continuous neuromuscular blockade. Six infants received oral trimethoprim/sulfamethoxazole for ≤14 days without adverse events, all from the fourth PICU case onward after MRBP identification. We did not give intravenous antibiotics prophylactically but initiated them after respiratory management for suspected secondary bacterial infection, including ventilator-associated pneumonia; 7 of 9 patients received ampicillin or ampicillin/sulbactam. Leukocytosis (>50,000 cells/μL) occurred in 3 infants, all treated with leukoreduction. Two patients developed pulmonary hypertension; 1 received inhaled nitric oxide. Median ventilation duration was 8 days and PICU stay 15 days for the PICU patients, longer than in our prior pertussis PICU cases (13 cases during 2012–2024). All patients survived without neurologic sequelae.
The temporal clustering of severe pertussis cases over a 2-month period suggests that increased community transmission contributed to the rise in critical cases. Although surveillance in Japan has reported sporadic MRBP cases, no prior studies have described a localized clinical outbreak. Despite the lack of prefectural-level susceptibility data, all 6 tested PICU cases were MRBP-positive, suggesting that a substantial proportion of circulating strains during this outbreak were resistant.
Pertussis most frequently affects children but also occurs in adults and adolescents. Outbreaks have been reported in fully vaccinated populations because of waning immunity, typically 5–10 years after vaccination (6,7). However, Japan does not currently require pertussis boosters beyond early childhood. From a public health perspective, a layered approach is essential: timely infant vaccination per the national schedule, booster doses for school-aged children to address waning immunity, and maternal tetanus-diphtheria-pertussis immunization during pregnancy to provide passive neonatal protection. Together, those strategies may help reduce secondary transmission within households and mitigate severe outcomes.
In severe infant pertussis, elevated leukocyte counts and pulmonary hypertension are associated with poor outcomes, underscoring the need for early monitoring and critical initiating of care-equipped setting (8–10). In the outbreak we describe, prolonged mechanical ventilation and extended PICU stays intermittently filled all PICU beds with MRBP-infected infant, placing a disproportionate burden on pediatric critical care capacity. This experience highlights how even localized outbreaks can strain systems in regions with only 1 PICU. In pertussis-endemic areas, considering antimicrobial-resistant strains, including MRBP, may support timely treatment decisions and improved outcomes.
Dr. Obara is a pediatric emergency and critical care researcher at Okayama University, Japan. He is also clinically involved in the care of critically ill pediatric patients in Okayama and Kumamoto. His primary research interests include pediatric resuscitation, severe infectious diseases, neurocritical care, and end-of-life care, including organ donation.