Effect of repeated mass drug administration on the transmission of yaws: a retrospective genomic epidemiology study.
BACKGROUND: Yaws, a neglected tropical disease caused by Treponema pallidum subspecies pertenue (T p pertenue), has evaded eradication, in part due to a high proportion of asymptomatic cases. Repeated mass drug administration (MDA), whereby an entire population is repeatedly treated irrespective of disease, could provide a solution. Here, we aimed to investigate the effect of MDA on the genomic epidemiology of T p pertenue. METHODS: We conducted a retrospective genomic epidemiology study on samples collected during a cluster-randomised trial of mass administration of azithromycin for yaws eradication in the Namatanai District of Papua New Guinea. Participants were in 38 wards (administrative units encompassing several villages) in three local-level government areas (LLGs). The experimental group received an initial round of MDA followed by two further rounds 6 months and 12 months after the first round. The control group received one round of MDA followed by two rounds of treatment targeting clinical cases and contacts only, on the same schedule as the MDA in the experimental group. A follow-up survey on both groups was done 18 months after the first MDA round. Swab samples were collected at each round from ulcerative and nodular skin lesions, and blood was collected by finger-prick for serological testing at 18 months. Metadata on ulcer size (cm) and duration (days) were recorded at each round, and treponemal and non-treponemal antibodies were recorded at 18 months. Samples from swabs positive for T p pertenue underwent library preparation and whole-genome sequencing. We examined the phylogenetic relationships between genomes, linking them with geospatial and patient metadata to understand the impact of MDA on T p pertenue diversity and transmission. FINDINGS: Swabs collected from 297 individuals with active yaws from April 30, 2018, to Nov 2, 2019, yielded 222 good-quality Tp pertenue genomes. We identified 20 sublineages of T p pertenue in the control group and 21 in the experimental group at the beginning of the study. At the end of the study, there were 13 sublineages in the control group and three in the experimental group, of which two persisted in both groups. Three sublineages not detected at baseline were observed in the control group after commencing MDA. The two sublineages that persisted in both groups had non-synonymous mutations in penicillin-binding proteins. One of these sublineages evolved macrolide resistance in three individuals and was associated with lowered treponemal antibody (p=0·0036) and longer ulcer duration (p=0·015). Despite the study taking place within a small island, sublineages were geographically clustered, with pairs of samples from the same ward (odds ratio 7·1, 95% CI 5·7-8·8; p<0·0001) or neighbouring wards (4·3, 3·3-5·4; p<0·0001) more likely to share the same sublineages compared with pairs from different LLGs. Additionally, older individuals were more likely to share sublineages than were younger individuals (1·5, 1·2-1·9; p<0·0001). INTERPRETATION: Repeated MDA was successful in reducing and maintaining the genetic diversity of T p pertenue at a low level but was associated with the development of macrolide resistance. Yaws re-emergence after MDA was attributed to multiple sublineages, of which the majority were detected in the population before MDA. Participants within the same ward were more likely to share sublineages than those that were more widely geographically separated, suggesting that re-emergence was driven by local transmission. These findings could inform future yaws elimination strategies. FUNDING: European Research Council, EU, Provincial Deputation of Barcelona, Barberà Solidària Foundation, Wellcome, and Fundació "la Caixa".