Exciting research paves the way for a genetic test for roaring

Cover image courtesy of The Image Is Everything

A research paper published earlier this year by a team at University College Dublin has made a major breakthrough in identifying genes that raise a horse’s risk of developing recurrent laryngeal neuropathy, the cause for laryngeal hemiplegia (roaring). Their work could be used to develop another tool in early detection of one of the most prevalent performance limiting conditions in Thoroughbreds.

A performance problem

Anyone who has been involved in racing long enough has come across a roarer, the bane of syndicator and yearling consignors alike. The neuropathy is characterised by a progressive loss of nerve function in the (typically) left side of the throat, preventing the horse from getting the full capacity of air into their lungs, particularly at a gallop.

While some horses with recurrent laryngeal neuropathy (RLN) will make a noise, others will not, but will still have their performance affected. Currently, the main option for treatment is unilateral arytenoid lateralisation (a tie-back procedure), where the affected part of the throat is permanently sutured open.

Endoscopy at rest is a routine and expected part of yearling sales, as well as many tried horse sales, but several studies have shown that the throat at rest can look very from in motion and that this isn’t an accurate diagnostic tool for identifying RLN or other disorders.

The gold standard for diagnosis is over ground endoscopy (otherwise known as a dynamic scope) and ultrasound of the laryngeal folds, but this can be an expensive and time-consuming endeavour for trainers and vets alike.

Now researchers at University College Dublin, led by Professors Emmeline Hill and Lisa Katz, have pinpointed a series of genes that their study shows makes a horse three times more likely to develop the problem.

Having spent decades working on performance and welfare limiting conditions in horses, the pair could see a dire need for a simpler, less intrusive way of picking up the disease.

“We recognise that it's at such a high instance in the population, and it appears to be such a problem that it's one of the things they're looking for at the sales,” Hill said. “So that was why we thought, ‘well, if we could develop a genetic test for it, then that would be useful’.”

A robust cohort of subjects

Hill and Katz are building off of a large body of research on the subject already; the disease has already been established as progressive, and is identifiable by a variety of diagnostic tools, and - key to this study - heritable in a high volume of cases. It is that last factor that they focused on.

Working with Irish flat racing trainer Jim Bolger - who is based in Coolcullen, County Carlow - the research team were able to include a high number of horses in a similar environment in their study and limit differences between subjects that could affect the data. Between 2012 and 2016, the team assessed over 200 racehorses to be included in their study.

“We are very lucky, over the years we've been working with a thoroughbred flat racehorse breeder and trainer, and he allowed access to his horses,” Katz shared. “So we knew that we would be able to look at a relatively large number of animals in a similar sort of environment, in which there was reduced variability in terms of management and exercise, and so forth.

“So basically it was considered that the only thing that would vary would be the horse, which is what we wanted from a genetic study.”

With Bolger’s assistance, the team could triage their subjects into either horses with RLN, control horses (no evidence of RLN or similar presenting diseases), and horses to exclude (horses who did not fall wholly into either group).

“In genetic studies, unless you've got thousands and thousands in your study cohort, you have to have a very rigid, stringent definition of your disease and control group to really identify the genetic involvement,” Katz said.

“Other researchers have hypothesised that this is not (caused by) a single genetic mutation. It's going to have multiple, different genetic mutations contributing towards the presence of the disease, so it's not going to be as simple as a ‘yes, no’ test.”

To achieve this, they used the gold standard of a dynamic endoscope, a resting endoscope, and ultrasound of the throat to fit horses into each category, using a very strict set of criteria.

Several studies have demonstrated that endoscopy video assessment is subjective and depends on the experience of the observer, so the same observer was used for every horse. All horses did the same, or similar enough, work when undergoing the dynamic scope, and all horses underwent ultrasound of the cricoarytenoideus lateralis muscle (part of the throat).

“Some researchers consistently found that before you start seeing functional problems in how the larynx works, you can scan that muscle, and you can pick up changes in the muscle that had very high sensitivity and specificity for predicting RLN,” Katz explained.

This was particularly important in establishing a control group of horses that definitely did not have RLN present (at the time of the study). Controls are the baseline to which all test subjects are compared - it is important that they are completely free of disease symptoms in order for the testing to work. Any horse that had RLN symptoms present on one or two points of examination, but not all three, were excluded.

Table: Criteria for sorting horses into study groups.

Significant findings confirm heritability

Once you have established your subject groups, what do you do next? The answer is a genome wide association study (or GWAS), using DNA extracted from blood or hair, that looks for variations within the genes of the horse. Katz and Hill’s team found several common variations in horses confirmed to have RLN that were much less common in their controls.

“We found in this study that the genetic contribution is up to 40 per cent, so this means that genes explain a reasonable portion of the reason for a horse developing RLN,” Hill said.

“We identified genetic markers (associated with RLN) and, by comparing the genetic profiles of the cases to the genetic profiles of the controls, we're looking for genetic markers that are at a significantly higher frequency in the cases compared to the controls. When we find them, they're considered to be risk markers.”

“The highest ranked (genetic variation) that differentiated between the controls and the disease groups was actually found on a gene that is involved in the nerve growth itself,” Katz said. Several other variations had also been previously identified as being associated with nerves - which makes sense, given where the disease develops.

“On their own, they weren't actually screaming ‘I'm the sole cause of this disease’, but when you put (the variations) together in a group, there was a huge difference between the disease and control groups,” Katz continued.

From this, the team could produce a polygenic risk score that identified six individual genetic variations (also referred to as single nucleotide polymorphisms) that, when found together, mean a horse is three times more likely to develop the disease.

Hill elaborated on the risk score; “It’s a test that takes several genes that you've identified to be associated with the trait, then you combine them into a prediction model, which should give you better accuracy at predicting the trait.

“These risk markers then can be potentially developed into screening tools for the disease, so that at an earlier stage, you get a read on the likelihood of a horse developing the disease.

“It's like an early warning system, and this is particularly valuable in a situation where you have a progressive disease, where current methods like scoping don't do very well at predicting future occurrence.”

What happens next?

The goal from here would be to develop a genetic test for use in the wider population - if the industry wants it.

“There has been some suggestion from the breeders associations (in Britain and Ireland) that it is a disease of interest,” Hill said. “If you have a genetic test, you could actually identify these horses that are at higher risk, and they might want to be scoped more often during training and kept an eye on for preclinical signs.

“We know that there are studies that have shown that the earlier the intervention - surgical intervention, and that's part of life - the better the recovery and the better the racing outcome for the horse. So that's the value of the genetic test in partnership with the existing veterinary tools.

“It's not going to absolve the need for veterinary inspection of horses, of course, but it will augment that and provide more power to the vets to advise their clients.”

“As a veterinarian, I'm all about trying to identify diseases early to try to implement treatment,” Katz said. “As more and more research is going into the treatment of RLN, people are coming up with less and less invasive treatments, like trying to stimulate the nerve itself. If you can identify the disorder early, you have a much better potential chance of treatment or early intervention that would be less invasive and taxing for the animal.”

Studies into the effects of nerve stimulation have been pioneered by Cornell University and the Royal Veterinary College in London, and it is promising; the partnership has observed that eight weeks of targeted electrical stimulation to the laryngeal muscle can halt muscle shrinkage and increase the diameter of muscle fibres, restoring function to the throat.

The hope is for clinical trials in the near future to prove its effectiveness as a far less invasive and perhaps more effective alternative to a tie-back procedure.

In the meantime, a larger scale study to confirm the findings would need the cooperation of the industry. A big part of this would involve testing horses that already have RLN to compare to their existing study results.

“It depends on the interest and requirement of the market,” Hill said. “If the market wants a test, we can develop it. We have the methodologies, we have the pipelines to develop something like this quite quickly. It would be a matter of months, not a matter of years (to produce a genetic test).”

Applications beyond racing

Both Hill and Katz think that a risk assessment for RLN would be useful to the breeding world, as a means of reducing the disease’s prevalence in the population. Hill was firm in pointing out that this is not about reducing a horse’s value or discarding him from the breeding population - it’s about making the right mating decision for him and his offspring.

“It would actually be better for that stallion if the mare owners know (the risk) and they are only sending mares that have a low risk to him, so that they are likely to have low risk offspring,” she said. “That would be better for the stallion in terms of his progeny’s performance.”

Katz agreed, saying, “I think it comes down to educating people about this, because I know 100 per cent that there are just superb animals in terms of performance who have this disorder, so I don't at all believe that this means that an animal isn't and will not be a wonderful performer. I'm trying to help the animal become even better and achieve its full potential.

“I think a genetic test like this will help inform people in terms of breeding decisions. Not saying ‘I'm never breeding to that stallion’, but maybe ‘I won't breed this mare to that stallion, I'll breed a different mare to him’.”

Both acknowledge it is probably the largest hurdle to overcome with reducing the occurrence of RLN, but they truly believe the test they could produce is in the best interests of everyone in the thoroughbred industry.

“The mission statement of the Horse Genome Sequencing Project, which was the foundation of equine genetic research, is to improve the health and welfare of horses,” Hill said. “This is one trait that is a performance and welfare limiting factor in thoroughbreds, and therefore having tools to assist with improving welfare outcomes surely speaks to the goals of the industry.”