Clinical genetics is a rapidly evolving field. Genetic screening has already shown valuable clinical utility, complementing existing best practices and strengthening the doctor–client–patient bond. The way veterinarians interpret and communicate these results, however, can involve additional considerations for both the clinician and the client.
As we learn more about how genetic variants relate to health outcomes, the way we classify results and describe risk will inevitably change to reflect a growing understanding of how complex genetic factors can influence health.
Types of genetic health results
Embark uses the terms clear, carrier, and at-risk to classify genetic results.
- Clear means that the dog does not have the variant allele.
- Carrier means that the dog has one copy of the variant allele and is not expected to have clinical signs of the disease. The term “carrier” only applies to conditions that are inherited recessively.
- At-risk means that the dog has one copy (dominant) or two copies (dominant or recessive) of the variant allele. We use the term at-risk because genetic testing does not diagnose disease. Other factors will determine if the at-risk dog becomes clinically affected.
If a dog is called clear for all tested genetic health risks, that does not mean that the dog is clear of all genetic diseases. Embark tests for several conditions where multiple genetic mutations cause the same or similar phenotype. For example, Embark tests for 20+ variants associated with Progressive Retinal Atrophy (PRA), but Embark does not test for every known variant. Moreover, there are almost certainly more variants that remain to be discovered. Mutations consistently occur (at a universally constant rate). This is a challenge to stay on top of genetic health considerations in dogs. It is also an opportunity to gain a better understanding of the etiology of complex diseases iteratively in a way that frames the future of diagnostics and drug development.
Veterinarians must consider dogs holistically when interpreting genetic results. Breed, age, sex, weight, diet, environment, and concurrent medical conditions, among other variables, can and should be taken into account. Complete physical examinations with an emphasis on the potentially affected organ systems should be performed. Further, not all breed-relevant conditions currently have a genetic test. Veterinarians must compose complete lists of differentials and perform diagnostic tests in conjunction with genetic screening.
The current terminology of clear, carrier, and at-risk came from a mostly “Mendelian” context in genetics; that is, a simple correlation between one genotype and one phenotype. As we (both at Embark and the field of canine genetics at large) increasingly resolve complex traits and diseases, this terminology will necessarily change.
In the future, this method of classifying genetic testing results will evolve to encompass polygenic risk, which is essential to understanding complex traits and diseases. Polygenic risk scores represent a total view of multiple variants involved in an individual’s heritable risk for developing a disease. As you can see, the way we classify genetic results—and the terms we use to describe risk—will continue to change as we better understand the multiple risk factors that can contribute to a phenotype. At Embark, we are working to restructure the way that complex information is delivered and contextualized.
Similarly, breeders and pet parents will be presented with new kinds of models to integrate in their breeding and wellness programs, respectively. This is new territory for all stakeholders in canine genetic health.
Communicating risk: penetrance and expressivity
The language of prognostic genetic testing includes a way to communicate an all-or-none probability that a dog will experience a given disease state, known as penetrance. A penetrance of 1 indicates a dog with the genotype will develop clinical signs in its lifetime with certainty. A penetrance of 0.5 indicates a dog with the genotype has a 50-50 chance of developing the disease state. Incomplete penetrance is the result of unknown genetic or environmental factors that modify outcome, or a stochastic effect of developmental biology. Future discoveries can improve the understanding of incomplete penetrance, leading to clearer predictions and recommendations.
As an example, MDR1 Drug Sensitivity is considered completely penetrant. Every canine with the variant will have an increased sensitivity to certain medications. However, Degenerative Myelopathy is considered incompletely penetrant. Most German Shepherds (GSDs) with the variant will develop signs of DM during their lifetime. However, most Yorkies with the variant will never develop signs of DM in their lifetime. For DM, the penetrance of the variant is breed-specific, which strongly suggests that there are other genetic factors influencing the outcome.
In contrast to penetrance, expressivity describes the variability of the disease state. Expressivity can reflect several types of variation:
- Variation in clinical presentation (severity) between dogs with the same genotype
- Age of onset when clinical signs first manifest
- Differential response to treatment
With Degenerative Myelopathy, there is variable expressivity for age of onset. For GSDs where the penetrance is high (meaning the dog will likely be affected by DM), clinical signs can develop across a range of age, from 8–12 years. Expressivity, like penetrance, may be influenced by many different genetic and environmental factors.
Continuously evolving process
Each of the prognostic genetic tests Embark offers should be viewed as a model—a simplification of reality. The current models take genotype and breed into account when providing a prediction of risk.
To improve models in the future, we must pull in empirical data on outcomes that are capable of supporting or refuting a model’s prediction. This process is simple and relatively straightforward for Mendelian models (a single outlier can break and inform the model). However, the process is enormously challenging for complex diseases (thousands of outliers may be needed to re-train the model).
Looking at examples of each model can help illustrate how the process differs between the two.
MDR1 Drug Sensitivity is an example of a simple model. Dogs with one or two copies of the variant will have a proportional change in their ability to tolerate certain drug dosages. This does not appear to vary based on age, sex, or breed. If there was an outlier, it would likely be easy to detect and simple to understand the biological basis for the outlier.
Degenerative Myelopathy is a more complex model. DM will require a much more elaborate data feedback loop to improve this model’s predictiveness. This model may need to take into account the genotype of multiple variants, breed, age, sex, nutrition, and lifestyle to accurately make a prediction. Outliers that do not have the predicted clinical outcome inform the evolution of the model. We can improve and rebuild the model when we understand why the outliers had the unexpected outcome. Fortunately, Embark is well-equipped to do the necessary work to train and retrain these models, due to our proactive communication with pet parents and veterinarians managing at-risk dogs.
Improving predictiveness
To improve the predictiveness and clinical utility of genetic screening, we need large volumes of data on both genotype and phenotype. This type of data will help us train the models as described above, and incorporate polygenic risk into the predictive models for genetic results.
Embark is uniquely positioned to conduct this type of needed research, through annual health and targeted surveys that connect genotype to phenotype. Epidemiological studies aid in understanding the distribution and breed relevance of health conditions. Modifiers of genetic disease and new discoveries can be found using our natural cohort of dogs as they traverse through life.
Partnering with practicing veterinarians provides us with invaluable clinical insight into how genotype relates to phenotype. By building models to better predict and protect the health of dogs, Embark is poised to increase the longevity and quality of life of dogs and their humans.
