The vast majority of research exploring the genetics of DCM has been performed on purebred American Dobermans, a high-risk population for DCM. Even in the Doberman, PDK4 (DCM1) and TTN (DCM2) are incompletely penetrant, meaning that while having one or two copies of these variants is thought to confer some increased risk of developing DCM, it is by no means predictive of disease. DCM is a highly complex disease that is modulated by many genetic factors, most unknown.
While the statistics have not been published, Dr. Kate Meurs, at NC State quotes 37% of Dobermans with the DCM1 (PDK4) mutation will develop disease, 50% with the DCM2 (TTN) mutation will develop disease, and 60% with the DCM1 and DCM2 mutations will go on to develop disease. We know that these statistics correlate more to disease in American lines of Dobermans than European lines of Dobermans (European dogs have a lower rate of clinical disease).
When determining what dogs to breed, there are several factors to take into consideration. Among these are temperament, adherence to the breed standard, genetic variant tests, non-genetic health screens such as an examination by an ophthalmologist and cardiologist, and the health and longevity of the line(s).
As previously mentioned, not only are the two DCM variants incompletely penetrant, there is high variability of predictiveness within lines/populations of Dobermans. Given those facts and the high percentage of Doberman Pinschers with either the DCM1 or DCM2 variants (or both), it is not recommended to remove all dogs with one or two copies of these variants from the breeding population as that will cause a bottleneck of genetic diversity and will likely lead to other health problems.
So, if a Doberman has either the DCM1 or DCM2 variant and is healthy and comes from lines with a low rate of clinically diagnosed of DCM, variant testing alone would not be a good reason to exclude it from a breeding program. However, given the statistics presented by Dr. Meurs, one step toward possibly reducing the clinical risk while aiming to not reduce breed diversity, would be to not breed dogs with the DCM1 variant to dogs with the DCM2 variant as to not produce puppies with both variants.
Additionally, we know that while present in other breeds of dogs, the variants do not seem to be clinically correlated to heart disease in breeds other than Dobermans.
To learn more, please see How do I interpret your results and how do I use them to make breeding decisions?