Figure 1

Figure 1. Geographic distribution of confirmed pertussis cases and MRBP cases in Peru, 2025. Regional shading shows cumulative confirmed pertussis cases since epidemiologic week 53 of 2025; darker blue indicates higher case...

In early 2025, we screened for MRBP in B. pertussis–positive nasopharyngeal swab samples from 68 patients (11) (median 3 months of age; range 1 month–39 years of age) by using direct PCR (Appendix Table 1) targeting the A2047G mutation in the 23S rRNA genes. We detected the A2047G mutation in 21 (31%) of 68 samples. Patients infected with MRBP were significantly younger (median 3.6 vs 36.1 months of age; p = 0.007) than patients infected with macrolide-susceptible B. pertussis strains. However, we observed no significant differences in patient sex or vaccination status between the 2 groups (Appendix Table 2). Most MRBP infections were in infants <6 months of age (62%). Among the 16 patients with available treatment information, 14 (88%) received azithromycin. MRBP cases were distributed across multiple regions, including Lima (n = 11), Cusco (n = 4), Arequipa (n = 3), Ancash (n = 1), Loreto (n = 1), and Puno (n = 1) (Table; Figure 1; Appendix Table 2).

From the 68 B. pertussis–positive samples, we obtained 7 culture-positive B. pertussis isolates, all of which were MRBP: 6 collected in 2025 and 1 in late 2024 (isolate no. tf518-24). We performed phenotypic susceptibility testing by using established methods (12). All 7 isolates were resistant to all antimicrobial drugs tested, with no zones of inhibition for erythromycin or azithromycin.

We performed whole-genome sequencing on all 7 MRBP strains (Appendix). All sequenced isolates were assigned to sequence type 2 (ST2) and clonal complex ST2. Six isolates shared the virulence-associated allelic profile ptxP3/ptxA1/ptxB1/ptxC4/fhaB1/fim2-1/fim3-1, whereas 1 isolate (tf098–25) carried the closely related ptxP29 promoter allele but retained the remaining alleles of this profile. We identified the prn150 allele in all sequenced isolates (Appendix), which is a genetic marker of the emerging MR-MT28 lineage (8). We did not identify any inactivating mutations in the ptx operon or fhaB, but 3 isolates (tf518-24, tf120-25, and tf098-25) had IS481-mediated prn disruption, consistent with a pertactin-deficient genotype (Appendix).

Figure 2

Figure 2. Maximum-likelihood phylogenetic tree of Bordetella pertussisisolates on the basis of genome-wide single-nucleotide polymorphism analysis, showing macrolide-resistant isolates from Peru within the emerging MR-MT28 lineage (red text) among globally...

Expanded whole-genome single-nucleotide polymorphism phylogenetic analysis incorporating globally representative MRBP genomes placed all isolates from Peru within the recently emerged MR-MT28 lineage. That expanded analysis also detected previously identified genetic markers, including the A2047G mutation, the prn150 allele, and the BP0685 G1959A substitution (8). The 7 MRBP isolates from Peru did not form a single monophyletic cluster. Three isolates (tf518-24, tf120-25, and tf098-25) clustered within the phylogenetic group 1 (PG1) sublineage and had IS481-mediated prn disruption, whereas the remaining 4 isolates clustered with the MR-MT28-other group and retained an intact prn coding sequence despite carrying the prn150 allele (Figure 2; Appendix).

We described the emergence of MRBP during the ongoing 2025 pertussis outbreak in Peru and documented circulation of the emerging MR-MT28 lineage in Latin America. Genomic analyses identified >2 genetically distinct MR-MT28 sublineages, including PG1-associated pertactin-deficient isolates, consistent with the ongoing international diversification and dissemination of this lineage.

Consistent with recent global genomic analyses, the close phylogenetic relationship between the 7 isolates from Peru and contemporary MR-MT28 strains is consistent with the ongoing international dissemination and diversification of this lineage (5,13). The presence of 2 genetically distinct MR-MT28 groups circulating in Peru suggests multiple international introductions of MRBP into Peru, rather than a single transmission event.

Pertactin-deficient isolates have predominantly been reported in countries using acellular pertussis vaccines, where loss of pertactin is considered an adaptive response to vaccine-induced immune pressure (5,13). Although Peru continues to use a whole-cell pertussis vaccine, 3 PG1 sublineage isolates had IS481-mediated prn disruption, suggesting introduction of an established pertactin-deficient MR-MT28 sublineage, rather than independent emergence under local vaccine-driven selection. Of note, 1 of those isolates (tf518-24) was recovered from an infant who received 2 doses of the pentavalent vaccine, indicating that a pertactin-deficient MR-MT28 strain was identified in a partially vaccinated infant.

We found that most patients with MRBP received azithromycin for treatment, which raises concerns about the effectiveness of macrolides as first-line therapy in Peru. Although B. pertussis has remained susceptible to sulfamethoxazole/trimethoprim in vitro (14) and possibly to other antimicrobial drugs (15), clinical evidence supporting alternative regimens is limited, underscoring the need for urgent clinical and epidemiologic studies.

A key limitation of our study is the lack of systematically collected clinical data linked to genomic findings within the current surveillance system, which precludes assessment of differences in disease severity and outcomes between macrolide-resistant and macrolide-susceptible B. pertussis infections during this outbreak. Addressing that gap will be critical to define the clinical impact of macrolide resistance and to inform evidence-based treatment strategies in high-prevalence settings.

Our findings underscore the urgent need to strengthen genomic surveillance and diagnostic capacity for B. pertussis to enable early detection of resistant strains and inform timely prevention and control strategies. In addition, reassessing empiric treatment recommendations and prioritizing prospective studies to evaluate the clinical effect of macrolide antimicrobial drug resistance will be critical to guide evidence-based public health responses.