6533b7d5fe1ef96bd1263c85

RESEARCH PRODUCT

A genomic epidemiology investigation of yaws re-emergence and bacterial drug resistance selection

subject

description

Abstract Background In a longitudinal study assessing the WHO strategy for yaws eradication using mass azithromycin treatment, we observed resurgence of yaws cases with dominance of a single JG8 sequence type and emergence of azithromycin-resistant Treponema pallidum. Here, we analyse genomic changes in the bacterial population using samples collected during the study. Methods We performed whole bacterial genome sequencing directly on DNA extracted from 37 lesion swabs collected from patients on Lihir Island, Papua New Guinea, between 2013 and 2016. We produced phylogenies and correlated these with temporo-spatial information to investigate the source of new cases and the emergence of five macrolide-resistant cases. We used deep amplicon sequencing of surveillance samples to assess the presence of minority macrolide resistance populations. Findings We recovered 20 whole Treponema genomes, and phylogenetic analysis showed that the re-emerging JG8 sequence type was composed of three bacterial sub-lineages characterised by distinct temporo-spatial patterns. Of five patients with resistant Treponema, all epidemiologically linked, we recovered genomes from three and found no variants. Deep sequencing showed that pre-treatment, the index patient harboured fixed macrolide-sensitive Treponema, while the post-treatment sample harboured a fixed resistant genotype, as did three of four contact cases. We also found no evidence of pre-existing minority Treponema drug-resistance variants in the general population. Interpretation In this study, re-emergence of yaws cases was polyphyletic, indicating multiple epidemiological sources. However, given the genomic and epidemiological linkage of resistant cases and the rarity of resistance alleles in the general population, it is likely that azithromycin resistance evolved only once in this study, followed by onward dissemination. Funding Wellcome (Grant #206194), Provintial Deputation of Barcelona (Grant #1940000465). Research in context Evidence before this study We searched PubMed using the terms “yaws”, “Treponema pallidum”, “mass treatment”, and “azithromycin”, without date or language restrictions. Mass drug administration (MDA) using azithromycin (single-dose 30 mg/kg, maximum 2 g) has been a successful strategy to reduce yaws prevalence in endemic communities, with trials conducted in both Papua New Guinea and Ghana. However, there are no long-term follow-up studies other than in Papua New Guinea, where a resurgence of yaws cases was observed 24-months after MDA. Multilocus sequence typing (MLST) showed a change in Treponema pallidum subspecies pertenue (TPE) to a single dominant JG8 type, and identified resistant strains with the same MLST type, but this method lacked the resolution to determine the epidemiological mechanism leading to yaws re-emergence and how macrolide resistance manifested. Added value of this study This study is the first to use whole genome sequencing (WGS) for a detailed investigation of TPE transmission and to apply deep amplicon sequencing to TPE macrolide resistance mutations. Leveraging the increased resolution of WGS over other techniques, we show that re-emergence of yaws after MDA was the result of at least three distinct bacterial sub-lineages with different temporo-spatial properties (likely due to several infected people missing treatment) rather than a single point-source. WGS also shows that patients with macrolide-resistant yaws had genomically indistinguishable TPE, indicating a recent common ancestor, and possibly a recent chain of transmission originating from the index case. Deep amplicon sequencing corroborates that PCR-tested macrolide sensitive samples do not harbour resistant minority populations at detectable levels. Implications of all the available evidence Our data suggest yaws re-emergence after MDA was driven by multiple sources, and therefore we recommend higher population coverage and intensive post-MDA surveillance. These data show resistance likely evolved once during the trial, rather than being widespread. We recommend careful monitoring of affected communities during post-MDA follow-up in order to rapidly detect new emergence of azithromycin resistance, and consideration for using alternative treatments for cases detected post-MDA.