Peru's worst whooping cough outbreak in decades turns up drug-resistant bacteria
Nearly a third of 68 sampled cases carried the mutation that blocks first-line antibiotics — and the resistant strain hit infants under six months hardest, during an outbreak that reached 4,976 cases and 76 deaths.
| Group | Value (value) |
|---|---|
| 2023 | 3.9 |
| 2024 | 7.4 |
| Aug 2025 | 57.5 |
The first-line treatment for whooping cough is a macrolide antibiotic — azithromycin, usually — given to shorten illness, reduce transmission, and lower the risk of death in infants [s1]. A report published on 1 October in the CDC journal Emerging Infectious Diseases finds that, during Peru's largest pertussis outbreak in decades, a substantial share of the bacteria causing it no longer respond to that drug [s1].
An outbreak that kept climbing
Peru's national pertussis incidence rose from 3.9 cases per million inhabitants in 2023 to 7.4 per million in 2024, then reached 57.5 per million by August 2025 [s1]. By the end of 2025 the country had recorded a cumulative 4,976 cases and 76 deaths [s1].
The burden was heavily concentrated. The Loreto region in the Amazon accounted for 3,922 of those cases and 57 of the deaths, falling disproportionately on Indigenous communities in Datem del Marañón, where pertussis vaccine coverage remains suboptimal [s1]. Peru's schedule calls for three primary doses of the pentavalent DTwP–HepB–Hib vaccine at 2, 4, and 6 months, DTwP boosters at 18 months and 4 years, and a maternal Tdap dose during pregnancy at 20–36 weeks of gestation [s1] — a program that depends on reaching every community to work.
Testing for resistance
Macrolide-resistant Bordetella pertussis is driven mainly by a single change, the A2047G mutation in the bacterium's 23S rRNA genes [s1]. It has emerged worldwide, with the highest reported prevalence in China and only scarce reports from Latin America until now [s1].
In early 2025, the investigators screened 68 patients whose nasopharyngeal swabs were positive for B. pertussis, using direct PCR to look for that mutation [s1]. They found it in 21 of the 68 samples — 31% [s1]. The resistant infections were distributed across several regions: Lima (11 cases), Cusco (4), Arequipa (3), and one each in Ancash, Loreto, and Puno [s1].
Culture confirmed the picture. Of the 68 positive samples, 7 yielded B. pertussis isolates, and every one was macrolide-resistant — 6 collected in 2025 and 1 in late 2024 [s1]. Phenotypic testing showed all 7 resistant to the drugs tested, with no zone of inhibition around erythromycin or azithromycin [s1].
Who it struck
The resistant strain did not spread at random across age groups. Patients infected with macrolide-resistant bacteria were significantly younger than those with susceptible strains — a median of 3.6 months versus 36.1 months (p = 0.007) [s1]. Most resistant infections, 62%, were in infants under 6 months old [s1] — exactly the group for whom pertussis is most dangerous and for whom effective antibiotics matter most. Among the 16 patients with treatment information available, 14 (88%) had received azithromycin [s1], the drug the resistant bacteria can shrug off.
Two introductions, not one
Whole-genome sequencing of the 7 resistant isolates pointed to more than a single resistant lineage taking hold. All belonged to sequence type 2, but the analysis revealed two genetically distinct groups, which the authors read as multiple separate introductions of resistance into Peru rather than one clone spreading from a single origin [s1]. The isolates carried the prn150 allele, a genetic marker of the emerging MR-MT28 lineage associated with macrolide resistance internationally [s1].
That detail matters for how the problem is likely to behave. A single imported clone can, in principle, be contained; resistance arriving by more than one route, against the backdrop of a large and still-active outbreak, is harder to wall off and more likely to become established. The resistant isolates belonged to clonal complex ST2, and six of the seven shared a single virulence-associated allelic profile, with one carrying a closely related promoter variant — a tight genetic cluster consistent with strains that are already circulating and evolving locally rather than appearing once and dying out.
What it changes
For clinicians, the finding complicates the reflex to reach for azithromycin, particularly in the youngest infants where the resistant strain concentrated — though the report stops at describing the emergence rather than rewriting treatment guidance. Its explicit recommendation is for strengthened surveillance: routine testing for macrolide resistance so that control strategies can be built on current data rather than the assumption, now shown to be unsafe in Peru, that the first-line drug still works everywhere [s1]. In a country where the outbreak had already reached nearly 5,000 cases and dozens of infant deaths, that gap between assumption and reality is not academic.
Sources
- Emergence of Macrolide-Resistant Bordetella pertussis, Peru, 2025 — Emerging Infectious Diseases (CDC), 1 October 2026
Sources
- Emergence of Macrolide-Resistant Bordetella pertussis, Peru, 2025 — Emerging Infectious Diseases (CDC) , October 1, 2026
Shigella found in one in five young children's diarrhoea, with resistance climbing
Surveillance across seven countries detected Shigella by PCR in 20.0% of diarrhoea episodes in toddlers, and found ciprofloxacin resistance reaching 37.2% — a warning for a disease with no licensed vaccine.
Whooping cough surged in Japan in 2025 as it ebbed elsewhere in the Pacific
National surveillance recorded 60,826 cases in Japan by July 2025 — about 15 times the previous year's 4054 — concentrated in school-age children, while four other countries had already peaked and declined.
One in three positive mpox PCR tests in DR Congo may be false, a study finds
An observational study across four DR Congo outbreak sites estimates that a third of positive mpox PCR results reflect environmental viral DNA, not infection — and that a stricter cutoff fixes most of it.
Nigeria confirmed 6,288 Lassa fever cases in six years, and one in six died
A national analysis of 2020–2025 surveillance found a case fatality rate of 15.5% that never improved, a season tied to heat and rainfall, and a falling share of suspected cases ever confirmed.