Phenotypic and genetic characterisation of disease subtypes of equine recurrent exertional rhabdomyolysis

Application
Identification of clinical subtypes of RER improves our ability to study the genetic basis and aetiopathogenesis of the disease, as well as seek potential treatments. This improves the welfare of the animals as well as having an impact on the huge financial losses suffered by the racing industry due to RER.

Introduction
Equine recurrent exertional rhabdomyolysis (RER) is a myopathy characterised by exercise-induced episodes of myofibre necrosis. RER is clinically heterogenous, with variation between clinical signs and severity of those signs between individuals and between episodes, with severe cases resulting in recumbency, kidney failure and even death. RER is a moderately heritable disease (h2 = 0.34–0.49) (Norton et al., 2016), but causal genetic variants have not yet been identified. Previous studies in another exertional myopathy, polysaccharide storage myopathy, identified disease subtypes with differing genetic basis (McCue et al., 2008, McCue et al., 2009). We aimed to identify possible disease subtypes within the RER syndrome.

Materials and methods
We applied principal components analysis, k-means clustering and hierarchical clustering methods to 209 RER horses retrospectively from the Royal Veterinary College’s Comparative Neuromuscular Diseases Laboratory diagnostic biopsy service, to identify clusters of cases based on clinical history and histological records. We then tested variables for association with clusters using Chi-square testing.

A support vector machine algorithm was trained successfully to differentiate between classic and non-classic RER cases, and this algorithm was applied to a dataset of 127 Warmblood horses (WB) and Connemara ponies (CP) (CP: 16 cases, 17 controls; WB: 50 cases, 44 controls). Genome-wide association studies (GWAS) were carried out within and across breed and disease subtype.

Using 69 samples of equine semimembranosus muscle from Arabian, Thoroughbred and WB horses (24 classic RER, 12 non-classic RER, 26 negative controls, and 7 with myofibrillar abnormalities), we carried out a blinded histological study comparing features identified by a specialist between disease groups from a standard panel of myopathy staining and immunohistochemistry, including for desmin. We selected features with p 
Results
Consistently, horses grouped into ‘classic’ and ‘non-classic’ RER subtypes: cases with no particular defining features, versus cases associated with gait abnormalities (p 
Conclusions
Overall, RER is a complex disease, likely consisting of multiple disease subtypes with possible distinct genetic associations.

Acknowledgments
This work was funded by the RVC Mellon Fund for Equine Research. We also wish to thank the referring vets and owners for submitting samples.