The Effect of Repeated Mass Drug Administration on the Transmission of Yaws: A Genomic Epidemiology Study

BackgroundYaws, a neglected tropical disease caused by Treponema pallidum subspecies pertenue, has evaded eradication, in part due to a high proportion of asymptomatic cases. A cluster-randomised trial in a yaws endemic area compared repeated mass drug administration (MDA) with one round of MDA followed by targeted treatment. Repeated rounds of MDA reduced active and latent prevalence of yaws but led to emergence and spread of azithromycin resistance to three children. Here we aimed to finely delineate the dynamics of T. pertenue sub-lineages over the course of this trial. MethodsWe performed whole genome sequencing directly on DNA from 263 swabs collected during this trial, recovering 222 good-quality T. pertenue genomes. We examined the phylogenetic relationships between genomes linked to geospatial and patient metadata. FindingsWe identified 29 fine-scale sub-lineages of T. pertenue, of which ten were eliminated by MDA, whilst 13 persisted in the control arm, one in the experimental arm, two in both study arms, and three were first observed after commencing MDA. The two persistent sub-lineages had non-synonymous mutations in penicillin binding proteins. One of these sub-lineages evolved macrolide resistance (N=3), and was associated with lowered treponemal antibody levels (p = 0.004) and longer ulcer duration (p = 0.015). Despite the study taking place within a relatively small geographic area (Namatanai District, in the Island of New Ireland, Papua New Guinea) sub-lineages were geographically clustered, and older children were more likely to share sub-lineages (p = 6x10-9). InterpretationOur findings show that the re-emergence of yaws after MDA was attributed to multiple sub-lineages. The majority of these sub-lineages were detected in the population prior to MDA, and participants were more likely to share sub-lineages within the same ward, suggesting that re-emergence was mostly driven by local transmission. These findings could help inform future yaws elimination strategies. FundingEuropean Research Council, European Union, Provincial Deputation of Barcelona, Barbera Solidaria Foundation and Wellcome. Research in contextO_ST_ABSEvidence before this studyC_ST_ABSWe searched PubMed on 24th July 2024 using the terms treponema * pertenue OR yaws, genom* OR sequencing, and "mass drug administration" OR azithromycin OR "mass treatment", without restrictions for language or date. Two studies were previously published on the current cohort from the Namatanai Province, Papua New Guinea, comparing three rounds of mass drug administration (MDA) with one round of MDA followed by targeted treatment. These studies used multi-locus sequence-typing and found that repeated MDA limited yaws to one sequence-type but resulted in three cases of macrolide resistance. A separate study used whole genome sequencing to find that after a single round of MDA on Lihir Island, Papua New Guinea, a rebound in yaws cases was caused by multiple sub-lineages of the same MLST, but evolution of macrolide resistance only occurred once. No studies have yet examined how repeated MDA affects the whole-genome diversity and evolution of yaws. Added value of this studyOur findings show that re-emergence of yaws after MDA was caused by multiple sub-lineages, most of which were already present in the population before MDA. In the group undergoing three rounds of MDA there was still a small rebound in cases six months after the third round, caused by two "persistent" sub-lineages with mutations in penicillin-binding proteins. One of these sub-lineages was associated with lower treponemal antibody and developed macrolide resistance. Sub-lineages were more likely to be shared between older participants and those in close geographical proximity. Implications of all the available evidenceThese data suggest that re-emergence is predominantly driven by cases missed by the initial round of MDA rather than by importation of new cases. Much more efficient population suppression was achieved using three rounds of MDA, and this is likely due to more comprehensively treating the population and eliminating latent cases. Geographical clustering of sub-lineages suggests that elimination by maintaining cases at a low enough prevalence to result in stochastic "fade out" could be feasible and achieved by repeated mass drug administration. Transmission was found to be most common amongst older children, and targeted approaches focusing on these groups may be beneficial. However, ongoing surveillance for macrolide resistance will be needed to achieve eradication.