Background Spastic hip subluxation or dislocation that is associated with an excessive coxa valga deformity is a common pathologic condition in children with cerebral palsy (CP) that is often treated with large bone reconstructive procedures. Guided growth techniques (such as stapling, plate, or transphyseal screw) have been widely used to alter the growth axis in patients with a lower-limb deformity but only a few reports have described their use in patients with coxa valga deformities. Questions/purposes (1) Does guided growth surgery using a transphyseal screw combined with adductor tenotomy prevent progressive coxa valga deformity and lateral hip subluxation in children with CP? (2) What factors influence the correction of coxa valga deformity and the success of hip stabilization? (3) What complications were associated with this operation and how often did children treated with it undergo reoperation? Methods From 2012 to 2016, at our institution, three authors (H-CH, KNK, K-WW) retrospectively studied data on children with CP who underwent guided growth of the hip for progressive bilateral hip subluxation associated with coxa valga deformities. A single percutaneous screw was inserted across the inferomedial portion of proximal femoral physis in an AP view and centered along femoral neck in lateral view under fluoroscopy guidance. During the period, we treated 25 consecutive children with CP who had progressive hip subluxation with coxa valga deformities. The indications for surgery were migration percentage > 30% and head-shaft angle > 155 degrees with at least 2 years growth remaining. Of those, 13 patients underwent guided growth alone, and 48% (12) underwent a combination of guided growth and adductor tenotomy. Of the 25 patients treated with this approach, 96% (24) were available for follow-up with complete data at a minimum of 2 years follow-up (mean 50 months; range 25 to 72). All children (17 boys and seven girls; 48 hips) underwent surgery at a mean age of 8 years (range 5 to 12). With regard to the gross motor function classification system, three patients were Level I, four patients were Level II, seven patients were Level III, seven were Level IV, and three were Level V. Radiographic parameters including the head-shaft angle, Hilgenreiner's epiphyseal angle, acetabular index, and Reimer's migration percentage were assessed before surgery and at the latest follow-up examination by one author (H-CH). Complications and reoperations were assessed by chart review. During the period in question, we generally offered secondary reconstructive surgery to patients who underwent a guided growth procedure once their subluxation progressed. Results With the data available, the coxa valga and lateral hip subluxation improved in terms of the reduction of head-shaft angle by a mean of 13 degrees +/- 7 degrees (95% CI 11 to 15; p < 0.001) and the reduction of the migration percentage by 10% +/- 11% (95% CI 7 to 13; p < 0.001). After controlling for potentially confounding variables like gender, gross motor function classification system, Hilgenreiner's epiphyseal angle and acetabular index, we found that longer follow-up duration (r = 0.234; p < 0.001) and a smaller preoperative migration percentage (r = -0.258; p = 0.004) were associated with larger changes in the head-shaft angle. In terms of complications, we found that the proximal femoral physis grew off the screw tip in 44% (21 of 48 hips) at a mean of 28 months. Among these, 31% of hips (15 of 48) in 33% of patients (eight of 24) underwent replacement with a longer screw. Among the 17% of hips (eight of 48) in 21% of patients (five of 24) who had progressive lateral subluxation and underwent secondary reconstructive surgery, we found that their preoperative acetabular index was higher (mean 29 degrees versus 21 degrees; p < 0.001), as was their head-shaft angle (mean 166 degrees versus 162 degrees; p = 0.045), and migration percentage (mean 54% versus 36 %; p < 0.001). Conclusions Although guided growth with single transphyseal screw did not create as large a degree of varus as proximal femoral osteotomy, it did stabilize the hip in children with cerebral palsy with migration percentage less than 50% in our series. It is a simple procedure that can be of benefit to children with cerebral palsy with unstable hip. Reoperation in patients where the physis has grown off the screw tip can be a problem; fortunately, it is a rather minor procedure to replace with a longer screw.
