New hope for old hips

Osteoarthritis is a common, often-debilitating disease of both animals and humans. It may involve any joint, but the joints most clinically affected are the hip, elbow, shoulder, knee, and spine. It is estimated that one in five dogs will show obvious signs of pain from osteoarthritis; in many more dogs, however, the signs will be subtle or subclinical. In dogs, the most common form of osteoarthritis is canine hip dysplasia. The radiographic prevalence of canine hip dysplasia is as high as 75% in some breeds and the disease can cause pain and dysfunction in affected animals. Most types of osteoarthritis, including canine hip dysplasia, are thought to be developmental diseases having complex inheritance, so-called polygenic diseases. It is understood that the incidence and severity of such genetic diseases can be influenced considerably by environmental factors,1-4 meaning any nongenetic factor, such as diet, lifestyle, housing, or trauma. In the past 15 years, landmark collaborative investigations have been conducted at the Nestlé Purina laboratories in St. Louis, Missouri, demonstrating the power and utility of environmental manipulation to offset or prevent a genetic predisposition to disease. These studies evaluated the effect of food restriction on development of hip joint laxity during growth and subsequent occurrence of osteoarthritis in joints during adulthood. Although not all results have yet been published, this lifelong investigation in a cohort of 48 Labrador retrievers is unique and represents our best understanding of the natural course of joint osteoarthritis, and especially canine hip dysplasia.5-7 The most recent unpublished study is presented below in abstract form.

ABSTRACT (UNPUBLISHED DATA):

Influence of Restricted Feeding and Age on Hip Osteoarthritis and Hip Score: A Life-long Study in Labrador Retrievers

Smith GK, 1 Power MY,1 Biery DN,1 Evans RH,2 McKelvie P, 1 Shofer FS,1 Gregor TP,1 Ballam JM,3 Mantz SL,3 Lust G,4 Houlton JEF,5 Lawler DF,3 Kealy RD3

1Department of Clinical Studies, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; 2Veterinary Pathology Services, Aliso Viejo, California; 3Nestlé Purina Research, St. Louis, Missouri; 4James A. Baker Institute for Animal Health, College of Veterinary Medicine, Cornell University, Ithaca, New York; 5Davies White Veterinary Specialists, Higham Gobion, Herts, England

There has been slow progress in reducing the incidence of canine hip dysplasia by selective breeding of normal dogs.8 The conventional diagnosis of canine hip dysplasia has been based on subjective radiographic findings of subluxation of the coxofemoral joint, or secondary osteoarthritis as seen on evaluation of the hip-extended, ventrodorsal radiographic view of the pelvis. In the United States this analysis is performed by the Orthopedic Foundation for Animals (OFA) scoring system after dogs are 2 years of age. In much of Europe a similar analysis is made after 1 year of age. Excessive hip laxity (Norberg angle measured from the hip-extended radiograph) can be marginally reduced by limiting caloric intake and this tightening of the joint prevents or delays the expression of osteoarthritis in some dogs predisposed to the condition.5-7 It has been generally assumed that the subjective scoring of hip phenotype at 1 or 2 years of age accurately reflects the true hip phenotype of the dog. No lifelong studies have been conducted to document the accuracy of the one- or two-year evaluation to predict the end-of-life hip phenotype. The purpose of the present investigation was to test the influence of food restriction on hip phenotype and to compare early hip screening (OFA and PennHIP scores at 2 years of age, and the British Veterinary Association/Kennel Club [BVA/KC] scores at 1 year of age) with end-of-life hip phenotype (radiographic osteoarthritis, OFA score, and histopathologic osteoarthritis).

Materials and methods

Forty-eight 8-week-old Labrador retriever puppies from seven litters were allotted by pairing to two groups of 24 dogs each. One group was fed ad libitum (full-fed) and each member of the other group (limit-fed) was given 25% less of the same food given to the full-fed pair-mate. Radiography (ventrodorsal, hip-extended) of the hip joints was done when the dogs were ages 30 and 54 weeks and yearly thereafter for life. Subluxation was measured using the Norberg angle on radiographs made with the dog in the standard, hip-extended position. The same radiographs were evaluated for evidence of canine hip dysplasia and osteoarthritis by an independent, blinded, board-certified radiologist (DNB) using criteria of the OFA scoring system. Radiographic hip scores at one year of age were provided by the current Chief Scrutineer (JEFH) using criteria of the BVA/KC. At 2 years of age, the hips were evaluated and scored using the PennHIP method. Aged dogs were euthanatized when dictated by quality of life. Histopathology of the hips was performed on 45 of the 48 euthanatized dogs.

Results

Restricted feeding had a profound positive effect on the hip phenotype of Labrador retrievers. Restricted-fed dogs had significantly lower incidence and severity of canine hip dysplasia and osteoarthritis compared with full-fed pair-mates. This health benefit continued for the life of the dogs. In the pooled sample of 48 dogs, the prevalence of hip osteoarthritis increased linearly throughout the study, from 15% at 2 years of age to 67% at end-of-life. For the full-fed dogs, end-of-life osteoarthritis prevalence was 83% and for the restricted-fed dogs, 50%.

At 2 years of age, OFA-type scoring found 19 of the 48 dogs in the study to be dysplastic while 29 dogs were scored as normal. The 19 dysplastic dogs remained dysplastic for life, with OFA scores increasing in severity for many of the dogs. However, of the 29 dogs scored as normal, 16 (55%) were scored dysplastic by end-of-life, representinga 45% false-negative rate of diagnosis at 2 years of age.

Histopathologically, 92% of the dogs scored as normal at 2 years of age using the OFA type scoring method had evidence of osteoarthritis, equating to 8% negative predictive value. From BVA/KC scoring criteria applied at one year of age, 59% of dogs having total hip scores less than 16 (the current mean for Labrador retrievers in Britain) went on to develop radiographic osteoarthritis, and 94% developed histopathologic evidence of osteoarthritis. The mean BVA/KC hip score for this sample of 48 Labrador retrievers was 14.4. PennHIP results showed that all the dogs included in this investigation were susceptible to osteoarthritis (distraction index [DI] > 0.36, range 0.36–0.92). Kaplan-Meier curves of disease-free interval showed that for dogs with a low DI the onset of osteoarthritis was much later in life than dogs with a large DI. For dogs with DI 0.4, the median disease-free interval was 12 years of age compared with dogs with DI > 0.6, whose median disease-free interval was only 3 years of age.

Discussion

This lifelong study demonstrated a profound effect of an environmental factor, diet, on a genetic disease, canine hip dysplasia; by keeping dogs lean, the onset of osteoarthritis was delayed and its severity and prevalence were reduced significantly. In addition, osteoarthritis prevalence in other joints of the lean dogs was decreased. Such findings are both statistically and clinically significant.

Theoretically, development of osteoarthritis at a primary site such as the hip could produce aberrant compensatory biomechanical forces acting on other joints, the altered weight bearing and ambulation leading to multiple joint osteoarthritis. In this study, hip joints were not always affected first; some dogs had radiographic evidence of shoulder or elbow lesions without hip joint involvement. Thus, alternatively it may be hypothesized that osteoarthritis has a systemic cause, with variable expression in different joints. This might explain in part the development of osteoarthritis in joints that are not subject to large forces associated with weight bearing, such as lumbar intervertebral joints in dogs and joints in the skeleton of sharks.9

Alternatively, a humoral substance from an affected joint may affect tissues in other joints. This concept is supported by the recent finding that extract of human arthritic bone tissue induced abnormalities in disease-free articular cartilage.10 Additional support for the concept of variability in tissue susceptibility is derived from a recent report that shoulder joints in young adult dogs at high risk for chronic hip dysplasia have histopathologic articular cartilage abnormalities similar to those found in the corresponding hip joints.11

Yet another hypothesis might be that osteoarthritis among different joints represents phenomena that are not interrelated or interdependent.

The linear increase in osteoarthritis incidence over the life of these Labrador retrievers refutes popular dogma that holds that hip osteoarthritis occurs either early in life, in the case of dysplastic dogs, or much later in the geriatric years in the case of "old age osteoarthritis." Dogs that were permitted to become overweight (mean body condition scores of 6.7) expressed osteoarthritis much more, 83%, than those kept lean (50%) (mean body condition score of 4.2).

The principal risk factor, if not the cause, for the development of hip osteoarthritis—irrespective of age—has been shown to be joint laxity.8,9 In the hip-extended radiographic view, this laxity is underestimated and often masked completely, leading to false-negative diagnoses. This means that some dogs appear phenotypically normal but are genotypically abnormal and, if bred, pass unwanted genes to subsequent generations. Evidence for this impression derives from the continued high frequency of chronic hip dysplasia in many breeds of dogs despite systematic mating of normal parents.

This lifelong study reported here provides additional evidence. Hip phenotypes in this sample of average Labrador retrievers (mean DI = 0.54, mean BVA/KC = 14.4) were much worse at the end of life than at 1 or 2 years of age. The normal subjective designation of hips at 2 years of age was wrong more than it was right (45% false-negative rate) when compared with end-of-life radiographic scores. When 2-year OFA-type score was compared with histopathologic scoring the negative predictive value of this screening method was found to be very low, 8%. Of the dogs satisfying current breeding standards in the United States and Britain, 92% and 94%, respectively, went on to develop osteoarthritis. Understandably BVA/KC scoring is not a pass/fail scheme and lower numbers are associated with better hip phenotypes, and presumably better genotypes. However, even if one examined the sample of dogs having combined hip scores 8, 94% of this pool also showed histopathologic evidence of osteoarthritis by end of life.

In contrast, the PennHIP DI at 2 years of age indicated that all dogs in this study were susceptible to osteoarthritis and therefore genotypically abnormal. This predicted susceptibility was borne out by the observed pattern of osteoarthritis expression later in life both radiographically and histopathologically—in fact, only one of 48 dogs was found to be free of both radiographic and histopathologic evidence of osteoarthritis. Importantly, diet did not influence the DI; this is an extremely desirable characteristic of a genetic test and suggests superior heritability for this selection criterion.

Clinical significance and recommendations:

The following clinical recommendations derive from the results of this life-long study:

• Perform early hip evaluation (screening) for breeding dogs as well as pet dogs.

• Keep dogs—especially those found to be susceptible to hip osteoarthritis—lean for life.

• Implement preventive measures, e.g., diet, early in life.

• Radiograph hips (and other joints) at intervals throughout life.

Acknowledgement: Nestlé Purina PetCare Research, St. Louis, Mo.

The disease and treatment

Osteoarthritis is a highly prevalent disease of complex inheritance that causes mild to severe pain and variable degrees of disability in affected dogs.

There are no medical or surgical cures. Medical or surgical treatments are palliative at best. The new generation of cyclooxygenase (COX-2) specific nonsteroidal anti-inflammatory drugs (NSAIDs) have great utility in the management of osteoarthritis. Glucosamine-chondroitin sulfate may have measurable although modest efficacy in treating osteoarthritis. New and multiple modalities of pain relief are under investigation in humans and dogs to provide comfort for the patient. The practicing veterinarian needs to keep abreast of this rapidly changing landscape. Reconstructive surgery (femoral head and neck excision or total hip replacement) may be the treatment of choice for severe osteoarthritis.

Prevention and control

Selective breeding of unaffected animals is the most effective means to control the frequency and severity of the disease. Genetic control of canine hip dysplasia requires:

• an accurate screening method keyed to a phenotype with optimal heritability,

• an organized screening program based on the phenotype identified by diagnostic testing,

• an associated database containing essential phenotypic and pedigree information, and

• trust and cooperation between breeders and veterinarians who perform the diagnostic procedure.

Alternatively, by imposing appropriate environmental conditions, such as weight control, the veterinarian in concert with the dog owner can reduce the frequency of disease expression and effect real benefit to the patient population. Several surgical techniques purport to prevent hip osteoarthritis; however, these methods have not yet received sufficient investigation to warrant widespread use. It is important to note that no surgical treatment has yet been definitively shown to prevent osteoarthritis either in animals or in humans.

References

1. Henricson, B. et al.: On the etiology and pathogenesis of hip dysplasia: A comparative review. J. Small Anim. Pract. 7:673–688; 1966.

2. Hutt, F.B.: Genetic selection to reduce the incidence of hip dysplasia in dogs. JAVMA151:1041–1048; 1967.

3. Jessen, C.R.; Spurrell, F.A.: Heritability of canine hip dysplasia. Proc. Canine Hip Dysplasia Symp. and Workshop, St. Louis, MO., 1972 ; pp 53–61.

4. Hedhammer, A. et al.: Canine hip dysplasia: A study of heritability in 401 litters of German Shepherd dogs. JAVMA 174:1012–1016; 1979.

5. Kealy, R.D. et al.: Effects of limited food consumption on the incidence of hip dysplasia in growing dogs. JAVMA201: 857–863; 1992.

6. Kealy, R.D. et al.: Five-year longitudinal study on limited food consumption and development of osteoarthritis in coxofermoral joints of dogs. JAVMA 210:222–225; 1997.

7. Kealy, R.D. et al.: Evaluation of the effect of limited food consumption on radiographic evidence of osteoarthritis in dogs. JAVMA217:1678–1680; 2000.

8. Corely, E.A.: Role of the Orthopedic Foundation for Animals in the control of canine hip dysplasia. Vet. Clin. North Am. (Small Anim. Pract.)201:857–863; 1992.

9. Sokoloff, L.: The pathology of osteoarthritis and the role of aging. The Aetiopathogeneis of Osteoarthritis (G. Nuki, ed.). Pitman, Tunbridge Wells, United Kingdom, 1980; pp 1–15.

10.Wastacott, C.I. et al.: Alteration of cartilage metabolism by cells from osteoarthritic bone. Arthritis Rheumatol.40:1282–1291; 1997.

11. Farquhar, T.: et al.: Variation in composition of cartilage from the shoulder joints of young adult dogs at risk for developing canine hip dysplasia. JAVMA210:1483–1485; 1997.