We analyzed 935 respiratory specimens collected from case patients clinically suspected of having pertussis in northern India during January 2019−August 2023. We sourced specimens from 2 cohorts: group I (n = 213), comprising patients hospitalized with acute respiratory illness at Postgraduate Institute of Medical Education and Research’s Advanced Pediatric Center; and group II (n = 722), obtained through the Vaccine-Preventable Disease Surveillance Network according to guidelines (9). We subjected 778 samples (178 from group I, 600 from group II) to real-time quantitative PCR (qPCR) (10) and examined 733 samples (101 from group I, 632 from group II) through serologic testing for pertussis toxin (PT) IgG using ELISA.

Figure

Figure. Yearwise trend of Bordetella pertussis and B. holmesii positivity from study of emergence of B. holmesii–associated pertussis-like illness, northern India, 2019–2023. Data based on...

Of 778 swab samples tested by qPCR, we determined 383 (49.2%) to be positive for Bordetella IS481. Among those, we identified B. pertussis monoinfection in 52 (13.6%) samples, B. holmesii monoinfection in 143 (37.3%), and co-infection with both species in 15 (3.9%) samples (Table 1). Of note, B. holmesii positivity surpassed that of B. pertussis in 2022–2023 in the North India catchment. (Figure).

Both species predominantly affected infants <1 year of age. However, B. holmesii was significantly more prevalent in the 5–10-year age group (≈30% compared with ≈9% B. pertussis; χ² = 16.22; p = 0.00102). The mean patient age for B. holmesii infection was 3.4 years and for B. pertussis was 1.9 years (Appendix Figure 2). We detected PT IgG with overall seroprevalence (positive + intermediate) in 232 (31.6%) of the 733 samples tested. We collected serology samples from 174 of the 210 cases that tested positive by qPCR.

Stratifying the qPCR and serology results of the samples for which qPCR and serology results were obtained, we identified 121 B. holmesii qPCR-confirmed cases, 43 of which we determined to be positive for PT IgG by ELISA. By that same method, we identified 41 qPCR-confirmed B. pertussis cases, determining 9 to be positive for PT IgG by ELISA (Table 2). Among 91 cases with known diphtheria-pertussis-tetanus vaccination status, 28 ELISA‐negative, unvaccinated case-patients (i.e., received no pertussis-containing vaccine [zero‐dose]) tested positive for B. holmesii monoinfection and 8 ELISA-negative, unvaccinated case-patients tested positive for B. pertussis monoinfection.

Our study highlights the emergence of B. holmesii as a key contributor to pertussis-like illness in northern India during 2021−2023. Species-specific qPCR revealed that B. holmesii detection rates overtook those of B. pertussis in the 2022–2023 period, particularly among children 5–10 years of age, suggesting an evolving epidemiology distinct from classic pertussis and echoing observations noted by researchers investigating other B. holmesii outbreaks (11–13). Our study demonstrates evidence of notable B. holmesii circulation in the India subcontinent, underscoring this pathogen’s potential role in co-infection and also as a primary etiologic agent, as demonstrated by the presence of this bacterium in unvaccinated PT IgG–negative case-patients (despite limitations in detailed clinical characterization of pertussis-like symptoms [e.g., cough duration, paroxysms] observed uniformly across cases).

In serology testing, PT IgG positivity (Table 1) among cases with B. holmesii monoinfection (qPCR-positive cases) might reflect recent B. pertussis exposure or vaccination. The absence of PT, pertactin, and fimbrial antigens from the B. holmesii genome and the lack of cross-protective immunity in whole-cell or acellular pertussis vaccines in animal models further reveals critical gaps in current immunization strategies (4,5). Although B. holmesii shares a filamentous hemagglutinin homologue and a conserved 66-kb pathogenicity island with B. pertussis, those shared elements have not yielded effective cross-protection or reliable serologic markers (14).

Our findings highlight the practicality of integrating molecular assays into routine surveillance to accurately distinguish Bordetella species and guide public health responses. Future research efforts should include more detailed clinical characterization of patients, population-based seroepidemiologic studies, and retrospective analysis of archived samples to clarify the historical prevalence of B. holmesii. Revisiting vaccine antigen composition to address nonpertussis Bordetella species might help close critical prevention gaps in pertussis-like disease control.