John H. Wedge, M.D., FRCSC
Surgeon-in-Chief
The Hospital for Sick Children
Associate Dean, Clinical Affairs
Faculty of Medicine, University of Toronto
Toronto, ON

I dread being asked to write or speak about Perthes' disease because there are so few undisputable or non-controversial "facts" that can be stated with reasonable certainty. Essentially, we know little more about this enigmatic condition, which predominantly affects males between the ages of four and 10 years of age and presumably has a multifactorial aetiology, than we did a quarter-century ago. This is not meant to be an exhaustive review article, so I refer the reader to three sensible and recent accounts of Perthes; the section in the latest AAOS OKU - Paediatrics series¹, the chapters in Lynn Staheli's marvellous book "Practice of Paediatric Orthopaedics"², and Wenger and Rang's beautifully illustrated "The Art and Practice of Children's Orthopaedics"³.  

Yes, magnetic resonance imaging produces pretty pictures of the epiphysis before early changes can be seen on plain radiographs and perhaps better predicts involvement of the proximal femoral physis and subsequent growth arrest, but falls short of predicting the course of the disease and the ultimate shape of the head of the femur when healing is complete. There are numerous classification systems, all with poor inter-observer reliability and which fail to provide guidance as to how any individual child should be treated. The Hering lateral pillar classification⁴ initially offered promise, suggesting that those who retained more than 50% of their height of the lateral portion of the epiphysis tended to do well regardless of treatment and those who lost more than 50% of the original height benefited from surgery. However, longer-term follow-up from the prospective multicentre Perthes' study⁵ has not confirmed this benefit of surgery in the latter group (pillar C) and that only those older than eight years in pillar B and in a new group called pillar B/C benefited. The necessity for this refinement of the Hering classification systems leads one to question the reliability and predictive value of it, as it appears to have become hopelessly arcane.

As far as the aetiology is concerned, there is little doubt that the vascularity of the epiphysis is impaired since dead bone is present, but post-mortem examinations in affected children suggests that more than a single infarct or disturbance of circulation is required to produce the clinical picture of Perthes' disease. It is more likely that avascular necrosis is the end stage of a pathological process rather than the primary event or cause. In the mid nineteen-nineties several studies suggested a hypercoaguable state might predispose a child to Perthes' disease but recent investigations have largely refuted this possible link⁶. Children who are exposed to passive smoking have a much greater frequency of Perthes' disease than those raised in a smoke-free environment⁷. However, Perthes' is also associated with low socio-economic status, that portion of society with the highest frequency of cigarette smoking. There is weak, if any, genetic influence on Perthes' so it is probable that multiple factors come together in a constitutionally vulnerable child to cause the disease. It has been known for decades that children with Perthes' lag well behind their chronological peers in skeletal development (greater than two standard deviations).

The prognosis for Perthes' worsens with advancing chronological age and those who are aged more than 10 years have an almost universally poor outcome even with aggressive treatment. Those less than five years old almost all do well without treatment but a small subset will end up with significant growth disturbance of the proximal femoral physis and a misshapen femoral head. The outcome as measured by the sphericity of the femoral head and the congruency of the hip joint at skeletal maturity worsens with increasing extent of epiphyseal involvement, although the correlation between the area and volume of the head affected and the end result is not very strong. The most important factors are the age at onset, the degree of collapse of the epiphysis and the length of time remaining until skeletal maturity for remodelling of the femoral head to occur. It is perhaps this latter factor, which explains why females appear to have a worse prognosis than males because of the more advanced skeletal maturation at onset.

Accepted principles of treatment are restoration and maintenance of range of motion of the affected hip and containment of the pliable femoral head in the acetabulum. These principles seem simple and straightforward but there is no consensus about how to apply them to any particular child with Perthes'. "Benign neglect", physical therapy, weight relief with crutches and/or calipers, serial Petrie casting, bracing, femoral osteotomy, pelvic osteotomy and combinations of each have their advocates and many authorities are passionate about their preferred method of treatment, but in truth have little proof of superiority over other options. Algorithms for treating Perthes' are very popular and some very elaborate. The sources mentioned in the first paragraph have flow charts which may be helpful to those infrequently encountering this relatively rare condition (1:10,000 population). If one can maintain satisfactory range of motion of the hip (>100° flexion, >30° abduction and an arc of rotation of 45° including at least 20° of internal rotation) during the phases of fragmentation and resorption of the dead epiphyseal bone, by whatever method, the outcome will most likely be favourable.

A novel treatment for Perthes' is the use of a hip distraction and hinge system with an external fixator to prevent collapse and extrusion of the epiphysis during the active phase of the disease. Numerous anecdotal reports of selected cases suggest that this technique may be of benefit but the only publication in the peer-reviewed literature of which I am aware failed to show significant benefit⁸. The protean nature of Perthes' disease requires comparison between appropriately matched children who receive a particular treatment with those who do not, before any claim for superiority may be confirmed. Give the logistical difficulty in meeting this condition, it is likely to be some time before we really know how Perthes' should best be treated in any given child. The universal application of any treatment method will inevitably result in a significant portion of children receiving treatment which is not necessary and from which they will not or cannot benefit.

My personal approach is to aggressively manage those children whose hip becomes stiff and irritable, initially with range of motion exercises and weight relief. I then progress to Petrie casting and an adductor tenotomy, when there is less than 30° of abduction while under anaesthesia, for those who do not respond. Once movement has improved, I am faced with the critical decision of whether to proceed with containment surgery. I prefer an innominate osteotomy in those between six and nine years and double-level osteotomy (femoral and innominate) in those older than nine years. However, the level of evidence to support this approach is not strong and therefore parental preference is very important in coming to a decision. Having made this disclaimer, Dr. Robert B. Salter and I have achieved Stulberg 1 or 2 femoral head shape in more than 80% of patients we have selected for surgery. When I detect resistance on the part of parents to surgery, I am reluctant to proceed and continue with nonoperative management rather than pushing surgery. A short course of bed rest and non-steroidal anti-inflammatory medication is reasonable initial management in the acute phase when the child has pain, a limp and stiff or irritable hip. This provides the parents time to become informed about Perthes' disease, typically from Internet web sites. Even though the information obtained from the Internet may be confusing, conflicting and frankly incorrect, those parents who want a "sure thing" from any proposed surgical operation may gain a better appreciation of how difficult it is to select the best form of management for their child and are then better equipped to participate in coming to the best decision.

In summary, I regret that I cannot give you exciting new information about this enigmatic and curious disorder, which has now been recognized for a century. Sophisticated imaging techniques have not been of much help in identifying those who may benefit from treatment. In fact, major breakthroughs leading to improved treatment will more likely come in the future from knowledge imported from scientific disciplines unrelated to our own. In the meantime, we must be careful not to overtreat yet be able to provide our best advice to parents with children with Perthes'. We must avoid pessimism and nihilism that leads to greater worry and the seeking of multiple unnecessary opinions from other orthopaedic surgeons.

References

1. Deluca P.A. Orthopaedic Knowledge Update - Paediatrics 2. American Academic of Orthopaedic Surgeons, Rosemount, Illinois 2002: pp 153-160.
   
   
2. Staheli L.T. Practice of Paediatric Orthopaedics 2001: Lippincott Williams & Wilkins, Philadelphia, USA: pp 146-151.
   
   
3. Wenger D. and Rang M. The Art and Practice of Children's Orthopaedics: 1992: Raven Press, New York.
   
   
4. Herring J.A. et al. The lateral pillar classification of Legg-Calvé-Perthes disease. Journal of Paediatric Orthopaedic 1992: 12: pp 143-150.
   
   
5. Herring J. et al. Legg-Calvé-Perthes disease. A multi-centre trial of five treatment methods. Abstract from Paediatric Orthopaedic Society of North America Annual Meeting: May 2003: p 26.
   
   
6. Thomas, D.P. et al. Perthes' disease and the relevance of thrombophilia. Journal of Bone and Joint Surgery 1999: 81-B: pp 691.
   
   
7. Mata S.G. et al. Legg-Calvé-Perthes disease and passive smoking. Journal of Paediatric Orthopaedics 2000: 20: pp 326-330.
   
   
8. Kocaglu et al. Ilizarov fixator for treatment of Legg-Calvé-Perthes disease. Journal of Paediatric Orthopaedics Part B 1999: 8: pp 276-281.