Hip joint subluxation after selective dorsal rhizotomy for spastic cerebral palsy

Transcription

1 J Neurosurg (Pediatrics 1) 103:10 16, 2005 Hip joint subluxation after selective dorsal rhizotomy for spastic cerebral palsy TUFAN HICDONMEZ, M.D., PAUL STEINBOK, M.B.B.S., F.R.C.S.(C), RICHARD BEAUCHAMP, M.D., F.R.C.S.(C), AND BONITA SAWATZKY, PH.D. Division of Pediatric Neurosurgery, Department of Surgery, and Department of Orthopedic Surgery, British Columbia s Children s Hospital, Vancouver, British Columbia, Canada Object. The effects of selective dorsal rhizotomy (SDR) procedures on the hip joints of children with spastic cerebral palsy (CP) are not well described. This study was performed to determine the incidence of hip subluxation in children with CP who underwent SDR at a single institution. Methods. The study group comprised 82 patients (164 hip joints) with a mean follow up of 3.9 years. Forty-four patients had spastic diplegia (53.6%), 35 had spastic quadriplegia (42.7%), two had spastic triplegia (2.4%), and one had spastic hemiplegia (1.2%). The mean patient age at SDR was years. Preoperative and postoperative hip radiographs were reviewed and the femoral head center edge (CE) angles were recorded. The mean pre- and postoperative right CE angles were 14.1 and 17.2, respectively, and those of the left were 13.6 and 15.1, respectively. Considering a change in CE angle greater than 5 as clinically significant, 72 hips (43.5%) remained unchanged, 63 (38.4%) improved, and 29 (17.7%) worsened. Of a number of preoperative variables, including age at time of surgery, Gross Motor Function Classification System (GMFCS) level, ambulatory status, extent of hip subluxation, preoperative scoliosis, and asymmetry of hip subluxation, only GMFCS level was associated with worsening of hip subluxation. Conclusions. Selective dorsal rhizotomy is more likely to have a positive effect or no effect on hip joint subluxation rather than to have a negative effect. More severe involvement, as measured using the GMFCS, may predispose to worsening of hip subluxation after SDR. KEY WORDS spasticity dorsal rhizotomy posterior rhizotomy outcome hip subluxation pediatric neurosurgery S Abbreviations used in this paper: AI = acetabular index; AP = anteroposterior; CE = center edge; CP = cerebral palsy; GMFCS = Gross Motor Function Classification System; MI = migration index; MP = Reimer migration percentage; SDR = selective dorsal rhizotomy. 10 ELECTIVE lumbosacral dorsal rhizotomy is a surgical procedure designed to reduce lower-limb spasticity and improve motor function in children with spastic CP. 2,10,17,24 In the last two decades, SDR has become accepted as a standard neurosurgical procedure for the treatment of spasticity associated with CP and is currently used at many centers around the world. 2,21,31 Children with spastic CP are known to be at risk of hip subluxation and dislocation. 6,7,13,14,19,22,28 The effects of the decreased spasticity that typically occurs after SDR, with attendant alterations in the balance of tone in muscles of the trunk and hips, may influence the development of hip joint deformities positively or negatively. The impact of SDR on the hip joints of the child, and in particular, on the later occurrence of structural changes to the hip joint, has not been well defined. The initial report on hip joint changes after SDR raised concern about rapid progressive subluxation of the hips. 12 Further reports documented a significant incidence of orthopedic surgery for hip subluxation after SDR, with a higher rate after longer follow up, but no reference was made to the preoperative status of the hips. 1,8 Systematic prospective studies of the hips before and after SDR have been reported from one center only, and these reports suggest that in general SDR resulted in stabilization or improvement in hip joint subluxation. 14,22 Focusing on a series of children with spastic CP who underwent SDR at a single institution, we performed the present study to determine the incidence of progressive hip subluxation and dislocation and to identify any factors that might correlate with the occurrence of hip joint changes after SDR for spastic CP. We also hoped to develop a model that would allow preoperative prediction of the probability of hip joint subluxation or dislocation developing after SDR. Clinical Material and Methods A retrospective analysis was performed of all children younger than 18 years of age who had undergone SDR at British Columbia s Children s Hospital from 1987 to Approval for the study was obtained from the Ethics Committee of the University of British Columbia. Patients were

2 Hip joint subluxation after selective dorsal rhizotomy identified through the prospective rhizotomy database that was initiated at the start of the SDR program. Information was extracted from the rhizotomy database about the sex, cause of spasticity, gestational age at birth, date of SDR, percentage of roots cut at each level, and preoperative ambulatory status. The patient records were reviewed retrospectively to assign a preoperative GMFCS level. 20 This system addresses the level of functional mobility of the child, with the mildest involvement being Level I, described as walks without restrictions; limitations in more advanced motor skills and the worst being Level V, selfmobility is severely limited even with the use of assistive technology. 19 In between, Level II is described as walks without assistive devices; limitations walking outdoors and in the community, Level III as walks with assistive mobility devices; limitations walking outdoors and in the community, and Level IV as self-mobility with limitations; children are transported or use power mobility outdoors and in the community. The GMFCS takes into account motor development as the child ages by having separate descriptions for each level in several age bands. Preoperative and postoperative pelvic AP radiographs of all patients were reviewed. Part of the planned protocol for children undergoing SDR included preoperative radiographs of the hips, repeated radiographs at 1, 2, and 5 years postoperatively, and thereafter as indicated clinically. No attempt was made to enforce this protocol, and often hip radiographs were not obtained as planned. These radiographs included AP plain radiographs of the pelvis showing the two hip joints and the upper half of the femoral bone. On the hip radiographs, the femoral head CE angle of Wiberg was measured (Fig. 1). This is the angle formed between the line (Perkin line) drawn perpendicular in the longitudinal axis (Hilgenreiner line) through the lateral bone edge of the acetabulum and the line drawn through the lateral bone edge of the acetabulum to the geometrical center of the femoral head (Fig. 1). Surgical Procedure Selective dorsal rhizotomy usually comprised partial rhizotomies of bilateral dorsal roots from L-2 to S-2. The amount of the dorsal roots cut at any level varied from 20 to 90%, and there was a trend over the years toward cutting smaller percentages of the roots and less of S-2 and L-4. The basis for selection of the dorsal nerve rootlets to be cut evolved over time. Initially, selection was based almost exclusively on electrophysiological responses to intraoperative electrical stimulation; since then, selection has been made on the basis of both clinical and intraoperative electrophysiological criteria. 32 Furthermore, as reported previously, the electrophysiological criteria changed over time. 32 Bilateral laminar cuts from L-1 to S-1 were performed for exposure in all patients, and the laminar flap was replaced at the end of the procedure. Statistical Analysis The data were exported from Filemaker Pro (Filemaker, Inc., Santa Clara, CA) to Microsoft Excel (Microsoft Corp., Redmond, WA). Descriptive statistical analysis was performed using Excel. Further regression analyses were performed to determine the contribution of a number of preop- FIG. 1. Hip radiograph (AP) showing the measurement of CE angle. 11

3 T. Hicdonmez, et al. erative variables to the extent of hip subluxation at time of most recent hip radiographs. This step was done to see if it might be possible to predict which patients might experience or have worsening hip subluxation after SDR. The preoperative variables included age at time of surgery, GMFCS level, ambulatory status, extent of hip subluxation of the worst hip, presence of scoliosis ( 10 ), and asymmetry of hip subluxation. The association of hip subluxation and the extent of the deformity at last follow up were assessed using the following variables present at that time: age, ambulatory status, GMFCS level, asymmetry of range of hip abduction, presence of scoliosis ( 10 ), extent of scoliosis, and duration of time of most recent hip radiographs after SDR. For the purpose of analysis, the valid study group comprised patients in whom hip radiographs were available preoperatively and 1 year or more postoperatively. For these analyses, the more problematic hip in each patient was used. The acceptable (that is, normal) CE angle was greater than A change of more than 5 in the CE angle was considered significant. Statistical analysis was performed using SPSS 11.0 for Windows (SPSS, Inc., Chicago, IL). Exploratory analysis was performed on all possible covariates, which were then tested for association with the outcome variable by using a multiple regression analysis. Results Demographic Data From February 1987 to February 2002, 193 children younger than 18 years of age underwent SDR at British Columbia s Children s Hospital. Of these 193 patients, 82 had preoperative hip radiographs and at least one postoperative series of hip radiographs at 1 year or later after SDR; these patients were designated as the study group. In the majority of the excluded patients, preoperative hip radiographs were not available for review, either because they had not been obtained or because they had been performed at another institution. A minority of patients were excluded because no postoperative hip radiographs were available for review. Of the 82 patients in the study group, 47 (57.3%) were boys and 35 (42.7%) were girls, with a mean age of 5.2 years (range years, median 4.4 years). Forty-four children (53.6%) had spastic diplegia, 35 (42.7%) had spastic quadriplegia, two (2.4%) had spastic triplegia, and one (1.2%) had spastic hemiplegia. Sixty-six children (80.4%) were born at less than 35 weeks gestation, with the mean gestational age being 31.5 weeks (range weeks, median 31 weeks). The preoperative functional ambulatory status in the TABLE 1 Number and percentage of patients at time periods of most recent radiological follow up for hip joint subluxation Yrs Post-SDR No. of Patients (%) 2 29 (35.4) (26.8) (25.6) (12.2) TABLE 2 Changes in status of 164 hips from pre- to post-sdr* No. of Hips (%) Hip Status CE Angle ( ) Pre-SDR Post-SDR normal (28) 73 (44.5) subluxation (67) 80 (48.8) severe sublux (4.2) 10 (6.1) disloc 40 1 (0.6) 1 (0.6) * Disloc = dislocation; sublux = subluxation. Same patient. children was as follows: 55 patients (67.1%) were in a wheelchair, 17 (20.7%) walked with a walker, one (1.2%) walked with crutches, and nine children (10.9%) walked independently, with or without aids, such as ankle foot orthoses. Preoperatively, 12 patients (14.6%) were at GMFCS Level V, 31 (37.8%) were at Level IV, 22 (26.8%) were at Level III, and 17 (20.7%) were at Level II. At the time of SDR, 52 children (63.4%) were 5 years of age or younger and 30 (36.6%) were older. The amount of the dorsal roots cut at any level varied from 5 to 85%. The mean radiological follow up was 4 years 2.7 years (range years; median 3.5 years; Table 1). At the time of the latest radiological follow up, the mean age of patients was years (range years, median 8.8 years). Hip Radiograph Measurements The mean right-side CE angle was 14.1 preoperatively and 17.2 after SDR at a mean radiological follow up of 4 years (range years). The mean pre- and postoperative left-side CE angles were 13.6 and 15.1, respectively. Using the CE angle, the following four radiological categories were described: 1) normal hip (CE 20 ); 2) hip subluxation (CE between 0 and 20 ); 3) severe hip subluxation (CE between 40 and 0 ); and 4) dislocated hip (CE 40 ). In the study group of 82 patients, 67 (81.7%) had subluxation in one or both hips. On analysis of the 164 hips independently in these 82 patients, the preoperative CE angles of 46 hips (28%) were normal, 110 (80%) had subluxation, seven (4.2%) had severe subluxation, and one hip was dislocated. Postoperative measurements revealed that 73 hips had a CE angle within the normal limits ( 20 ), 80 hips had subluxation, 10 hips had severe subluxation, and one hip (of the same patient) was dislocated (Table 2). To assess the changes in the extent of hip subluxation preoperatively to postoperatively, improvement was defined as greater than 5 of increase in CE angle, worsening as more than 5 of decrease in CE angle, and stable as CE changes with either decrease or increase of less than 5. Of the 164 individual hips examined, 72 hips (43.5%) remained stable, 63 (38.4%) improved, and 29 (17.7%) worsened. In many patients, the two hips changed differently from each other (Table 3). Predictors of Worsening in Degree of Hip Subluxation After SDR Further analysis was performed to determine if any asso- 12

4 Hip joint subluxation after selective dorsal rhizotomy TABLE 3 Status of each hip joint from pre- to post-sdr assessed radiologically* Variable (Side) No. of Patients (%) improved (both) 15 (18.3) unchanged (both) 19(23.2) worsened (both) 5 (6.1) improved (one); unchanged (other) 24 (29.2) improved (one); worsened (other) 9 (11.0) unchanged (one); worsened (other) 10 (12.2) total 82 (100) * Improvement and worsening were defined as a change in CE angle of greater than 5. ciation could be found between the extent of hip subluxation at time of latest hip radiographs and a number of preoperative variables, including age at time of surgery, GMFCS level, ambulatory status, extent of hip subluxation of worst hip (CE angle), preoperative scoliosis of more than 10, and asymmetry of hip subluxation ( 10 of range of movement between right and left). On univariate analysis, the only variable that correlated with an increased risk of worsening of hip subluxation (CE angle decrease 5 ) was the GMFCS level, with a higher level (corresponding to a worse functional status) predisposing to the development of worse subluxation (Table 4). Multiple regression analysis was then performed using the same variables to determine if it might be possible to create a model that would predict which patients might experience worsening hip subluxation after SDR. The only variable to show a significant relationship was the preoperative GMFCS level, but the amount of variation explained by the total model was only 16%, indicating the existence of many other factors not considered in this model that might have contributed to the outcome. Factors Potentially Associated With Presence of Hip Subluxation at Most Recent Follow Up The extent of hip subluxation at the most recent follow up was analyzed with respect to the following variables present at that time: age, ambulatory status, GMFCS level, asymmetry of range of hip abduction ( 10 between right and left), presence of scoliosis ( 10 ), extent of scoliosis, and duration of time between SDR and most recent hip radiographs. The only variable that was associated was the GMFCS level, with the more severely affected children at time of their most recent follow up having more hip subluxation than the others. Discussion Selective dorsal rhizotomy of the lumbosacral roots is an established neurosurgical procedure that may benefit some children with spastic CP. The procedure has been shown to decrease spasticity and to increase the range of movement in the lower limbs of children with spastic CP as well as improve their motor function. 3,18,25,31,33 35 Reports addressing the complications after SDR are infrequent, 1,36 and the impact of SDR on the hip joints, which is the subject of this study, has not been well characterized in the literature. The reported incidence of hip dislocation in children with CP ranges from 2.6% to as high as 60%, depending on the population being studied and the method of evaluation. 13,16,27,28 Scrutton, et al., 29 reported that as early as 18 months, migration percentages were significantly greater in children with CP than in the normally developing population. Subluxation is reported to be the second most common orthopedic problem in CP, after pes equinus deformity. 9,23 Hip subluxation and dislocation can affect posture and gait in ambulatory patients and can cause lower-extremity fractures, pain, and skin ulcerations in nonambulatory patients. 13,16 Griffiths, et al., 13 reported that in 79 children with CP who were 5 to 16 years old, subluxation of the hip joint was found in 14% of patients and dislocation in an additional 6%; furthermore, the AI and neck shaft angle increased in relation to the degree of the femoral head migration. Scrutton and Baird 28 followed 166 children with bilateral CP, and reported that by age 5 years, 43 hips in 31 of the 166 children had subluxation greater than 32%, as evidenced by the Reimer index. Sauser, et al., 27 reported hip changes in a series of 69 patients with CP. Thirty-six patients (52%) had subluxation or dislocation of one or both hips. The vast majority had Grade I (Reimer index up to 33%) or II (Reimer index up to 67%) hip subluxation, but nine hips had complete dislocation. 27 In the current study, 81.3% of 82 children with spastic CP selected for SDR had hip subluxation based on the CE angles at the time of SDR, with the majority having subluxation of both hips. This relatively high rate of subluxation probably reflects the severity of the underlying spasticity in the patients chosen for SDR. The natural history of hip deformity in patients with CP and spasticity, especially involving the hip adductors and flexors, is progressive subluxation leading to dislocation. 5,7, 13,37 The spasticity of the adductors and the iliopsoas muscle causes a shift in the axis of rotation of the femur from the center of the femoral head to the lesser trochanter. Therefore, lateral hip migration in growing children with CP tends to be progressive. 14 The risk of progression of subluxation is linked to many factors, such as the severity of neurological involvement, ambulation, and acetabular dysplasia. 16,19 The risk of progression of hip subluxation appears to be highest in the first few years of life. In one study, Vidal, et al., 38 determined that without surgical intervention, the MP, a measure of the amount of lateral subluxation of the femoral head) increased approximately 5.5% per year until age 5 in children with the potential to ambulate, whereas it increased 7% per year in children lacking such potential. In a review of 64 subluxed hips in 45 patients with CP in a 19-year follow up, Bagg, et al., 5 reported that subluxed hips progressed to dislocation in almost 33% of cases. Risk factors for progression to subluxation included young age, severity of CP, and an MI greater than 50%. 5 Using the AI to assess and compare the development of hip pathologies at 30 months and 5 years of age, Scrutton and Baird 28 concluded that all hips with AIs of at least 30 at 30 months had a hip problem requiring medical or surgical intervention by 5 years. In the light of these and other studies, orthopedic intervention has often been recommended before the age of 5 years to prevent the longer-term complications of hip subluxation, the progression of which in adolescence and adulthood has been studied infrequently. Miller and Bagg 19 determined that hips with an MP less than 30% remained stable until adulthood, whereas hips with 13

5 T. Hicdonmez, et al. TABLE 4 Number of patients with one hip having a CE angle improvement or worsening more than 5 after SDR according to preoperative GMFCS level* No. (%) of Patients Mean Change (range) No. (%) of Patients Mean Change (range) Total No. w/ Improved CE Angle for Hips w/ CE Angle w/ Worse CE Angle for Hips w/ CE Angle GMFCS Level of Patients of 1 Hip by 5 Improved by 5 of 1 Hip by 5 Worsened by 5 I 0 II 17 7 (41.2) 9.9 (6 15) 4 (23.5) 11.0 ( 7 to 15) III (59.1) 11.3 (7 21) 2 (9.1) 7.0 ( 6 to 8) IV (61.3) 13.3 (6 36) 10 (32.3) 12.8 ( 6 to 21) V 12 7 (58.3) 18.3 (12 31) 6 (50.0) 18.0 ( 7 to 32) * For the hips that improved or worsened by greater than 5, the extent of worsening is also detailed. For any patient, if both hips worsened or improved by greater than 5, the hip that changed the most was used in determining the extent of change. In degrees. MPs between 30 and 60% (mild to moderate dislocation) worsened in 25% of patients through adolescence. Hips with an MP between 60 and 90% were treated surgically to prevent subsequent dislocation. 19 These prior reports indicate that hip subluxation occurs commonly as part of the natural history of spastic CP and is often progressive. This situation makes assessing the effects of SDR on hip deformity complicated, because at least two major interacting factors have an impact on the development of hip and lower-limb deformities: the natural history of the hip in CP and the effects of the reduction of spasticity by the SDR. In theory, the decrease in spasticity after SDR could be expected to reduce the deforming forces of the spasticity and any subsequent hip subluxation and dislocation, yet the pathogenesis of the hip subluxation in CP is multifactorial. 14,37 Other factors include muscle imbalance caused by the overactivity of the hip flexors and adductors, pelvic obliquity, coxa valga, persistent femoral anteversion, and asymmetry in the spasticity and range of movement around the hip joint. 7,19,22 Hip Changes After SDR The relationship of SDR to progressive hip subluxation has been controversial. Greene, et al., 12 first expressed concern about an adverse effect of SDR on hip subluxation. They reported rapid progression of hip subluxation in the 1st year after SDR in seven hips (six patients) and surmised that this phenomenon might be fairly common. They recommended that patients be followed closely after SDR for hip subluxation in the 1st year after surgery. They also suggested that because the L-1 root was generally not cut during SDR, the rapid subluxation might relate to residual hip flexor spasticity in the presence of reduced tone in the hip adductors and extensors. This factor could result in imbalance in muscle tone around the hip joint, and together with preexisting bone changes in the proximal femur and acetabulum could lead to rapid progression of hip subluxation after SDR. 12 Their hypothesis was in keeping with prior work on the pathogenesis of hip subluxation in spastic CP, which suggested that the increased tone in the iliopsoas muscles might play a key role. 7 In two retrospective series of patients with SDR, the incidence of orthopedic surgery for progressive hip subluxation was 12% in patients followed for more than 2 years 1 and 25% in patients followed for more than 5 years, 8 suggesting that the incidence of progressive hip subluxation increased with longer follow up. None of these studies reviewed the overall outcome of the SDR on the hip joints; rather they looked only at the incidence of orthopedic intervention. A number of prospective studies have compared changes in the hips before and after SDR. Stempien, et al., 37 reported, in abstract form only, that in 45 patients with hip displacement prior to SDR, 86% remained radiologically stable 1 year later. Park, et al., 22 reported that of 134 hips in 67 children with spastic diplegic and quadriplegic CP, 75% remained unchanged, 17% improved, and 8% worsened after SDR at a follow up ranging from 6 to 46 months. The children in their study ranged from 2 to 11 years at the time of SDR. The same group examined in their series the effect of SDR on migration of the femoral head in children with spastic quadriplegia who ranged in age from 2 to 9 years. Of the 90 hips in 45 patients, 9% improved, 80% remained unchanged, and 11% worsened, at a mean follow up of 20 months (range 7 50 months). 14 Arens, et al., 4 reported that no case of subluxation or threatening subluxation of the hips that had been present preoperatively progressed further toward dislocation after reduction of spasticity in a series of 51 children observed up to 7 years after SDR. Hodgkinson, et al., 15 reported on a series of 20 children with spastic quadriplegic CP who all underwent SDR after 5 years of age and noted no change in the extent of hip subluxation at 1 year after SDR. In the present series of 164 hips in 82 children with spastic diplegia who underwent SDR at 2.7 to 14.6 years of age, the average extent of hip subluxation improved after SDR, as measured by the CE angle. On the basis of the CE angle results, 43.5% of hips remained unchanged and 38.4% improved radiologically at a mean follow-up time of 4 years (range 1 12 years). Thus, as has been the case in other reported prospective studies, even accounting for longer follow up, the vast majority of hips (82.3%) did not deteriorate radiologically after SDR and more hips improved than deteriorated. The rate of progressive hip subluxation is toward the lower end of the range generally reported for children with spastic CP who have not undergone SDR, as detailed previously. 13,16,27,28 Furthermore, the relatively high rate of improvement in hip subluxation after SDR is different from what might have been expected without SDR, indicating that SDR is more likely to be beneficial than harmful for children with hip subluxation. Factors Related to Progressive Hip Subluxation After SDR A number of factors have been identified in the literature 14

6 Hip joint subluxation after selective dorsal rhizotomy that might predispose patients with CP to experience progressive hip subluxation. These factors include the presence of hip dysplasia or severe subluxation; 5,19 a young age, with more progression in first 5 years of life, 38 more severe CP, with a 60% rate of hip subluxations or dislocations in dependent sitters compared with 7% in independent ambulators; 16 asymmetrical range of motion in the hips; 7,9 scoliosis; 9 and intellectual delay. 9 In our study, no correlation was found between age at SDR and outcome of hip subluxation. This result is consistent with the findings of Park, et al., 22 who found no difference in outcome in children younger than 4 years of age compared with older children 5 to 11 years of age. Of the aforementioned factors that have been reported to predispose patients to progression of hip subluxation, the only one that was of any significance in our study was the severity of CP, as measured using the GMFCS. It is important to note, however, that no matter what the severity of CP, hip subluxation was more likely to improve than worsen after SDR. With respect to rhizotomy procedures, Greene, et al., 12 suggested that rapid worsening of hip subluxation might relate to residual hip flexor spasticity in the presence of reduced tone in the hip adductors and extensors, as a result of cutting the L-2 roots and below but leaving the L-1 roots intact. The results of our study tend to refute Greene and colleagues hypothesis. Only four of the 82 children in our study had L-1 rootlets cut and all had L-2 rootlets cut, yet hip subluxation generally improved or remained stable. Strengths and Limitations of This Study Our study is distinctive for the following reasons: it has the largest number of patients reported in the literature, the follow-up period is longer than in other reports, and the pre- and postoperative data were collected prospectively as part of a set of guidelines. Limitations include the lack of a comparison group of controls who did not undergo SDR and the relative brevity of the follow up. A large number of patients who had undergone SDR did not meet the study criteria because either pre- or postoperative hip radiographs were not available for review. This omission could create bias if patients who did and did not have radiographs were selected in some fashion. In fact, it had been expected that appropriate radiographs would have been obtained in all patients, but preoperative hip radiographs were not acquired in many patients at our institution. Thus, bias is unlikely. Another potential concern is the choice of the CE angle as the outcome measure in this group of patients. This measurement relies on the accurate identification of the center of the femoral head, which may be more difficult to define in younger children in whom the head of the femur may not be circular. Hence, the MI was developed and is considered by some to be more reliable in younger children. 9,26 Nevertheless, in a report on the effectiveness of varus osteotomy in patients with CP, ranging in age from 3 years to 15 years (mean 7.7 years), Settecerri and Karol 30 compared pre- and postoperative indices and found similar results when using both the CE angle and MI. They noted no difficulty in using either of the measurement tools. Thus, we think that the use of the CE angle as the outcome measure in this study is acceptable and that the results are valid. Conclusions The results of this study suggest that the hip subluxation and dislocation are consequences of the natural history of spasticity in children with CP. Selective dorsal rhizotomy is more likely to improve hip subluxation than to cause deterioration. The outcomes after SDR are better than one would expect from the known natural history. More severe spastic CP, as measured using the GMFCS, may predispose to worsening of hip subluxation after SDR, as is also the case in patients who have not undergone SDR. Acknowledgments We wish to thank Ruth Milner and Jeremy Hamm at the Children s and Women s Health Center statistical consulting unit for performing and consulting on the statistical analyses. We thank also Hyeon Sook Kim for assistance with GMFCS grading. References 1. Abbott R: Complications with selective posterior rhizotomy. Pediatr Neurosurg 18:43 47, Abbott R: Sensory rhizotomy for the treatment of childhood spasticity. J Child Neurol 11 (Suppl 1):S36 S42, Abbott R, Forem SL, Johann M: Selective posterior rhizotomy for the treatment of spasticity: a review. Childs Nerv Syst 5: , Arens LJ, Peacock WJ, Peter J: Selective posterior rhizotomy: a long-term follow-up study. Childs Nerv Syst 5: , Bagg MR, Farber J, Miller F: Long-term follow-up of hip subluxation in cerebral palsy patients. J Pediatr Orthop 13: 32 36, Beck AJ, Gaskill SJ, Marlin AE: Improvement in upper extremity function and trunk control after selective posterior rhizotomy. Am J Occup Ther 47: , Carr C, Gage JR: The fate of the nonoperated hip in cerebral palsy. J Pediatr Orthop 7: , Carroll KL, Moore KR, Stevens PM: Orthopedic procedures after rhizotomy. J Pediatr Orthop 18:69 74, Cornell MS: The hip in cerebral palsy. Dev Med Child Neurol 37:3 18, Fasano VA, Broggi G, Zeme S, Lo Russo G, Sguazzi A: Longterm results of posterior functional rhizotomy. Acta Neurochir Suppl 30: , Fredensborg N: The CE angle of normal hips. Acta Orthop Scand 47: , Greene WB, Dietz FR, Goldberg MJ, Gross RH, Miller F, Sussman MD: Rapid progression of hip subluxation in cerebral palsy after selective posterior rhizotomy. J Pediatr Orthop 11: , Griffiths GJ, Evans KT, Roberts GM, Lloyd KN: The radiology of the hip joints and pelvis in cerebral palsy. Clin Radiol 28: , Heim RC, Park TS, Vogler GP, Kaufman BA, Noetzel MJ, Ortman MR: Changes in hip migration after selective dorsal rhizotomy for spastic quadriplegia in cerebral palsy. J Neurosurg 82: , Hodgkinson I, Berard C, Jindrich ML, Sindou M, Mertens P, Berard J: Selective dorsal rhizotomy in children with cerebral palsy. Results in 18 cases at one year postoperatively. Stereotact Funct Neurosurg 69: , Lonstein JE, Beck K: Hip dislocation and subluxation in cerebral palsy. J Pediatr Orthop 6: , McLaughlin J, Bjornson K, Temkin N, Steinbok P, Wright V, Reiner A, et al: Selective dorsal rhizotomy: meta-analysis of three randomized controlled trials. Dev Med Child Neurol 44: 17 25,

7 T. Hicdonmez, et al. 18. McLaughlin JF, Bjornson KF, Astley SJ, Hays RM, Hoffinger SA, Armantrout EA, et al: The role of selective dorsal rhizotomy in cerebral palsy: critical evaluation of a prospective clinical series. Dev Med Child Neurol 36: , Miller F, Bagg MR: Age and migration percentage as risk factors for progression in spastic hip disease. Dev Med Child Neurol 37: , Palisano RJ, Hanna SE, Rosenbaum PL, Russell DJ, Walker SD, Wood EP, et al: Validation of a model of gross motor function for children with cerebral palsy. Phys Ther 80: , Park TS: Selective dorsal rhizotomy: an excellent therapeutic option for spastic cerebral palsy. Clin Neurosurg 47: , Park TS, Vogler GP, Phillips LH II, Kaufman BA, Ortman MR, McClure SM, et al: Effects of selective dorsal rhizotomy for spastic diplegia on hip migration in cerebral palsy. Pediatr Neurosurg 20:43 49, Parrott J, Boyd RN, Dobson F, Lancester A, Love S, Oates J, et al: Hip displacement in spastic cerebral palsy: repeatability of radiologic measurement. J Pediatr Orthop 22: , Peacock WJ, Arens LJ: Selective posterior rhizotomy for the relief of spasticity in cerebral palsy. S Afr Med J 62: , Peacock WJ, Staudt LA: Selective posterior rhizotomy: further comments. J Child Neurol 6: , 1991 (Letter) 26. Reimers J: The stability of the hip in children. A radiological study of the results of muscle surgery in cerebral palsy. Acta Orthop Scand Suppl 184:1 100, Sauser DD, Hewes RC, Root L: Hip changes in spastic cerebral palsy. AJR Am J Roentgenol 146: , Scrutton D, Baird G: Surveillance measures of the hips of children with bilateral cerebral palsy. Arch Dis Child 76: , Scrutton D, Baird G, Smeeton N: Hip dysplasia in bilateral cerebral palsy: incidence and natural history in children aged 18 months to 5 years. Dev Med Child Neurol 43: , Settecerri JJ, Karol LA: Effectiveness of femoral varus osteotomy in patients with cerebral palsy. J Pediatr Orthop 20: , Steinbok P: Outcomes after selective dorsal rhizotomy for spastic cerebral palsy. Childs Nerv Syst 17:1 18, Steinbok P, Gustavsson B, Kestle JR, Reiner A, Cochrane DD: Relationship of intraoperative electrophysiological criteria to outcome after selective functional posterior rhizotomy. J Neurosurg 83:18 26, Steinbok P, McLeod K: Comparison of motor outcomes after selective dorsal rhizotomy with and without preoperative intensified physiotherapy in children with spastic diplegic cerebral palsy. Pediatr Neurosurg 36: , Steinbok P, Reiner A, Beauchamp RD, Cochrane DD, Keyes R: Selective functional posterior rhizotomy for treatment of spastic cerebral palsy in children. Review of 50 consecutive cases. Pediatr Neurosurg 18:34 42, Steinbok P, Reiner AM, Beauchamp R, Armstrong RW, Cochrane DD, Kestle J: A randomized clinical trial to compare selective posterior rhizotomy plus physiotherapy with physiotherapy alone in children with spastic diplegic cerebral palsy. Dev Med Child Neurol 39: , Steinbok P, Schrag C: Complications after selective posterior rhizotomy for spasticity in children with cerebral palsy. Pediatr Neurosurg 28: , Stempien L, Gaebler-Spira D, Dias L, Storrs B, Cioffi M, Feathergill B: The natural history of the hip in cerebral palsy following selective posterior rhizotomy. Dev Med Child Neurol (Suppl 62) 32:5 6, Vidal J, Deguillaume P, Vidal M: The anatomy of the dysplastic hip in cerebral palsy related to prognosis and treatment. Int Orthop 9: , 1985 Manuscript received October 10, Accepted in final form January 27, Address reprint requests to: Paul Steinbok, F.R.C.S.(C), Division of Pediatric Neurosurgery, British Columbia s Children s Hospital, 4480 Oak Street, Room K3 159, Vancouver, British Columbia, Canada V6H 3V4. psteinbok@cw.bc.ca. 16