Choosing a Corgi Breeder: Should I Look for “DM Clear”?

I really thought long and hard about whether or not I wanted to talk about this at all. For one thing, as is the oft-repeated refrain around here, I’m not an expert in anything. But the top reason is that it’s an incredibly touchy subject. It’s a very hard disease to watch a beloved pet go through. In case you haven’t heard of it before, degenerative myelopathy (DM) is a disease that affects the spinal cord. Specifically, it is a breaking down of the myelin—the covering around nerve fibres that helps conduct nerve impulses. It affects the hind limbs, first resulting in weakness and then gradual paralysis. If you have heard of ALS in humans (Lou Gehrig’s disease), it is similar to that. You may have seen pictures and videos of corgis who have special carts for their hind legs, which are paralysed. That is usually DM.

The disease is always fatal, but mercifully it strikes near the end of a dog’s natural lifespan, so the dog may still live to be into the teens.

If you are looking to get a corgi puppy, you will see DM listed in many sources, mentioned in some podcasts, and shouted about on social media as one of the necessary tests that the parents of the puppy should have had. Many will say you should look for a breeder who breeds “DM clears.” You may also see the phrase “triple clear” thrown about, and DM is one of three tests that is meant by that. Obviously, no one wants their dog to go through the horrible disease process that I just described. So that sounds like what you need to look for…right?

But then you go to the Pembroke Welsh Corgi Club of America (PWCCA) website, and DM is not one of the tests that are required in their Code of Ethics for their breeders, and they have an article recommending against the use of the test in breeding decisions. (We’ll break that down in a minute.)

What?! It’s understandably very confusing for those of us trying to navigate buying our first corgi and doing so responsibly.

But for those who have lost a corgi to DM, or to something that at least looked like DM? It is more than that. It is infuriating. It is deeply hurtful, and it sounds like their experience and that of their beloved dog is being dismissed by breeders who care more about their pedigrees and bloodlines than people’s beloved pets. That is where a lot of people are coming from, and that is why this is an incredibly touchy subject that goes beyond facts and science. And that is why I wasn’t sure I wanted to touch this at all.

My Promise to You, and My Request

There are things that can be said of DM which are perfectly logical, especially to those of us who have not been personally touched by this issue, but come off as incredibly cold and cruel to grieving owners, as well as others who have never lost a dog to DM but have seen the effects and are worried about it. “It’s an old age disease,” “all dogs die of something,” “your dog might not even have had DM—did you get a necropsy?” and “DM is painless to the dog.” That last one is the kind of thing that a person might take comfort in, but the way that it is said can be dismissive of the person’s pain at watching their beloved pet deteriorate and having to make big adjustments and constantly evaluate quality of life. Plenty of things that do not cause physical pain are still devastating.

In this post, I am going to try to stay far away from statements like that, and simply present some science, and I am going to try very hard to do so in a sensitive way.

So please, if you identify as someone whose dog has DM, or if you have lost a dog to this disease, please continue reading with an open mind. In return, I want to promise that in this post, I will not argue that DM is not a terrible disease. I will not argue that your experience was not heart-rending and anguishing. I will not argue that health is a less important consideration than any of the other considerations. Please do not read this as me diminishing, dismissing, down-playing, or pretending that I understand your experience.

How is DM Diagnosed?

There is only one way DM can be definitively tested for and diagnosed. There are classic, microscopic changes to the spinal cord which confirm a diagnosis of DM, and these can be found during a necropsy. Unfortunately, this means the only way to tell for sure whether a dog has DM is after the dog has passed away.

While the dog is still alive, a presumptive diagnosis of DM (a strong, educated suspicion) can be made once all other neurological diseases have been ruled out as the cause for the symptoms.

If It Looks Like DM and Walks Like DM…It Could Still Be Something Else

Diseases that can mimic DM include intravertebral disc disease (IVDD), lumbosacral disease, vascular events, and neoplasia. (Neoplasia is the term for various abnormal cell or tissue growths, which can be benign or malignant. Even the benign ones could cause problems by pressing on important stuff…like the spinal cord, for instance.)

To illustrate why distinguishing DM from these other diseases is critical, let’s take IVDD as an example. IVDD absolutely occurs in the corgi breeds, along with many other breeds. It is “a condition where the cushioning discs between the vertebrae (bones) of the spinal column either bulge or burst into the spinal cord space. This is commonly called a herniated disc or slipped disc.” The symptoms can include lameness or weakness in the rear legs, difficulty walking, avoidance of jumping, incontinence, and paralysis—all of which can look very similar to DM. This condition is painful, so the dog may also show signs of being in pain. While there is a genetic component to IVDD, there is currently no genetic test, as more than one gene appears to be involved, and there is also a huge environmental/lifestyle component. IVDD is treatable, and there are a variety of approaches that the vet may take depending on the situation.

Since these other diseases are painful and often treatable, even curable, whereas DM does not have any proven effective treatments, it is important to rule them out before deciding a dog has DM. Unfortunately, vet care is very expensive, and many people simply cannot afford to do a battery of tests on their dog and rule out everything else to land on a disease that is not curable anyway. Especially when the dog is nearing the end of life. So the actual number of DM cases is unknown, and, unless confirmed by necropsy, it is possible that some of the reported cases were actually a different disease. Again, that doesn’t diminish anyone’s experience, and I’m not trying to be dismissive. The amount of people who identify as having or having had a dog with DM is not conclusive as to the actual prevalence of the disease—but their dogs were sick with something, and these other diseases can be just as bad or worse.

What About Genetic Testing?

Okay, but what about that DM test everyone is talking about? The test that you should see before you buy a puppy? The test that will guarantee your dog never develops DM?

What we’re talking about here is genetic testing, and we have to understand how that actually works. I am obviously, clearly, not a geneticist. I’m not even someone who has a good grasp on genetics. So I’m going to explain it in terms that I myself can understand, and I’m sorry to anyone reading this who actually knows about genetics (yeah, I’m talking about you! You know who you are) because it’s probably going to be cringey. But here we go. What’s a genetic test?

What’s In A Gene?

Let’s go back to high school biology class for a minute. Every cell in your body contains DNA molecules that contain the instructions for everything. Every piece and process that goes on in your body. Along that DNA molecule, there are certain segments that are known as genes. These are heritable bits of information made up of two alleles—one from each of your parents. Alleles are variants of the gene. The pairs of alleles that you have inside your genetic code make up your genotype. Your phenotype is the alleles that are expressed—the ones that show outwardly, as it were. You can carry genes in your genotype that you can’t see by looking at your phenotype.

Still with me? Cool. If you inherit the same allele (same version of a gene) from both of your parents, you are homozygous for that gene. If you inherit different alleles (different versions of that gene) from each parent, you are heterozygous for that gene. If you are homozygous, then there is no question which allele will show up in your phenotype. There is only one option there. But if you are heterozygous, then which allele will it be? Well, that is where dominance and recessiveness come in. In the case of heterozygosity, the dominant gene will be expressed in the phenotype. But you still carry the recessive gene in your genotype. And you could pass it on.

The classic, easy-to-understand example is human eye colour, which used to be thought of as a single-gene trait but now apparently turns out not to be? But I’m just going to pretend it’s a simple one-gene trait to help us all grasp this, okay? Cool. So brown eyes are the most dominant, and blue eyes are recessive. Let’s say your mom has Brown eyes (B) and your dad has blue eyes (b). You get one copy of Brown and one copy of blue. So your genotype is Bb. Your phenotype is Brown eyes. But you still carry that blue eye gene. So you have a child with someone who has blue eyes—or maybe, like you, someone who outwardly has Brown eyes but whose genotype is Bb. It’s possible for your baby to have blue eyes—if their genotype is bb. Make sense?

But, as the fact that eye colour is apparently no longer just one gene illustrates, it’s way more complicated when you’re dealing with very complex organisms with traits that involve more than one gene working together.

There is also the fact that trait inheritance is complex—genes don’t exist in tidy little packages. They often pull other genes with them. So in the example where you and the other parent are both Bb, what determines whether the resulting child is BB, Bb, Bb, or bb? Is it an equal 25% chance? More and more, geneticists are saying it’s not that simple. The allele you get for this gene might depend on the allele you get of a different, maybe even seemingly unrelated gene. Does that make sense?

So it’s not a separate dice roll for each gene, going down the DNA strand neatly choosing which alleles you get, completely separate from everything else. Roll the dice: brown hair. Roll the dice: brown eyes. Roll the dice: oof—you got grandpa’s nose! Roll the dice: freckles. Instead, it’s a tangled web where this gene affects that one which in turn affects these ones. So maybe it’s more like, roll the dice: brown hair—ope, looks like you’re getting that freckle allele then, which also means you’re getting grandpa’s nose, but only because you also got brown eyes together with the freckles, because if you had instead gotten blue eyes… I don’t know enough about it even to provide accurate examples, so that’s all made up. But I think it’s like that, just super complicated and connected in ways that no one has a complete grasp on yet.

When Genes Make You Sick

If you have one healthy copy but one diseased copy, (heterozygous), you will still never have the disease. However, you are a carrier, and you could either pass a healthy copy or a diseased copy to your offspring. If you have two diseased copies (homozygous for the mutation) then you are at risk.

Does that mean you will have the disease? Do two mutated alleles always cause disease? Well, once again, it gets complicated.

Identifying The Culprits

First of all, how are genetic tests even developed? How do you figure out that this gene causes this disease? All of us have genetic mutations. Not all of these cause disease—more on this in a minute. So how do you link a gene to a disease?

Like most things in science: painstakingly, systematically, one tiny step at a time. I’m trying to get to the point here, so here is a Wired article that summarizes it well. With monogenic conditions such as Huntington’s disease, you’re simply finding the one faulty sequence that all people with that disease share, and which no one without that disease has. The actual work to pinpoint that must be, as I said, painstaking. But as a concept, that’s relatively simple.

As you can imagine, though, it’s rarely that simple, as that Wired article goes on to explain. Most diseases are not monogenic—even the ones where there is obviously a hereditary component. Sure, many diseases can be linked to a gene or a few genes, and companies will be quick to sell you a test for that gene (with the disclaimer that it is only partially predictive, of course) and call it a genetic breast cancer test, for example. But that gene that they are testing for is recognised to only be one piece of the puzzle. Because unlike with the monogenic diseases, these mutations are present in a lot of people who will never get the disease. Obviously, the actual disease risk must be determined by that gene in combination with other genes, but those haven’t been identified yet. And more confounding still, environmental factors likely play a role, and that is much more difficult to suss out.

But the fact that a genetic mutation is only partly predictive of the disease, with the rest of the puzzle still being a blank void, will never discourage a company from selling you a test. They will just slap a disclaimer on there, reminding you that the test is not diagnostic or definitive. And then they’ll label it as a Breast Cancer Risk Test, even though a more accurate name would be “test for mutations in xyz genes that have been found in a high percentage of people with breast cancer, though not exclusively in people with breast cancer, and are currently being researched, but on their own are clearly not causative.” Of course, that’s a lot less catchy.

Bringing It Back to Dogs

Okay. That little science lesson from the least-qualified person in the room is over now. Just as in humans, some diseases in dogs are monogenic. For instance, von Willebrand’s disease (vWD), which is a bleeding disorder that your corgi’s parents should be tested for, is an example of this. But many diseases are a complex amalgam of environmental and genetic factors, and it’s not clear exactly which genes are involved. Where genes have been identified as playing a role, it is clear they are only one piece of the puzzle.

The Sod1 Mutation

In 2008, a big step forward in DM research was achieved. Here is a link to the article in the Proceedings of the National Academy of Sciences if you want to dig in, but I will try to summarize. Drs. Tomoyuki Awano, Gary Johnson, Claire Wade, Martin Katz, Gayle Johnson, Jeremy Taylor, Michele Perloski, Tara Biagi, Izabella Baranowska, Sam Long, Phillip March, Natasha Olby, G. Diane Shelton, Shahnawaz Kahn, Dennis O’Brien, Kerstin Lindblad-Toh, and Joan Coates (sorry for the long list, I just wanted to give them all credit!) performed Genome-wide association (GWA) mapping, basically looking for a gene that was associated with dogs who had DM. They found something on the gene called Sod1, and this was bolstered by clinical similarities between DM and the human disease ALS. Mutations in Sod1 account for around 20% of familial ALS cases and around 5% of sporadic cases (ones without family history.) So combined with the GWA mapping, this looked promising.

Looking at the Sod1 gene, they found a mutation called E40K missense mutation. Homozygosity for this mutation was present in all of the dogs they tested who were symptomatic for DM, out of a sample size of 38 corgis. They all had two mutated alleles making up this gene. It is important to note that of the asymptomatic control group, 10 out of 17 were also homozygous for the mutation. Even so, there was enough evidence to strongly link the E40K mutation with the disease. And they found two copies of the mutation in DM symptomatic dogs of the other breeds they studied, not just corgis, so this appears to hold true across all the breeds which experience DM.

However. They take care to state: “All of the strictly diagnosed DM affected dogs were A/A [that is, the E40K missense mutation of Sod1] homozygotes; however, several of the aged A/A homozygotes were symptom free. Thus, DM appears to be an incompletely penetrant autosomal recessive disease.” (italics mine.) In other words, many dogs who had two mutated copies of the gene, despite being of the age where symptoms start to appear, did not have DM. This means that the mutation they identified is not entirely predictive. Here is an in-depth explanation of incomplete penetrance. In part, it explains: “Reduced penetrance is likely to be a consequence of the combination of a variety of different genetic and environmental factors.”

Bottom line: If we are seeing a significant percentage of the population which have the mutation but do not have the disease, it is evident there are other factors at play. And that is what scientists found with this sod1 mutation.

The Problem

The problem is that sod1 is a very, very ancient mutation in the dog genome. A huge percentage of both purebred and mixed-breed dogs are either heterozygous (carriers) or homozygous (at risk) for the sod1 mutation thought to cause DM. In fact, it is the most frequent mutation identified in genetic screening. Some breeds with high percentages of this mutation do not have one single documented case of DM disease. Despite the prevalence of the mutation, there are only a handful of breeds which seem prone to DM, including the pemb. So, while identifying the sod1 mutation was a big step forward in DM research, it is only part of the puzzle, and the scientists who discovered it never claimed it was the whole puzzle. That’s not how science works.

But genetic testing companies were quick to market the test as a “DM Test.” Since the mutation these tests are looking for is not on its own predictive of actual DM risk, it is not really accurate to call it a DM test, any more than it is accurate for those human tests to be called a breast cancer test. It’s not how science works, but that is how companies work. And it’s up to consumers to look at these things with a critical eye.

Now Let’s Talk Corgis

Okay, but we aren’t talking about a breed where the mutation is present but there has never been a case of actual DM. We know Pembroke Welsh corgis do get the disease, and we know that every single one of the symptomatic corgis tested in that study out of 38 did have this mutation. So isn’t that enough reason to only breed clears and carriers, just to be on the safe side? Isn’t it worth throwing some otherwise good dogs out of the gene pool if it could save families from pain?

Maybe, but to make that decision we would have to know the whole picture. Because genetic traits are so intricately and inextricably linked, because you cannot simply take out one gene as though it exists in a vacuum, we don’t know the whole picture. We don’t know what problems would rear up if only clears and carriers were bred. What we do know is that some of the best dogs structurally, well-put-together so that they don’t break down but can move, and move, and move as a corgi should, are “at risk” according to the DM test. However, none of their close relatives or recent ancestors had DM, despite having the gene.

This is not cosmetics we are talking about. This is not prioritizing beauty over health. We’re talking about structure. Poor structure can cause dogs to blow out their cranial cruciate ligaments, which is very painful, and the surgery for which is very expensive. Being built incorrectly can put pressure on the spine, causing disc problems which, as mentioned earlier, can look like DM but which unlike DM are painful to the dog. Poor structure can result in lameness much earlier in life than the symptoms of DM would show up.

How many dogs are we talking about eliminating? More than half of the Pembroke Welsh corgi gene pool test as homozygous “at risk” for the sod1 mutation linked to DM. But on average, only 1.1% of those at risk dogs will ever develop DM.

Well, so what. Why not just breed those at risk dogs to clears? Then their offspring will be carriers and we can slowly eliminate the sod1 mutation from the gene pool, right?

No. For one thing, only around 12.2% of Pembroke Welsh corgis were found to be clear. To breed all dogs to only those 12.2% would create a bottleneck. When you have a large gene pool, you can have all sorts of recessive mutations that rarely come to our attention, because the odds of two carriers producing offspring just aren’t that high. But in a small gene pool, those odds increase. You eliminated affecteds and most carriers of this one mutation, but now that’s all you have of another, maybe more deadly mutation. And now you have lost the genes that allow you to get yourself out of that corner.

Secondly, we don’t know what else that gene is linked to. The fact that some of the most perfect examples of the breed and the healthiest dogs carry two mutated copies of this gene should set off alarm bells if you’re talking about completely eliminating it. Why is this ancient mutation so widespread across breeds and mixed breeds? Is it linked with something beneficial that we would not want to lose?

What if the mutated sod1 gene were found to be linked to a lower risk of hip dysplasia, or cancer, or something else? What if dogs with this mutation live longer, or are less susceptible to certain viruses or parasites? What if, what if. We just don’t know. But until we find out, we shouldn’t talk about eliminating genes.

The Role of Testing

Genetic testing is a tool. It is important to look into these things, and it allows us to do so. Sometimes, it allows us to easily ensure that none of the puppies in a litter will get a certain disease, and that’s fantastic. But often, it’s not that simple. To sum up, this is what Dr. Jerold Bell concludes:

Due to the high frequency of the sod1 mutation but the low frequency of clinical disease in the breed, it is recommended the Pembroke Welsh Corgi breeders not use the results of the sod1 mutation in making breeding decisions unless there are confirmed DM affected close relatives (and therefore a high risk of carrying the other required mutation(s) for clinical disease). 

87.8% of Pembroke Welsh Corgis test either as heterozygous carrier or homozygous “at risk” for the sod1 mutation. If breeders attempt to eliminate the sod1 mutation or only breed to the 12.2% of dogs not carrying the mutation, this will cause a severe and possibly disastrous restriction of the breed gene pool and breed genetic diversity. 

Canine Degenerative Myelopathy and Genetic Testing in Pembroke Welsh Corgis by Jerold S. Bell, DVM

Just as family history of cancer in combination with genetic testing can give you a better picture of your overall risk (though unfortunately, still not the whole picture), breeders can study the pedigrees of their dogs and see that, although this dog has two copies of the mutated sod1 gene, there is not a family history of DM. Does this line lack some other gene that is necessary in order for the homozygous sod1 mutation dog to get DM? Do they possess a different gene that prevents them from getting DM? Are there environmental factors also at play? We just don’t know, but educated guesses can be made.

Continuing in this vein: