Pre-operative hamstring length and velocity do not explain the reduced effectiveness of repeat hamstring lengthening in children with cerebral palsy and crouch gait

Abstract

BackgroundHamstring lengthening surgery (HSL) is often performed to correct crouch gait in patients with cerebral palsy (CP). However, crouch can recur over time, and repeat HSL may be ineffective. One possible reason is that the hamstrings in repeat HSL patients are neither short nor lengthening slowly and would therefore not benefit from HSL. Research questionThis study aimed to determine whether the hamstrings are short and/or slow preoperatively only in patients with primary, and not repeat, HSL. MethodsWe compared pre- and postoperative dynamic semimembranosus muscle-tendon lengths for children with CP who had primary (N = 15) or repeat (N = 8) HSL to a group of control participants (N = 10). Outcome measures were compared between visits (pre- vs. postoperative) and groups (control, primary HSL, repeat HSL) using mixed model analysis. ResultsPreoperatively, hamstrings were shorter and slower than normal on average in both HSL groups (p < 0.001); all but 3 limbs (primary 26/28, repeat 13/14) had hamstrings that were shorter and/or slower than controls by more than two standard deviations. Postoperative improvements were observed in the primary HSL group for popliteal angle, initial contact knee flexion, minimum stance knee flexion, and dynamic hamstring length (p ≤ 0.001). The repeat HSL group improved only in dynamic hamstring length (p = 0.004) and worsened in passive knee extension (p = 0.01) and minimum hip flexion in stance (p = 0.04). Hamstrings in both surgical groups on average remained shorter and slower than controls postoperatively (p ≤ 0.001). SignificanceThe fact that repeat HSL is less effective in improving knee motion is not due to a lack of short or slow hamstrings preoperatively. However, in recurrent crouch, short or slow hamstrings do not usually indicate hamstring dysfunction, and correction of other deformities such as rotational malalignment, fixed knee flexion contractures, patella alta, weak calf muscles, and/or loose heelcords should be considered rather than repeat HSL.