Are hips stable in children with cerebral palsy?

Abstract

This commentary is on the original article by Terjesen on pages 951-957 of this issue. In 2010, the mechano-biological principles of proximal femoral geometry and muscle forces were implemented on a three-dimensional finite element model of the proximal femur, in order to predict changes in morphology over time in both typically developing children and children with cerebral palsy (CP).1 The simulation relies on the assumption that cyclic octahedral shear stress promotes growth and cyclic hydrostatic compressive stress inhibits growth. This finite element model predicts an increase in femoral anteversion and coxa valga for load conditions in children with CP when compared with typically developing children. In this study, growth was simulated over a period of approximately 5 months. From three population-based studies2–4 we know that hip displacement is common in children with CP (with an overall incidence of approximately one-third); it is often present by 2–3 years of age; range of motion is a poor indicator of hips at risk; and it is directly related to gross motor function as graded by the Gross Motor Function Classification System (GMFCS). In these studies the cut-off values for the migration percentage were set at slightly different levels: 30–33% for hip subluxation and 90–100% for hip dislocation.2–4 In a new study, Terjesen has carefully evaluated and reported a population-based hip surveillance programme for children with CP in Norway.5 A large number of children (n=335) were included and the data were collected and analysed meticulously. Hip displacement occurred in 26% of all children (subluxation in 22% and dislocation in 4%) and in 63% of those at GMFCS levels IV and V. The mean migration percentage was 20.4% at the initial radiographs and 34.0% at the end of the study. The rate of increase in migration percentage increased with decreasing functional level, from 0.2% per year at GMFCS level I to 9.5% at GMFCS level V. Radiographic hip surveillance from age 1 to 2 years was recommended for non-ambulatory children with CP on a regular basis. Current management options for hip displacement in children with CP include reconstructive procedures such as varus derotation osteotomy of the proximal femur,6 preventive surgery such as adductor releases,7 or a combination of both.8 A large recently published study7 demonstrated that the success of preventive surgery for hip displacement was directly related to GMFCS level. ‘Success’ was defined as the absence of subsequent surgical procedures during the study period of 7 years and a migration percentage of less than 50%. The success rate was 32%. However, the success rate was predicted by GMFCS level, with 94% at GMFCS level II decreasing to 14% at GMFCS level V. Finally, the question remains as to when the hip is stable or unstable in patients with CP. The data from current studies support the fact that hips in children with CP on a high functional level (GMFCS levels I, II, and III) are usually stable, or easily made so – but that hips in non-ambulatory children with CP (GMFCS levels IV and V) are usually not stable! Hip surveillance is important and needs to start by the age of 2 years in non-ambulatory children. The main goal of systematic hip screening is to prevent hip dislocation. Future research, based on the understanding of the biomechanics of the proximal femur, should focus on specific management guidelines for each GMFCS level or at least for ambulatory and non-ambulatory children with CP.