A systematic review of whole-genome sequencing studies of Mycobacterium bovis that determine the direction of transmission between badgers and cattle

Abstract Background: Tuberculosis, caused by Mycobacterium bovis, is a bacterial disease affecting cattle and other domestic species, wild mammals, and with resilience to survive in the wider environment for many months. Whole-genome sequencing (WGS) methods can distinguish closely related isolates and local transmission clusters. However, the relatively slow evolutionary rate of M. bovis, with substitutions accumulating more slowly than individual transmission events occur, results in limited genetic divergence between epidemiologically linked isolates and a weak temporal signal, greatly limiting the ability to determine precisely ‘who infected whom’ and to estimate transmission rates. Differences in surveillance, sampling intensity, geographic coverage, and detection probability between cattle and the European badger Meles meles warrants a review to examine the strength of evidence, to better inform disease control considerations. Methods: Following PRISMA guidance, we searched PubMed/MEDLINE, Web of Science, Scopus, and Embase for peer-reviewed studies using WGS data from sympatric cattle and European badgers. Studies were assessed according to their sampling design, temporal and geographic coverage, genomic diversity, analytical framework, treatment of unsampled and environmental transmission pathways, and integration of genomic and epidemiological data. Evidence was considered separately for host sharing, local persistence, maintenance, spillover, and transmission direction. Results: The reviewed studies consistently showed close genetic relationships and local sharing of M. bovis strains between cattle and badgers. However, estimates of transmission direction varied substantially between studies with large variation in sampling quality and analytical approaches. Some studies reported patterns compatible with greater badger-to-cattle transmission, while others indicated the opposite, and some emphasized within-species spread and cattle movement. Directional estimates were potentially sensitive to unequal host sampling, differences in surveillance processes, weak temporal signal, incomplete observation of infection chains, the treatment of environmental or unsampled transmission pathways, and especially model assumptions. Close genomic similarity supported shared local transmission histories but generally did not establish direct transmission or its direction. Collectively the evidence base is considered insufficiently rigorous for findings to be definitive and this is recognised by most authors. Conclusions: Current WGS studies demonstrate shared local M. bovis strains and epidemiological connections, but directionality and reservoir status remain highly uncertain and particularly sensitive to sampling and modelling assumptions some of which show signs of confirmation bias. Existing caveated estimates of transmission are presently insufficient to inform generalised conclusions about variation in interspecies transmission rates and their use to validate badger intervention in disease control strategy.