Morphometric Analysis of the Femur in Cerebral Palsy: 3-dimensional CT Study

Background: The cause of hip disorder in cerebral palsy (CP) has been thought to involve muscle imbalance, flexion, and adduction contracture of the hip joint, acetabular dysplasia, and femoral growth abnormalities. The aim of this study was to quantitatively evaluate the 3-dimensional femoral geometry and subluxation/dislocation of the hip in spastic CP using 3D-CT reconstructed images of the pelvis and the femur, focusing on the femoral growth abnormalities in CP. Methods: Between June 2006 and September 2009, 186 hips in 93 bilateral spastic CP patients, including spastic diplegia (SD) in 73 patients and spastic quadriplegia (SQ) in 20 patients, who had not received any surgical treatment, were investigated using 3D-CT at our hospital. There were 59 boys and 34 girls with an average age of 5.3 years (range: 2.6 to 6.8 y). As an index for the femoral geometry, the neck-shaft angle, the femoral anteversion, and the femoral offset were 3-dimensionally measured. The center of the acetabulum and the femoral head were determined to calculate the CT migration percentage as the distance between these centers divided by the femoral head diameter. To elucidate the factors related to hip subluxation/dislocation, the relationships between the neck-shaft angle, the femoral anteversion, the femoral offset, and the CT migration percentage were investigated. Results: The mean neck-shaft angle was 150.4 +/- 9.4 degrees (range: 129.4 to 173.2 degrees). The mean femoral anteversion was 44.4 +/- 13.6 degrees (range: 5.8 to 84.0 degrees). The mean CT migration percentage was 22.4 +/- 22.7% (range: 3 to 129%). There was positive correlation between the CT migration percentage and the neck-shaft angle (r = 0.49). Hips with large CT migration percentage tended to show coxa valga. There was an inverse correlation between the neck-shaft angle and the femoral offset (r = -0.90), but no correlation between the CT migration percentage and the femoral anteversion (r = 0.26), between the femoral offset and the femoral anteversion (r = -0.25), or between the neck-shaft angle and the femoral anteversion (r = 0.23). The neck-shaft angle, the femoral anteversion, and the CT migration percentage were significantly larger, and the femoral offset was significantly smaller, in patients with the Gross Motor Functional Classification System (GMFCS) level IV/V (nonwalking children) and SQ type, than in patients with GMFCS level II/III (mostly walking children) and SD type. Conclusions: The 3-dimensional femoral geometry in CP patients can be analyzed quantitatively using 3D-CT regardless of the abnormal spastic posture. Our data indicate that 3-dimensional evaluation is accurate and useful for analysis of the femur and acetabulum in CP, and that the extent of coxa valga and femoral anteversion is more severe in the patients with GMFCS level IV/V and SQ type. Level of Evidence: Level IV.