Retrospective analysis of the epidemiology and clinical presentation of West Nile virus infection in horses in South Africa, 2016 - 2017

Abstract

West Nile virus (WNV) has gained international attention in recent years as a globally emerging disease, particularly after large epidemics occurred in North America in the past 20 years. Although endemic to South Africa, it has only been recognised as a significant cause of neurological disease in either humans or horses since 2008. This retrospective study provides an epidemiological and clinical description of WNV disease in horses in South Africa during 2016–2017, when 54 cases, most of which occurred during 2017, were diagnosed by passive surveillance at the Centre for Viral Zoonoses (CVZ), University of Pretoria. Cases were followed up and then statistically compared to a randomly selected set of 120 WNV-negative controls from the CVZ database of the same time period, which complied with similar case descriptions. Clinical presentation of WNV cases was found to be remarkably similar to international trends, with 89% neuroinvasive disease and 39% case fatality rate, mostly displaying typical, significant neurological signs: ataxia (74%), hindleg paralysis (35%), paresis (30%), total paralysis (28%), tremors / muscle fasciculations (19%), foreleg paralysis (17%) and laminitic stance (9%). Approximately half of the cases exhibited pyrexia. Cases that had only neurological signs were more likely to die while cases with pyrexia, with or without neurological signs, were more likely to recover. Most of the cases were in Thoroughbred, Warmblood or Arabian horses, while local or mixed breed horses were the least represented. Cases occurred mostly in WNVunvaccinated horses less than 5 years old, specifically in the late summer and autumn months after heavy rain in the temperate to warm Eastern parts of South Africa. Cases were located mainly in Gauteng, KwaZulu-Natal Midlands and the Northern Cape with fewer cases in the Free State and Western Cape provinces. In the multivariable logistic regression analysis, the odds of WNV infection was associated with season (higher during March-April vs. all other times), altitude (higher at 1293–1466 m vs. other categories), breed (lowest in mixed and local breeds), younger age and failure to vaccinate against WNV. Based on these findings, risk-based recommendations may be made to horse owners; in particular, vaccination against WNV, which is currently the most effective prophylactic measure available to reduce disease, severity of clinical signs and mortality.