We present three children with primary subacute epiphyseal and metaepiphyseal osteomyelitis. The diagnosis was delayed because of subtle radiological findings and mild general symptoms. Primary epiphyseal osteomyelitis is extremely rare. We believe that this is the first time that the MRI findings have been presented. In the first case they revealed a perforation into the knee and therefore an intra-articular epiphyseal approach was used for debridement. In the second and third cases the metaepiphyseal lesions showed considerable physical involvement and a metaphyseal approach was chosen. We believe that in this condition MRI is essential both for diagnosis and in the planning of surgical treatment.

Introduction: Autopsy findings (Jensen and Lauritzen 1976, Catterall et al. 1982) as well as own MRI studies (Lange et al. 1996) indicate that in Perthes’ disease there is an early cartilaginous enlargement of the femoral head. Lack of concomitant acetabular enlargement will lead to loss of containment and subluxation. We divided the transverse acetabular ligament (TAL) to promote expansion of the acetabulum for prevention of femoral head extrusion and loss of containment.

Material and Methods: We report 13 patients with Perthes’disease belonging to Catterall group III or IV. The operation was performed when MRI showed a labrum lift near horizontal position indicating risk of loss of containment (Meiss 2001). There was an average cartilaginous head enlargement of 11 % in comparison to the uneffected side as measured by the Maximum Oblique Diameter. Division of the TAL was performed through an antero-medial approach (Ludloff 1913, Wein-stein 1993). A window of about 1,5 x 1 cm was created in the capsule which was left open.

The TAL was divided but not removed. In all cases strict non-weightbearing was imposed postoperatively (use of a wheel chair and crutches) until well into the regeneration phase. The average period of non-weight-bearing was 1 year and 10 months. An abduction pillow was worn at night.

Results: The results after an average follow-up of 4,2 years (range 2,6 -5,3 years) were evaluated according to Stulberg (1981) and Catterall (1982) with emphasis on the radiographic appearance (sphericity of the femoral head, joint congruity, containment [acetabulum head index], articulo-trochanteric distance). The result was excellent in 2 cases, good in 8, satisfactory in 2, and poor in 1 case. Two patients had a bad compliance for non-weightbearing and underwent additional bony procedures. The outcome was satisfactory and poor.

Conclusions: Our data indicate that the combination of the division of the TAL and a strict conservative treatment gives very satisfactory results in Catterall group III and IV cases that present with signs of risk of loss of containment on MRI.