Macrolide-Resistant Bordetella pertussis Surfaces During Peru’s 2025 Pertussis Surge
Peru’s 2025 pertussis outbreak did more than strain pediatric services and public health surveillance. It exposed the arrival of macrolide-resistant Bordetella pertussis across multiple regions, including a concentration of cases in infants at the highest risk for severe disease and death. Among 68 pertussis-positive patients tested early in the outbreak, 21 carried the A2047G mutation in 23S rRNA genes, the genetic change most closely linked to macrolide resistance. The mutation rate reached 31%, a level that immediately raises concern for empiric azithromycin use in a setting where prompt treatment is usually intended to reduce transmission and protect vulnerable infants.
The finding lands in the middle of a national outbreak that had already accelerated sharply. Reported incidence rose from 3.9 cases per 1 million inhabitants in 2023 to 7.4 in 2024, then climbed to 57.5 per 1 million by August 2025. By the end of the year, 4,976 cases and 76 deaths had been recorded, with Loreto carrying the heaviest burden. In Datem del Marañón, Indigenous communities were affected disproportionately, reflecting the same vulnerability seen in many pertussis outbreaks where access gaps and incomplete vaccine coverage leave infants exposed before they can complete the primary immunization series.
The age profile of the resistant infections was especially concerning. Patients with macrolide-resistant pertussis were significantly younger than those infected with susceptible strains, with a median age of 3.6 months versus 36.1 months. Most resistant infections occurred in infants younger than 6 months, the stage of life when pertussis can progress quickly from cough to apnea, feeding difficulty, pneumonia, or respiratory failure. In this age group, the value of early antimicrobial therapy is not limited to symptom control; it is also meant to blunt bacterial carriage and interrupt spread within households and maternity wards. Once resistance enters the picture, that strategy becomes less dependable.
One sentence captures the bedside implication plainly: azithromycin may no longer be a safe assumption for empiric pertussis treatment in parts of Peru.
Among the 16 patients with available treatment data, 14 received azithromycin. That pattern is clinically understandable because macrolides remain standard first-line therapy in many settings. Yet the culture data offered a hard contradiction. All 7 recovered isolates were resistant to every antimicrobial drug tested, with no inhibition zones for erythromycin or azithromycin. Those results are not abstract microbiology; they imply that the standard drug choice given to most of these patients may have delivered little or no antimicrobial effect against the circulating strain.
Genomic Evidence Points to More Than One Introduction
Whole-genome sequencing made the outbreak signal sharper rather than flatter. All 7 sequenced isolates belonged to sequence type 2 and the ST2 clonal complex, and all carried the prn150 allele, a marker associated with the emerging MR-MT28 lineage. The genetic profile also included the A2047G mutation and a BP0685 G1959A substitution. Yet the Peruvian isolates did not collapse into a single branch of the phylogenetic tree. Instead, they split into two genetically distinct MR-MT28 sublineages.
That split matters because it argues against a single local expansion event. Three isolates clustered in the PG1 sublineage and carried IS481-mediated prn disruption, consistent with pertactin-deficient organisms. The remaining four clustered with the MR-MT28-other group and retained an intact prn coding sequence despite carrying the prn150 allele. In practical terms, Peru was not dealing with one resistant clone spreading in one direction; the genomic pattern fits multiple introductions of an already established resistant lineage into the country.
The presence of pertactin-deficient strains adds another layer of concern. Pertactin loss has often been described in settings using acellular pertussis vaccines, where vaccine-driven immune pressure may favor strains that alter surface antigens. Peru continues to use a whole-cell pertussis schedule, yet three PG1 isolates showed IS481-mediated prn disruption. The most plausible explanation is importation of a pertactin-deficient MR-MT28 sublineage rather than independent local evolution under the country’s vaccine program. One isolate came from an infant who had already received two doses of the pentavalent vaccine, showing that partial immunization does not rule out infection by these adapted strains.
Treatment Assumptions and Surveillance Gaps
Macrolides remain the backbone of pertussis treatment because they reduce carriage, limit spread, and can lower the risk of death in infants. The problem is not theoretical resistance in a distant database. It is resistant pertussis already circulating in a country facing its largest outbreak in decades. Alternative options exist in vitro, including sulfamethoxazole/trimethoprim, but the clinical evidence supporting substitute regimens remains limited. That leaves clinicians in a difficult position: continue a familiar drug class with declining microbiologic reliability, or move to less validated alternatives when the stakes are highest in infants.
Surveillance infrastructure will decide how well systems can respond. This study linked direct PCR detection of the A2047G mutation with phenotypic culture and whole-genome sequencing, a combination that revealed both resistance and transmission structure. Yet the surveillance system still lacked systematically collected clinical data tied to genomic findings. Without those data, no one can say from this outbreak dataset whether macrolide-resistant infections produced worse outcomes, more prolonged illness, or different hospitalization patterns than susceptible infections. The missing link is clinical correlation, and it remains essential if treatment policies are going to change on evidence rather than alarm.
The outbreak also shows why genomics cannot sit apart from routine diagnostics. A resistant organism can move through multiple regions before standard practice notices the shift. Here, MRBP appeared in Lima, Cusco, Arequipa, Ancash, Loreto, and Puno, suggesting broad geographic dissemination rather than a single contained cluster. In a disease as contagious as pertussis, that distribution raises the probability of further household spread, especially when infants are infected before they can complete the vaccine series.
Public health responses now face a narrow but urgent task: expand genomic surveillance, preserve phenotypic susceptibility testing, and revisit empiric treatment guidance for pertussis in Peru. The study does not show whether resistance worsened severity, and it does not prove which alternative antimicrobial strategy should replace macrolides. What it does show, with unusual clarity, is that resistant Bordetella pertussis has arrived during a major outbreak, spread across regions, and entered the case mix of the youngest patients.
The next question is not whether resistance exists. It is how quickly surveillance, diagnostic capacity, and treatment recommendations can catch up before more infants are treated with drugs that no longer match the organism in front of them.