Specific Problems in Breeding:  Dilemmas for the Breeder



The scenarios below are my attempts to play with the real-world scenarios that real-world breeders get presented with -- as such, they reflect particular dilemmas, not in order of importance or in any other order, but simply as I happened to think about them, either because they are relevant to Cavaliers or because somebody asked me a question or made a comment that got me thinking about a specific topic.  These certainly are not answers for the age.  They just represent my attempts to think about health and other issues at this point in time, and I think best in print.  As somebody or other said -- I've always remembered the quote, but have no idea about the attribution -- "How can I know what I think until I read what I wrote and see what I said?"



A.  What should I do if my otherwise good-quality Cavalier bitch OFAs borderline for hips when she's two?


    1.  Check her with PennHip.  Also, consider waiting six months and then sending in another set of x-rays.  You might find a different result, either because of a real change in hip morphology or because of differences in readers.  OFA radiograph readers aren’t gods.  Three readers checked your radiograph and two out of three ruled that your girl was borderline, but the third reader might have been right when he thought she was “good.”  OFA suggests that bitches not be x-rayed within a few months of their heats -- the hormone flux associated with being in season apparently affects the laxity of the hip joints.  So if your timing was wrong, you probably want to wait a bit and repeat the x-rays.


     2.  Use the data you got from various assessment techniques to assess her true hip structure.  If she has a problem, is it with the depth of her acetabulae?  The roundness of her femoral heads?  The tightness of fit?  Joint laxity?  What precisely?  If  you decide to breed her, or a sibling of hers, you will want to select a male who does not double up on her specific weakness, even if his hips are good overall.


     3.  Assess her sibs (all her sibs – even the pets) and the sibs of her parents.  If she is the only one with borderline hips, everyone else being good, then her genetic quality for this trait is probably acceptable.  If sibs or uncles or aunts have worse hips then she does, she is probably not a good genetic risk for this trait.  Going sideways in the pedigree is a very important step when assessing a polygenic trait.


     4.  Assess her structural quality, her genetic quality with regards to heart, patellas, eyes, and anything else for which you can collect data, and her type and temperament.  Have somebody who’s not a personal friend assess her too, with a clear unbiased judge’s eye.  Is she the kind of bitch that definitely belongs in a breeding program?


     5.  Make your decision about whether to breed her or not on the basis of what’s best for your line and for the breed as a whole.  If you choose to breed her based on your careful assessment of her worth and of the possible risk to her puppies, which may very well be reasonable given the rarity of clinical symptoms for a dysplastic Cavalier, then get preliminary X-rays on all her pups before you sell them, including the pets, so you can accurately track her record as a producer without depending on the pet owners to do the X-rays two years after buying their pups.  The OFA suggests that good prelims strongly correlate with good X-rays at the standard two years (www.offa.org).






B.  What should I do if my outstanding bitch produces a puppy that presents with classic SM symptoms at one year of age?


    1.  It has been suggested that SM is controlled by a two-gene recessive system.  In this case, a homozygous normal dog would be genotype GGZZ, a carrier might be GgZZ or GGZz or GgZz (all with normal phenotypes), and an affected dog would be ggzz.


    This suggestion is actually not consistent with clinical presentation in Cavaliers.  What we are seeing is, if I understand the reports correctly, something like this::

        a) puppy hydrocephalus, which is thought to be an extreme presentation of SM in some cases;

        b) juvenile presentation (6-18 months) of fairly severe symptoms;

        c) adult presentation (after 1 year and up to 6+ years) of moderate or even mild symptoms;

        d) mildly affected parents giving rise to more severely affected offspring.


If the genetic situation was the simple one outlined above, with the trait purely recessive (as is suggested by the abbreviated reports that are available at the moment), then all clinical presentation should be identical, or nearly so -- if not, then something must be creating the variable expression.  What?  There is no suggestion that environmental triggers can do anything to switch expression from minimal all the way over to hydrocephalus.  At minimum, therefore, we cannot have a completely recessive simple trait. 


We might have a situation where there is a two-gene system but dominance is not complete, so that "carriers" are sometimes (often?  always?) showing symptoms themselves.  It's possible, for example, that GGZZ dogs are normal, GGZz dogs also normal, GgZZ dogs minimally affected, GgZz dogs somewhat more severely affected, and ggzz dogs severely or very severely affected.  That would be somewhat consistent with the pattern described above, but would suggest that many moderately- or severely-affected animals should have at least one mildly- or moderately-affected parent.  It also suggests that the absolute numbers of affected Cavaliers should be quite high.  No one, as far as I know, has suggested that this is the case.


One expects, given the normal pathways by which genes are expressed, that a simple one- or two-system is likely for metabolic defects, but not for complicated anatomical traits.  SM counts as a complicated anatomical trait, because it appears to depends on the size and shape of the back of the skull, the size of the foramen magnum, the size of the brain and the cerebellum, and probably various spinal characteristics, among perhaps other things.


It seems to me to be likely that other factors are influencing the expression of this trait, and it seems to me that these 'other factors' are most likely to be, or include, additional genes.  In other words, given the information above, this problem looks to me like it might be more consistent with threshold polygeny.  If this were the case, the genetic situation might be something like this:


A dog with a perfect genotype for this trait is GGCCMMDDSSZZ.  Carriers may carry recessives for any of these genes, so there are a zillion different genotypes for carriers.  Affected dogs must be homozygous recessive for at least two genes (GGccMMDdSSzz, for example.  More homozygous recessive genes in the genotype means more severe presentation and earlier onset.   A hydrocephalus puppy would thus be, perhaps, ggccmmDdsszz.  Juvenile presentation might be GGccMmddSszz.  And adult onset with mild symptoms might be ggCcMmDDssZz.


All affected dogs therefore must tip the threshold (at least two genes homozygous recessive) but then severity of symptoms depends on how many more of the genes involved are also homozygous recessive.  There would be a wide range in possible severity, which is consistent with the situation I understand we are seeing clinically.  In this type of inheritance, most parents could be normal but still contribute deleterious alleles -- perhaps lots of deleterious alleles.


The range of presentation in Cavaliers is also, incidentally, consistent with a trinucleotide repeat disorder.  No such genetic mechanisms have been described in dogs as far as I know, but there's absolutely no reason to believe none exist.  Trinucleiotide repeat disorders usually become more severe as generations pass and, in humans, usually or always show neurological effects (although not in the way SM does -- SM is caused by a more mechanical kind of damage to the spinal cord, compared to the modes of action for most trinucleotide repeat problems).  The best known trinucleotide repeat disorder, Huntington's disease, tends to "look" recessive when it first turns up in a pedigree and then "looks" more dominant as generations pass, although neither term is strictly applicable.  Penetrance also tends to increase through the generations within any given family.  I have no idea whether Cavalier pedigrees are showing any such patterns.  Also, parents with mild or moderate symptoms tend to produce offspring with much more severe symptoms.  This sort of pattern is showing up in the Cavalier pedigrees studied, but of course a simple accumulation of recessive alleles as mildly-affected animals are bred to one another could produce the same effect.


    2.  Taking this for the moment as a threshold polygenic trait, there is every reason to assume that essentially every Cavalier now alive that descends from well-known dogs of the past, which is nearly all of them, carries some of the recessives responsible for the problem.  However, animals from different lines might carry different recessives, and thus be safe to breed together (GGCCmmDDSSZZ would be quite safe to cross to ggCCMMDDSSZZ, because no puppy could double up on recessives for either recessive gene).


If your bitch has produced an affected puppy, she is an obligate carrier (and so is the sire).  Doing any close linebreeding with this bitch would not be a good idea, because her relatives are likely to be carrying the same recessives she is.  However, if she is outstanding, and given that we do not want to constrict the gene pool more than necessary, you might be justified in carefully selecting an outcross sire.  In this case, you would probably want to select an older sire whose breeder you trust when she tells you that as far as she knows he has not produced any SM.  If you bred this bitch to that dog, and carried on her genes with one puppy, then you should have diluted the chance that you carried on affected alleles from her by one-half.  If you did another outcross with the puppy, you should dilute that chance by one-half again, and so forth.


    3.  There's no way of knowing how extensive SM alleles are in Cavaliers or how likely you are to actually get away from the alleles your bitch is clearly carrying by outcrossing.  For a few years, until we know more about the distribution of and mechanisms behind the problem, breeding carriers is going to be particularly risky.


    4.  Some parents of SM-affected puppies apparently go on to develop symptoms later in life.  Your bitch might not be a carrier -- she might in fact be affected, but minimally.  You might want to carefully assess her against a complete list of SM behaviors and signs and try to decide if she looks like she might be absolutely clear of any trace of SM behavior.  If she is affected, breeding her would produce obligate carriers no matter how careful you were.  You are going to need to keep VERY careful records and be absolutely scrupulous in choosing sires for her or her daughters, because it would be disastrous to select a sire carrying the same recessives she is carrying herself.  It would be safest to retire this bitch from your breeding program.  At most you may want to get one outcrossed daughter to carry on her line.  It's clear that you're going to risk producing affected puppies if you breed her.  On the other hand, it's also pretty clear that very few Cavaliers can be assumed to be clear for SM at this point.  We have to breed something.


    5.  Don't hold your breath waiting for a genetic test to become available.  Tests for polygenic traits are exceedingly difficult to develop.  The very best situation -- that this is in fact a two-gene system or that there are only a few "major" genes tipping the threshold switch -- is not necessarily what we are facing.  And even that is harder to deal with than any single-gene system.






C.   What should I do if a thyroid test for my really superb Dalmation bitch shows TGAA antibodies?


       1.  There are divergent opinions on whether lymphocytic thyroiditis, which is the autoimmune form of hypothyroidism in which antibodies are produced that attack thyroglubin, typically progresses to clinical hypothyroidism over the course of time.  If so, this implies that "idiopathic hypothyroidism may in fact be the end-stage of lymphocytic thyroiditis, that lymphocytic thyroiditis is merely a subclinical stage of hypothyroidism.  The pool of data that supports this conclusion is small.  However, the hypothesis makes biochemical sense because it's hard to see how the presence of antibodies against thyroglobulin would not negatively affect thyroid function to at least some degree.  See www.beaconforhealth.org/Addison's-Aronson.htm


    2.  Other sources state that some normal dogs have TGAA antibodies.  There is a shortage of really clear information.  Yes, of course some normal dogs certainly will be positive for antibodies.  The question is, do they stay normal over time?  Apparently the data are not there to answer this question one way or the other.  I e-mailed the author of one particularly good site and it turns out she (a DVM and breeder of Welsh Springer Spaniels) is following an extended family of Springers in order to determine whether anti-body positive animals consistently go on to develop clinical hypothyroidism ( www.artwork.net/cok13/thyroid_study.htm ).  They do seem to, although in one odd twist one dog tested positive for some time and then became negative and has not (yet) developed hypothyroidism.  I can think of a couple of possible explanations for this.  Maybe the dog has one copy of the normal allele and produces some correct protein but not enough to give consistent test results?  However, the complete switch to always positive to always negative is not consistent with this possibility.  Maybe some kind of modifier "switched on" the correct allele later than normal, or switches it on and off for some reason?  Maybe some environmental cue changed?  I'm not a big fan of this last idea, but it could be, I guess.


    3.  Diseases that appear to be linked to hypothyroidism in at least some breeds include diabetes, Addison's disease, and van Willebrand's disease.  You might check this bitch's family tree very carefully for any of these problems.  Addison's apparently can kill fast and leave no tracks for a normal autopsy to find.  I'd suspect Addison's disease in any case in which a 3-5 year old animal suddenly died for no clear reason, particularly if the animal also had a nervous temperament.


    4.  Lymphocytic thyroiditis is "strongly suspected" of being caused by a simple autosomal recessive -- by some researchers.  This is based on the above-cited work with the Springer Spaniels.  Hypothyroidism apparently typically is diagnosed earlier and progresses faster in Welsh Springers than in other breeds, suggesting that it may very well be controlled by a different genetic mechanism in this breed than in others.  Lymphocytic thyroiditis is apparently also suspected of being an autosomal recessive in borzoi and beagles.  However, other sources use the exact same citations to support a polygenic mode of inheritance (!).  This means that the only thing to do is actually get the papers cited and take a personal look, which I have not yet done.  The papers are: 


Conaway DH et al.  1985.  The familial occurance of lymphcytic thyroiditis in borzoi dogs.  Am J Med Gen 22: 409-414.


Benjamen SA et al.  1996.  Associations between lymphocytic thyroiditis, hypothyroidism, and thyroid neoplasia in beagles.  Vet Path 33: 486-494.


Hypothyroidism is apparently thought or known to be an autosomal recessive in humans and mice.  It is definitely possible that multiple different simple mutations all lead to clinical hypothyroidism.


    5.  For the purpose of making decisions in the immediate future, it seems reasonable to guess that the specific problem of lymphocytic thyroiditis in Dalmatians, identified by a positive test for thyroglobulin antibodies, is caused by a single autosomal recessive and that it is an early stage of clinical hypothyroidism.


    6.  That would mean the bitch is homozygous recessive for this trait.  If the bitch is stellar, and given the large proportion of Dalmatians that are hypothyroid, and also given the easily-treated nature of the disease, you may want to breed her anyway (if she was a Cavalier, you might not, because hypothyroidism is rare in Cavaliers and why change that?)


You would probably want to select:  an older male, so his thyroid status can be accurately assessed; a male who has been used fairly extensively for breeding so you can also assess the thyroid status of is progeny; a male who is TGAA negative and who has produced, as far as you can tell, only TGAA-negative offspring.  It would be good to assess him as a proven-clear stud, as you could do, using Padgett's tables, for example.  A stud that has produced no affected puppies after ten litters, for example, is probably clear (assuming a high carrier frequency in Dalmatians; this would not be the case for Cavaliers).  It would be good if there was no Addison's disease, diabetes, or vWD in his family, either.


    7.  You would then treat this bitch's offspring as obligate carriers and breed them as carefully as their mother, to dilute their genetic risk further.  You may want to take a carrier daughter to a TGAA-positive dog to see what you get.  Or there would be other kinds of test-crosses you might do, in order to serve the same purpose (determine the mode of inheritance of this trait for your breed).  If you are sufficiently careful in tracking the descendents of this bitch, and given the late age of onset of hypothyroidism, you should be able to offer a reasonable guess about the mode of inheritance in, say, six to fourteen years.










Inbreeding and the practical breeder --


HWE and the breeding of domestic animals --