Periodically, I check on the progress of research on DM, since it has such bearing on our beloved Corgis, especially Pems.   

The summary of the research for those who don't want to read the whole long post are this:  

The good news:  Modifier genes have been found which may determine whether or not At Risk dogs become symptomatic.  

The bad news:  Two genetically Clear dogs and several genetically Carrier dogs have been confirmed to have DM by necropsy, meaning no one can now say for certain that only breeding Clear dogs will eliminate the risk of DM.  (This breeding option was never considered smart for all Corgis, but was a viable option in many other breeds)

For some previous discussions on DM here at MyCorgi, check:

http://mycorgi.com/forum/topics/degenerative-myelopathy-2

http://mycorgi.com/forum/topics/dm-question?id=1150197%3ATopic%3A14...

The primer (those who have followed the research can skip this bit and go right to "New Developments"):

DM is a progressive degenerative disease of the spine, similar to ALS in humans.  It is always fatal, though many Corgis are elderly when they do get it and so they may die of other causes first.  While technically painless, it can be stressful for the dog and the owner, since the dog loses feeling and function beginning in his hind end and working forward.  It is relatively common in Corgis, though the overall incidence seems to be in the low single percents.

http://www.caninegeneticdiseases.net/DM/basicDM.htm

There is a simple recessive genetic mutation that highly correlates with DM in dogs.   The GG allele is called "Clear" and is not (or was not) associated with the disease;  a dog with GG can only pass on the "good" G gene.

The AA dog is considered "At Risk" of developing the disease and has two copies of the "bad" allele.  However, while most dogs who get DM are AA, most dogs (and this is important) who have the AA presentation do NOT get DM.  So unlike vWD, where dogs with two copies always have the disease, something else must be going on.  Either there are modifier genes (now seen as likely) or environmental factors, or both.

The AG dog is a "carrier" and should not get the disease (usually) but can pass on either the A allele or the G allele to offspring.

Incidence of carrier and at risk dogs is very high in Corgis, with only about 11% of Corgis showing as Clear.  This has made many breeders cautious about using the test (from what I've heard) until more information becomes available.  

http://www.offa.org/stats_dna.html?dnatest=DM

To complicate matters, DM is a disease diagnosed by symptoms and elimination, and many other spinal problems can mimic DM.  Like Alzheimer's, the only way to firmly diagnose is upon necropsy. 

New Developments:

As an aside, from what I know of complex genetic disease in humans, I was initially skeptical that DM was just a simple recessive, and the researchers seemed from the beginning to feel it was polygenic.  For starters, too many At Risk dogs never get sick.  Secondly, in some breeds age of onset is around 7, in others not til 11 or 12.  A gene that modified age of onset is likely;  it is possible that ALL At Risk dogs would get sick if they lived long enough, for instance.  

Here's what's new:

Predictably, some modifier genes have been found.  Research is ongoing:

http://www.akcchf.org/research/funded-research/1372.html

"Three loci on chromosomes 25, 27 and 37 were significantly associated with ALS in PWC. The two loci on chromosomes 27 and 37 are well defined and each contains an excellent candidate gene expressed in the brain."

Modifier genes might be more useful in breeding decisions, or it may turn out there are so many in such complex combinations that they prove to be difficult to test for.

From 

http://www.chessieinfo.net/degenerative-myelopathy.htm

“The inheritance mode(s) of the other gene(s) involved with DM are unknown at this time, but current information suggests that another gene may be required to create symptoms of disease. This gene is also suspected to be a simple recessive, which may be a "switch" which turns on the expression of the disease, causing symptomatic DM."

If there is indeed a simple recessive "switch" that may be very helpful.  You can see how hard it would be to avoid DM just by knowing which dogs in which lines had symptoms, if indeed there are two separate recessive genes needed to express the disease.  The possible genetic combinations even within the same letter would be vast.

The second development is a bit more troubling:  Necropsy has now confirmed cases of DM in two dogs that are Clear, and several that are Carrier, which clouds the initial genetic picture.

http://www.offa.org/dnatesting/dmexplanation.html

"Among the hundreds of dogs studied so far at the University of Missouri, only two dogs with test results of N/N (Normal) have been confirmed to have DM.

Carriers are far less likely to develop DM, but we have confirmed DM in a few carrier dogs.

Although almost all dogs in the research study with confirmed DM have had A/A DNA test results, recent evidence suggest that there are other causes of DM in some breeds."

Some more info from the Chessie site, linked above:

"One of the dogs tested Carrier but confirmed as having DM was a Chesapeake Bay Retriever, which may indicate that we have more than one form of the disease in our breed."

From the research link above:

"We identified a new SOD1 mutation in Bernese Mountain Dogs and currently are investigating a GSD with DM for a mutation in a different gene."

And the most important thing to remember right now, considering the high incidence of genetically At Risk dogs in our breed:

"Dogs are listed as At Risk, rather than Affected, because having this gene alone does not cause the disease. Greater than 90% of Fox Terriers have tested At Risk or Carrier, yet there has never been a single case of DM reported within that breed, which lives a very long time."

So there it is in a nutshell, with more research ongoing.  I know that in a genetic disease of humans that I follow closely for personal reasons, there have been dozens of genes uncovered that seem to play a role, with a clear understanding not yet reached on how the genes interplay to determine who gets the disease and who does not;  age of onset;  triggers; and severity.  The same may well hold true for this disease in dogs.  

In the meantime, all we can do is be educated and wait.  I would personally caution anyone to be careful of advice which seems to offer a black-and-white answer for the role of breeders.  For now, it would seem that those who have dogs with DM in their lines will be making different breeding choices than those who don't have DM in their lines (meaning dogs who actually got sick, not dogs who carry the gene).   What is right for one breeder will be wrong for another.  There are no easy answers to this situation.  It would be nice if we had a more definitive genetic test out there, but we perhaps should not expect one right away.