A new canine hip scheme for breeders?
A proposed new canine hip scheme was launched at a meeting in mid-January, with some promotional material on Facebook and an article in Our Dogs (January 27) entitled ‘New hope for breeders of ‘dysplastic’ dogs’. Dog breeders will need to critically appraise the proposed ultrasound imaging diagnostic technique to see if it really offers them ‘new hope’ or whether the claim that it will provide breeders with many benefits is going to turn out as giving ‘false hope’.
A key part of the publicity was the qualifications of its innovator, a consultant orthopaedic surgeon. I have no doubt that he is one of the world’s leading experts in human hip problems, but I am sceptical that his techniques will find much application in dogs. For example, highlighted in the publicity material was his skill in correcting existing hip problems in human babies that are born with such problems by ‘dynamic splinting with a ‘pavlik’ harness’. This may be extremely valuable to humans as it has helped abnormal human baby hips grow to normality, especially when applied to a baby born eight weeks prematurely with dislocated hips that achieved normal hips by the time it would have been born full-term. Can you imagine applying splints to a six to eight week old puppy?
We need to be clear what the technique’s intention is for dogs, so the first crucial question needs to be: has the ultrasound imaging technique been developed for the purpose of correcting dysplastic hips by splinting/surgery thereby hiding the true natural hip status which is needed to determine a dog’s breeding value; or is it a technique that can be used by breeders to improve hip status in dogs by breeding from dogs that had ‘clear’ hips at six to eight weeks of age, and if so how will it do this?
The procedure is to screen six to eight week old puppies by ultrasound imaging which it is claimed will identify congenital, genetic ‘dysplasia’.
Let me start with the problem of scanning at six weeks. I do not doubt that ultrasound scanning at this age will diagnose any hip abnormalities that are present at six weeks, but this is long before skeletal maturity, before the bones and the joints have developed and matured, and bears little relationship to how the hips will be at skeletal maturity because developmental and environmental effects will intervene, as well as genetic effects realising their potential. The BVA/KC Hip Dysplasia Scheme requires “dogs to be at least one year of age [at assessment] in order to ensure skeletal maturity” (Dennis, 2012). (The claim in the ultrasound publicity that dogs are x-rayed in the BVA/KC Scheme from 18 months of age is incorrect).
The next question is how are the dogs scanned? For ultrasound scanning, puppies are held standing in the natural load bearing position without anaesthetic. Are they held perfectly still? Do they move or wriggle? Are all puppies ultrasound scanned in exactly the same position? The ultrasound publicity criticises the BVA/KC Scheme for manipulating a dog under anaesthetic into an unnatural position for x-raying with the dog positioned on its back and its hind legs pulled out caudally (rearwards). But what this does is to enable precise positioning of the pelvis and hips, and allows an excellent view of the hip joint in which all relevant areas can be seen clearly. All dogs are x-rayed in this position, and therefore the method is repeatable and consistent from dog to dog, enabling valid comparisons to be made (Dennis, 2012).
Then we need to know: what is measured by ultrasound scanning (there has to be a measure of something so that pups’ hips can be compared) and what is the result that is given? There was no mention in the original publicity material of making any measurement or of a definition of ‘normal’, but a later PUPscan Facebook posting added: “From the ultrasound images measurements of the shape and development of the ball and socket joint will be returned to the breeder and/or the new owner”. So the next question is: has the standard of normality in six to eight week old pups been defined so that individual puppy scans can be referenced against it, or is it just a subjective judgement by the scanner?
The ultrasound scan will result in giving a certificate of normality (at six to eight weeks old!) or a diagnosis in the event that a problem is found. In other words there are two ‘grades’ – normal and problem. The two-grade ultrasound system is even less than the old BVA/KC Scheme which had three grades – normal, breeder’s letter (near normal) and fail, and which was abandoned because of its lack of differentiation, being replaced with the current scoring scheme devised by Malcolm Willis in 1978 for German Shepherd Dogs and with the input of Prof Lawson, the chief scrutineer of the BVA panel at that time, extended to all breeds from ‘83 onwards.
The scoring scheme looks at nine radiographic features with a numerical ‘score’ being ascribed to abnormalities for each of the nine features, thereby describing the extent of any abnormality. It uses a scale from 0 (ideal) to six (worst) for eight of the features, with a maximum of five for one feature. This means that a dog can score from 0 to 53 for one hip or from 0 to 106 for both hips combined. The scheme thus allows a wide range of abnormality to be diagnosed (Dennis, 2012), yet the ultrasound publicity claims that ‘the present hip scoring system does not give a diagnosis’ (PUPscan Facebook, 2017)! Hip scoring is unlike most other hip schemes in which only a handful of different categories exist, but the proposed ultrasound scheme has the least of any scheme as there are only two categories, so hip scoring is a very powerful, differentiating method.
I have covered the genetics of HD in previous articles, but I will give a quick reminder. HD is a polygenic condition (controlled by many genes) and polygenic conditions usually show considerable phenotypic variation within a population. This is because what we see and measure in the phenotypic variation is not necessarily exclusively genetically produced. What we see has been influenced by a genetic component and an environmental component.
The genetic component can be further sub-divided into two sections called additive and non-additive, and the non-additive section is further divisible into what are called dominance and epistasis component parts (dominance refers to how different alleles of the same gene interact with one another; epistasis is how different genes interact with each other).
For complex polygenic traits we need to know in the broad sense what proportion of the variation in the population is genetic in origin. This is known as heritability. However, strictly speaking, heritability is the proportion of the variation in the population that is additive (ie not due to dominance or epistasis) and can thus be transmitted. The heritability of different traits can vary from 0 to 100 per cent. Traits of high heritability are those above 50 per cent; those from 20 to 50 per cent are of medium heritability; and those below 20 per cent are of low heritability (Willis, 1989). There are various estimates of the heritability of HD that show that it is on average about 30 per cent genetic (therefore of medium heritability) and 70 per cent environmental, but this will vary from breed to breed and is dependent on which scheme results were used in the calculations – for example a highly differentiating scoring scheme will give a more accurate estimate of heritability than a grading scheme with few grades.
In its publicity the ultrasound system claims that it can diagnose congenital genetic problems. We have to be clear what the difference is between ‘congenital’ and ‘genetic’. ‘Congenital’ means present at birth (eg in human babies a ‘hole in the heart’); ‘genetic’ means inherited. And yes some genetic conditions are evident at birth (eg mongolism caused by having an extra 21st chromosome), but many are not (eg lens luxation).
The publicity material contradicts itself when it says that it can identify congenital genetic HD at six to eight weeks, because it then goes on to say that “a hip that is demonstrated to be abnormal at six to eight weeks may be caused by intra-uterine abnormalities for example too little amniotic fluid, or by too little space [in the uterus] for the [foetal] pup to grow”. In other words the uterine environment may have caused the problem to be present at birth, and hip genetics may have had nothing to do with it! Ultrasound scanning therefore cannot possibly identify genetic causes.
The ultrasound publicity material has denigrated the BVA/KC Scheme almost at every opportunity. It criticises the present hip scoring system because it “does not give a diagnosis of genetic information” and “a score does not give a genetic diagnosis” (see above). That cannot be a criticism because no system that just looks at the physical condition (phenotype) can diagnose the genetic status, whether that be by x-ray imaging or ultrasound imaging. So the claims by the ultrasound publicity that ultrasound imaging “gives us the opportunity to identify genetic abnormalities” and that it “identifies genetic dysplasia” are therefore false. In this regard ultrasound imaging is the same as x-ray imaging – you can’t see the genes on x-ray plates or on ultrasound pictures!
It is claimed in the publicity that “the aim [of ultrasound scanning] is to find an answer as to why a hip score does not include a diagnosis to justify that score;” that “scores are not diagnoses and that KC/BVA hip scores do not diagnose problems” (see the last paragraph of the procedures section above). In my opinion all of this is not true and is misleading. The KC/BVA Scheme describes normality or abnormality in nine anatomical features of the hip joint. It attributes a score to the features based on objective definitions of each score, which have been tabulated by Dennis (2012). It thus gives as accurate a diagnosis of the condition of the hips as is possible.
The ultrasound publicity criticises the KC/BVA Scheme for not being able to say why there is an abnormality. Physical assessment schemes can determine what the abnormalities are, but no physical assessment scheme, x-ray or ultrasound, can say WHY there is an abnormality. Physical assessment schemes cannot give reasons and therefore cannot differentiate between genetic and environmental effects.
The ultrasound publicity denigrates the KC Estimated Breeding Value (EBV) for “not yet being able to demonstrate genetic abnormalities”, but such a statement only demonstrates the lack of understanding in the publicity material. EBVs do not “demonstrate genetic abnormalities” and never will because EBVs are not designed to do this. EBVs are designed to estimate the breeding value of a dog for good/bad hips, based on computations that incorporate the known scores of relatives, the degree of relationship of the relatives to the subject dog and the heritability of HD, thereby removing environmental effects from the computation to leave only the estimate of genetic status.
The reason that EBVs are only estimates is because the breeding value will change as more scores of relatives are added into the computation – the greater the amount of data put in, the more accurate the estimates become and this is expressed as a ‘confidence’ per cent, in other words how confident you can be with the estimate.
EBVs for hips and elbows are relatively new in dogs but they are the most powerful tools that have been made available to dog breeders, and they are proven to work in other livestock breeding, and given time they will lead to enormous improvement of hips and elbows in dogs. However we have to acknowledge that an EBV Scheme or an HD Scheme on its own will not improve hips. It is how breeders use the information generated by the schemes that will lead to improvement. The HD Scheme is therefore a tool for breeders to use – as either hip scores in breeds that do not yet have EBVs, or as hip scores computed into EBVs for some breeds, and if they use that information correctly it will lead to improvement.
The publicity material for ultrasound scanning claims that “x-raying dogs at full skeletal maturity has not reduced the prevalence of ‘dysplasia’ confirming that this label [ie the HD Scheme] is not fit for purpose”. This is a very serious, unfortunate and uncalled for slur on the BVA/KC Scheme because it is simply not true. The KC Dog Health Group Annual Report (2015) shows clearly the improvements that have been achieved for the breeds that have a high throughput of annual scoring. Progress is slow because that is the very nature of polygenic conditions which have moderate heritability, but whenever any condition has a genetic component influencing it then selective breeding will work and improvements will be made.
The PUPscan Facebook (2017) publicity material also denigrates the KC/BVA Scheme as “the current antiquated system of x-raying”. Despite previous criticisms from me about inconsistencies in the scoring of one dog in two different countries, I do consider the BVA/KC HD scheme is the best in the world, and I wish it could be adopted globally. As a new initiative to harmonise DNA testing worldwide has been announced by the International Partnership For Dogs (IPFD), of which the KC is a founding partner, perhaps there is hope that a harmonised HD Scheme could be advanced based on the BVA/KC Scheme.
The ultrasound publicity statement said that “[ultrasound] screening at six to eight weeks gives us the opportunity to identify the true genetic abnormalities”. This is in my opinion seriously flawed, untrue and cannot be substantiated scientifically and the statement was made despite the self-contradiction that the influence of the uterine environment is not a genetic condition of the hip (PUPscan Facebook, 2017).
It seems to me from the publicity material in Our Dogs and on Facebook, which is the only publicly available information on which an assessment of the ultrasound scheme can be made, that the only aspect that is claimed as a benefit is to “allow peace of mind” to breeders by obtaining a “certificate of normality” for the hips of pups at six to eight weeks of age, thus enabling “pups to go to their new homes with a clear certificate” thereby leaving breeders free of “litigation further down the line by dogs’ owners if a problem occurs with the dogs’ hips”.
The publicity material itself also states that “A problem found later on is almost certainly caused by other non-genetic factors such as trauma (often associated with poor husbandry), abnormal nutrition or Perthes disease”. In other words breeders will be able to say to owners that ‘your puppy was scanned normal at six to eight weeks so anything that is the matter with it in adulthood must have been caused by you the owner’. This is not something that makes an expert, caring breeder. It seems to me that this is nothing more than a cop-out, a pathetic excuse for a condition that has a genetic component for which a breeder is responsible. A second PUPscan Facebook announcement (www.facebook.com/woofhouseltd/#!/PUPscan) backed away from this by saying “The tests are simply designed to check things out at between six to eight weeks, not predict a score of lifetime’s wellbeing”.
I feel desperately sorry for breeders faced with this publicity, and I would urge them to get answers to all the issues I have raised before deciding whether ultrasound scanning at six to eight weeks of age is the “new hope” panacea for HD. n
Dennis, R. (2012). Interpretation and use of BVA/KC hip scores in dogs. In Practice, April 2012, Volume 34, Pages 178-194.
PUPscan publicity (2017). Story. www.facebook.com/permalink.php?story.
Willis, M B (1989). Genetics of the Dog. Published by H F and G Witherby Ltd, London.