Risk Factors for Hip Displacement in Children With Cerebral Palsy: Systematic Review

REVIEW ARTICLE Risk Factors for Hip Displacement in Children With Cerebral Palsy: Systematic Review Blazej Pruszczynski, MD,* Julieanne Sees, DO,w and Freeman Miller, MDw Background: When hip displacement in children with cerebral palsy (CP) is identified early, treatment is more successful. The standard test is a radiograph of the pelvis measuring the migration index (MI). Our study aims to review published literature of the natural history of hip dislocation among children with CP and to define related risk factors to develop screening criteria for early recognition. Methods: The review included 10 studies with sample sizes greater than 20 children with CP below18 years who had hips with no surgical intervention or dislocation at initial presentation, minimum 2-year follow-up, and recorded MI, pattern, and Gross Motor Function Classification System (GMFCS) level. Results: On the basis of this review, we suggest screening with 1 radiograph for GMFCS I and II, or, if MI > 30%, an annual radiograph between ages 2 and 8 years, followed by a radiograph every 2 years until the age of 18 years. For GMFCS III, IV, and V, we recommend an annual radiograph if MI < 30% or 1 every 6 months if MI > 30% between ages 2 and 8 years, followed by radiograph every 2 years until the age of 18 years. Conclusions: Applying a practical surveillance program for children with CP can prevent hip dislocation, provide early treatment, and ultimately lead to consistently better outcomes than those of neglected hip dislocations. The GMFCS level has a strong impact on subluxation risk and that the risk continues to the end of growth. Level of Evidence: Level III systematic review. Key Words: cerebral palsy (CP), migration percentage (MP), subluxation, dislocation, gross motor function classification system (GMFCS), radiograph (J Pediatr Orthop 2016;36:829 833) Cerebral palsy (CP) is one of the most common causes of physical disability in children, accounting for 2% to 2.5% per 1000 live births. 1,2 Hip joint dislocation is a common disease in children with CP especially in those children with severe neurological involvement and may From the *Department of Orthopedics and Pediatric Orthopedics, Medical University of Lodz, Lodz, Poland; and wdepartment of Orthopedics, Nemours/Alfred I. DuPont Hospital for Children, Wilmington, DE. Source of funding: none. The authors declare no conflicts of interest. Reprints: Freeman Miller, MD, Department of Orthopedics, Nemours/ Alfred I. DuPont Hospital for Children, P.O. Box 269, Wilmington, DE 19899. E-mail: Freeman.Miller@nemours.org. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. progress from subluxation to dislocation. In early stages of hip subluxation, children may have no signs of pain or discomfort, and pain may be present with the dislocation of the hip. 3,4 The prevalence of complete dislocation has been reported to be between 10% and 15% 5,6 ; subluxation has been estimated between 25% and 60%. 7,8 Hip displacement in children with CP occurs due to contractures and spasticity of adductors, hip flexors, and medial hamstrings. 9 The muscle imbalance causes secondary osseous deformity with increased femoral anteversion and acetabular dysplasia and leads to hip joint instability. In later stages, there is increased stiffness and limited range of motion of the hip joint, which may make toileting and maintaining personal hygiene difficult. Further problems with positioning, sitting, standing, walking, and spinal deformity may also be impacted by hip displacement. 2,10 Early stages of subluxation are silent; as the subluxation becomes more severe, some develop pain, and/or decrease function with increasing spasticity as the hip migrates laterally. 11 The Gross Motor Function Classification System (GMFCS) categorizes the ambulatory ability and motor function of children with CP for surveillance for impending deformities, such as progressive hip disease. 12 The treatment of hip displacement in children with CP is less invasive and more successful in hips with less hip degenerative change and less displacement. In addition, pain from degenerative changes may be prevented. When the hip is highly dislocated and the patient has been exposed to the pain for a longer time, surgical results are much more likely to be nonadequate. Moreover, it is important to understand the natural history of the hip displacement and the risk factors involved with further displacement. 2,4 7,13,14 To reliably treat hips early, they need to be identified early, which means there needs to be some screening or surveillance program in place. The gold standard test for demonstrating hip displacement in children with CP is a radiograph of the pelvis measuring the displacement using the method of Reimers. 15 A radiographic surveillance program should focus on the highest risk patients at the age of greatest risk. A similar program is well developed for newborn screening for early identification of developmental hip dysplasia in infants. Children with CP, however, develop hip displacement later and are not identified with early infant screening. The aims of this study are to assess published medical literature of the natural history of hip dislocation in children with CP and to evaluate the evidence using a systematic review to define risk factors related to J Pediatr Orthop Volume 36, Number 8, December 2016 www.pedorthopaedics.com 829

Pruszczynski et al J Pediatr Orthop Volume 36, Number 8, December 2016 progression of hip displacement in children with CP to develop screening criteria for early identification. METHODS Study Design Our initial goal was to do quantitative evidence synthesis, which is a variation of systematic quantitative review (meta-analysis), to synthesize the evidence on natural history of hip displacement in children with CP. However, we found that quantitative evidence synthesis was not possible to conduct because of the limited published specific data and data variation; therefore, we proceeded with a simple summary data review. Eligibility: Inclusion and Exclusion Criteria The following inclusion criteria were established for the published literature: (a) children below 18 years with CP, (b) nonoperated and non Botox-treated hips, (c) nondislocated hips at initial presentation, (d) recorded migration index (MI), (e) minimum of 2 years follow-up, (f) reasonable sample size (more than 20 patients in the overall study), (g) anatomic pattern of CP, (h) recorded type of CP, and (i) GMFCS or classification of mobility in walkers, independent sitters, and dependent sitters. We excluded papers with poor quality and small overall sample size (< 20) because of their tendency to provide invalid evidence on the natural history of CP hip displacement. The statistical analysis techniques used in the papers were examined for the adequacy to answer the research question. Finally, we determined whether there was proper inference from the data. Definition: Type of CP CP is a static lesion occurring in the immature brain. Descriptive patterns in CP are commonly based on the anatomic involvement. The patterns include hemiplegia, involving one half of the body; diplegia, involving primarily the lower extremities with mild upper extremity involvement; and quadriplegia, involving all 4 extremities. The term triplegia has been suggested for individuals whose pattern includes having hemiplegia on 1 side and diplegia of the lower limbs. Other terms include paraplegia, which may imply purely a lower extremity paralysis; monoplegia, where 1 limb is primarily involved; and pentiplegia, where an individual has no independent head control. 11 Functional Scale The GMFCS is the standard classification system for mobility assessment and ambulatory ability in CP. 12,16 It is a 5-level system based on self-initiated movement with particular emphasis on sitting control and walking. Levels are based on functional abilities and limitations; the need for assistive devices, such as walkers, crutches, canes, and wheeled equipment, for mobility; and, to a lesser extent, the quality of the child s movement. 12 For each child, the focus of the GMFCS is on determining the suitable level that best represents the daily performance of that child at home, at school, and in the community. Statistical Analysis Data for the MI progression measurement were extracted from papers. Categorical variables were summarized with frequency and percentages, whereas the continuous variables were summarized with mean and SD. RESULTS Databases were used to search published literature through 2013 that had full-text publication of the containing data on the natural history of hips in children with CP. With these search terms and the eligibility criteria, we identified 10 studies that are included in a final analysis. Of the studies that met our inclusion criteria, 10 were eligible for the review. Because of the lack of direct data in the published papers (except 2: Terjesen 14 and Ha gglund et al 6 ), we were not able to perform any comparison between papers regarding the change of MI measurements (in the time frame), GMFCS assessment, and type of CP. Overall Risk for Hip Subluxation Related to GMFCS Level The risk of developing hip subluxation is directly related to the severity of motor impairment as assessed by the GMFCS. By combining 4 studies 2,6,8,14 that reported indication for subluxation of either greater than 30% or 33% migration, we could summarize the outcome of 1082 children and show that there is almost a true linear relationship in increased risk as the severity of GMFCS increases (Figs. 1, 2). All 4 of these papers also evaluated the impact of anatomic pattern involvement, and, as this closely mirrors GMFCS and anatomic pattern has less clearly defined diagnostic criteria, we have elected to focus on GMFCS as the measure of motor disability. There are reports of hip dysplasia in children with type-iv hemiplegia in GMFCS I and II, but there is no evidence FIGURE 1. Risk of developing hip subluxation related to GMFCS (1 = 100% risk of subluxation). GMFCS indicates Gross Motor Function Classification System; MI, migration index. 830 www.pedorthopaedics.com Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

J Pediatr Orthop Volume 36, Number 8, December 2016 Risk for Hip Displacement in Children With CP FIGURE 2. Rate of progression of hip subluxation by GMFCS (data reported as % migration per year change). GMFCS indicates Gross Motor Function Classification System; MI, migration index. of the incidence; moreover, the age at which it occurs may be later. 17 Risk Related to Age To determine the age at highest risk for the development of hip subluxation we combined data on 182 patients from 2 studies 1,4 to report the age at which hips reached 30% or 33% migration. These data show a clear peak between ages 3 and 5 years old (Fig. 3). There was further information in 1 paper 7 that evaluated the age at the time of hip dislocation in 38 hips. These data show a later peak, around ages 5 and 6 (Fig. 3). Rate of Increase in the MI As these children start with normal hips and then develop hip displacement in childhood, we wanted to identify the rate at which this displacement occurs. The rate of displacement was reported in 2 papers 2,6 (379 patients) showing a linear relationship increasing from 0.2% per year in GMFCS I to 12% per year in GMFCS V (Fig. 1). Also, 1 report found a rate of increase of 8.2% per year in children younger than 5 and 4.4% per year in children older than 5. 2 Another study looking at an older population found from mean age 7 to age 26 a mean increase of 7% per year with most of this occurring in those with a migration over 50% in childhood or adolescence. We could not find any other data on the rate of change in MI relative to hip position. Muscle Tone and Movement Disorder as a Risk for Hip Subluxation or Dislocation None of the authors observed hip subluxation or dislocation in ataxic type of CP. 2,13,16 The last authors calculated the relative risk for hip subluxation as 0.11 (confidence interval, 0.03-0.45) compared with diplegia. 2 In spastic diplegic patients, Ha gglund et al 18 found hip FIGURE 3. Percent of the total population reaching the subluxation or dislocation category (vertical axis is mean % of population reported, horizontal is age in years). MI indicates migration index. subluxation in 33.7% and Soo et al 2 in 19.2%. Soo et al 2 stated that in their CP population, 40% of hips are subluxated in dystonia, and 44.4% are subluxated in hypotonia. There was no clear pattern in reports related to spasticity, hypotonia, and dystonia that was a clear trend except all groups were at risk, except ataxic type pattern seems to be at very low risk. Hip Range of Motion as a Risk Factor Early reports on screening focused on limited hip abduction as an important screening method. 15,19 There are reports indicating that hypotonic hips also subluxate 4,12 ; therefore, limited hip abduction is not a good screening tool in this population although limitation of abduction is a strong indicator for surveillance in those patients who are spastic or dystonic. Weak Points of Articles The studies of Terjesen 14 and Hägglund et al 18 had higher number of patients classified as GMFCS I (about one-third of group). In both Ha gglund papers, the quadriplegic group is the minority (5 times fewer than others respectively). 6,18 In Soo et al, 2 twice as many patients were classified as GMFCS I compared with the other groups, and almost one-third of the cohort had surgery with prevention or reconstruction. Most papers also do not clearly state the age at which the GMFCS is assessed as this is not a reliable measure for 2- to 3-year-old children because of the individual variation in development at that age. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. www.pedorthopaedics.com 831

Pruszczynski et al J Pediatr Orthop Volume 36, Number 8, December 2016 DISCUSSION Our overall assessment of the literature is that there are good quality data published showing a strong association of GMFCS level to risk of developing hip displacement. 2,6,8,14 There are also relatively strong data showing the age at greatest risk for developing hip displacement in childhood is between ages 3 and 5 years. 1,4 The data for risk in adolescence and young adulthood are largely absent except for 1 report, which shows an increased risk of progression if migration is over 50%. 3 The only data on age of dislocation suggest a peak in childhood at age 5 and 6 but with some risk throughout the teenage years. With the goal of developing recommendations for a surveillance program using radiographic measurement of hip displacement as defined by Reimers, 15 we can now consider the optimal ages that should be screened. We generally classify MI > 90% as dislocated; however, in some papers this is not so clearly defined, and we are going with the definition used in each paper. The hip risk groups are GMFCS III, IV, and V between the ages of 2 and 8 years old. If the goal is to identify hip displacement between 30% and 40% migration, then annual radiographs beginning at age 2 years are recommended based on the risk factors from GMFCS and age (Fig. 2). Assigning GMFCS for children aged 2 or younger is very difficult; therefore, it is better for all children with CP around age 2 to have a radiograph. This can also reveal displacement in the rare child with developmental dysplasia that was missed at birth or the rare GMFCS I and II child that may have displacement. If this radiograph shows <30% migration and the child is independently ambulating (GMFCS I and II), further screening is not indicated based on the published data reviewed. The frequency of radiographs, based on the assumption that the previous radiograph was completely normal (< 30% migration), should be annual based on the highest reported mean rate of subluxation (12 degrees per year for GMFCS V). This means at the highest risk, the subluxation would still be under 45% at diagnosis. If, however, the MI is 30% to 40% and there is a decision to continue to monitor, then the radiographs should be increased to every 6 months, especially for GMFCS V and for children below 5 years. 2 On the basis of the published literature, it is not possible to make recommendations for monitoring beyond age 8 years, except to say that in those with significant subluxation (over 50%), migrating will likely progress, 3 and dislocations also continue to occur throughout adolescence 7 (Fig. 3). On the basis of this limited information, it seems reasonable to continue radiographs every 2 to 4 years until skeletal maturity for GMFCS III, IV, and V. Those hips with subluxation that have not been treated are at higher risk and should have closer follow-up. 3 Therefore, based on the best current literature of natural history, our recommendation for a surveillance program for the hips of children with CP is presented in Table 1. Children with type-iv hemiplegia in GMFCS I and II seem to be a small group at risk in the high functioning subset that TABLE 1. Recommendations for Identifying Frequency of Radiographic Hip Displacement Surveillance in Children With CP Age 2-8 y Age 8-18 y GMFCS I, II One radiograph None GMFCS I, II (MI > 30%)* Annual radiograph Every 2 y GMFCS III, IV, V Annual Every 2 y GMFCS III, IV, V (MI > 30%)* Every 6 mo Annually *Authors suggestions using only 1 AP supine pelvis radiograph. CP indicates cerebral palsy; GMFCS, Gross Motor Function Classification System; MI, migration index. might benefit from more screening into adolescence, but there is no objective data that would allow us to make a specific recommendation on frequency. 17 Limitations These recommendations are based on the reported literature of the natural history, and some elements, such as the rate of progression of the subluxation in relation to the degree of subluxation, have almost no published data. There is a subset children with type-iv hemiplegia (unilateral) in GMFCS I and II who are at higher risk, but data on natural history are not available to allow an objective recommendation. 17 On the basis of the presumption that early intervention is used to prevent hip dislocation, it is very unlikely that a true natural history study with sequential radiographic follow-up will be untaken. Moreover, the rate of hip subluxation in a child with GMFCS V who is already 50% subluxated is likely higher than if the hip has 25% migration. In addition, there is no attempt in this review to define the optimal degree of subluxation at which time surgical intervention or any other intervention should be applied. Although surveillance only has validity if early intervention is planned, the timing and specific intervention is beyond the scope of this review. SUMMARY Hip dislocation may be preventable with a surveillance program 7 and early treatment of the hips. 20 The Swedish surveillance program is very successful; only 1 hip dislocation has been reported because the parents refused surgical treatment. 18 Surveillance is practical and effective, providing consistently better outcomes than those of neglected hip dislocations. With a well-applied surveillance program and early treatment, hip dislocation in CP may become as rare as developmental hip dislocation. REFERENCES 1. Stanley F, Blair E, Alberman E. Birth events and cerebral palsy: facts were not presented clearly. BMJ. 2001;322:50. 2. Soo B, Howard JJ, Boyd RN, et al. Hip displacement in cerebral palsy. J Bone Joint Surg Am. 2006;88:121 129. 3. Bagg MR, Farber J, Miller F. Long-term follow-up of hip subluxation in cerebral palsy patients. J Pediatr Orthop. 1993;13: 32 36. 832 www.pedorthopaedics.com Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

J Pediatr Orthop Volume 36, Number 8, December 2016 Risk for Hip Displacement in Children With CP 4. Graham HK. Painful hip dislocation in cerebral palsy. Lancet. 2002;359:907 908. 5. Cooke PH, Cole WG, Carey RP. Dislocation of the hip in cerebral palsy. Natural history and predictability. J Bone Joint Surg Br. 1989; 71:441 446. 6. Ha gglund G, Andersson S, Düppe H, et al. Prevention of dislocation of the hip in children with cerebral palsy. The first ten years of a population-based prevention programme. J Bone Joint Surg Br. 2005;87:95 101. 7. Elkamil AI, Andersen GL, Ha gglund G, et al. Prevalence of hip dislocation among children with cerebral palsy in regions with and without a surveillance programme: a cross sectional study in Sweden and Norway. BMC Musculoskelet Disord. 2011;12:284 291. 8. 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. 2001;43:586 600. 9. Miller F, Slomczykowski M, Cope R, et al. Computer modeling of the pathomechanics of spastic hip dislocation in children. J Pediatr Orthop. 1999;19:486 492. 10. Flynn JM, Miller F. Management of hip disorders in patients with cerebral palsy. J Am Acad Orthop Surg. 2002;10:198 209. 11. Miller F. Cerebral Palsy. New York, NY: Springer; 2005. 12. Palisano RJ, Cameron D, Rosenbaum PL, et al. Stability of the gross motor function classification system. Dev Med Child Neurol. 2006;48:424 428. 13. Terjesen T. Development of the hip joints in unoperated children with cerebral palsy: a radiographic study of 76 patients. Acta Orthop. 2006;77:125 131. 14. Terjesen T. The natural history of hip development in cerebral palsy. Dev Med Child Neurol. 2012;54:951 957. 15. 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. 1980;184:1 100. 16. Minear WL. A classification of cerebral palsy. Pediatrics. 1956;18: 841 852. 17. Rutz E, Passmore E, Baker R, et al. Multilevel surgery improves gait in spastic hemiplegia but does not resolve hip dysplasia. Clin Orthop Relat Res. 2012;470:1294 1302. 18. Hägglund G, Lauge-Pedersen H, Wagner P. Characteristics of children with hip displacement in cerebral palsy. BMC Musculoskelet Disord. 2007;8:101 107. 19. Silver RL, Rang M, Chan J, et al. Adductor release in nonambulant children with cerebral palsy. J Pediatr Orthop. 1985;5:672 677. 20. Presedo A, Oh CW, Dabney KW, et al. Soft-tissue releases to treat spastic hip subluxation in children with cerebral palsy. J Bone Joint Surg Am. 2005;87:832 841. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. www.pedorthopaedics.com 833