CHILDREN’S ORTHOPAEDICS
Avascular necrosis as a complication of the treatment of dislocation of the hip in children with cerebral palsy A. Koch, M. Jozwiak, M. Idzior, M. Molinska-Glura, A. Szulc From University of Medical Sciences, Poznan, Poland
We investigated the incidence and risk factors for the development of avascular necrosis (AVN) of the femoral head in the course of treatment of children with cerebral palsy (CP) and dislocation of the hip. All underwent open reduction, proximal femoral and Dega pelvic osteotomy. The inclusion criteria were: a predominantly spastic form of CP, dislocation of the hip (migration percentage, MP > 80%), Gross Motor Function Classification System, (GMFCS) grade IV to V, a primary surgical procedure and follow-up of > one year. There were 81 consecutive children (40 girls and 41 boys) in the study. Their mean age was nine years (3.5 to 13.8) and mean follow-up was 5.5 years (1.6 to 15.1). Radiological evaluation included measurement of the MP, the acetabular index (AI), the epiphyseal shaft angle (ESA) and the pelvic femoral angle (PFA). The presence and grade of AVN were assessed radiologically according to the Kruczynski classification. Signs of AVN (grades I to V) were seen in 79 hips (68.7%). A total of 23 hips (18%) were classified between grades III and V. Although open reduction of the hip combined with femoral and Dega osteotomy is an effective form of treatment for children with CP and dislocation of the hip, there were signs of avascular necrosis in about two-thirds of the children. There was a strong correlation between post-operative pain and the severity of the grade of AVN. Cite this article: Bone Joint J 2015;97-B:270–6.
A. Koch, MD, PhD, Orthopaedic Surgeon, M. Idzior, MD, PhD, Orthopaedic Surgeon, M. Jozwiak, MD, PhD, Orthopaedic Surgeon, Professor, A. Szulc, MD, PhD, Orthopaedic Surgeon, Professor, Department of Pediatrics, Orthopedics and Traumatology University of Medical Sciences, ul. 28 czerwca 1956r 135/147 61545 Poznan, Poland. M. Molinska-Glura, PhD, Statistician, Department of Computer Science and Statistics Poznan University of Medical Sciences, ul. Dąbrowskiego 79 (second floor), 60-529 Poznan, Poland. Correspondence should be sent to Dr A. Koch; e-mail: [email protected] ©2015 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.97B2. 34280 $2.00 Bone Joint J 2015;97-B:270–6. Received 5 May 2014; Accepted after revision 27 October 2014
270
Avascular necrosis (AVN) of the epiphysis of the femoral head is a severe complication of the treatment of developmental dysplasia of the hip (DDH).1-3 There is little information about AVN as a complication of the treatment of children with cerebral palsy (CP) and dislocation of the hip.4,5 There is no detailed information about the classification of AVN in these children and therefore no data about the severity of the changes.5-8 The most common system used for the classification of AVN in children is that of Kalamachi and MacEwen,4 which was described in 1980 and modified by Kruczynski in 1996.3 According to Kruczynski,3 AVN grades I and II involve mild to moderate changes of the femoral head, and grades III to V involve severe changes (Fig. 1). The purpose of our study was to evaluate the incidence of AVN in children with CP who have been treated for dislocation of the hip and to identify risk factors associated with this complication.
Patients and Methods In this retrospective cohort study, we reviewed the medical records of 620 children with CP
who were treated surgically between 1998 and 2008. A detailed analysis of all those who had been treated for dislocation of the hip was performed. Inclusion criteria were: a predominantly spastic form of pathology assessed by the Ashworth scale (modified by Bohannon and Smith9), high subluxation or dislocation of the hip (Reimer's hip migration10 percentage (MP) > 80%) on one or both sides, Gross Motor Function Classification System (GMFCS)11 level IV or V, open reduction of the hip as primary surgical treatment, and followup of more than one year. Exclusion criteria were: athetotic or dyskinetic forms of CP, subluxation of the hip joint (MP < 80%), GMFCS levels I to III, and open reduction of the hip not as primary surgical treatment. A group of 81 consecutive children (40 girls, 41 boys) meeting the inclusion criteria were selected. All underwent open reduction of the hip combined with a derotation-varus shortening femoral osteotomy (DVSO) and a Dega transiliac osteotomy.12-16 The mean age of the children at the time of surgery was nine years (3.5 to 13.8). The mean follow-up was 5.5 years (1.6 to 15.1). The 81 children included 115 dislocated hips. There were 47 THE BONE & JOINT JOURNAL
AVASCULAR NECROSIS AS A COMPLICATION OF THE TREATMENT OF DISLOCATION OF THE HIP IN CHILDREN WITH CEREBRAL PALSY
Fig. 1a
Fig. 1b
Fig. 1c
Fig. 1d
271
Fig. 1e
Classification of avascular necrosis as described by Kruczynski. Example radiographs one year after surgery show changes to the femoral head with a) no signs of fragmentation, b) with fragmentation, c) lateral and d) medial part of epiphysis and e) whole growth cartilage.
children (47 dislocated hips, 24 girls and 23 boys) with unilateral dislocation and 34 (68 dislocated hips, 16 girls and 18 boys) with bilateral dislocation. There were 20 children (nine unilateral and 11 bilateral dislocations) with GMFCS level IV and 61 (38 unilateral and 23 bilateral dislocations) with GMFCS level V. The first stage involves bilateral release of the adductors, psoas, and hamstrings, distal lengthening, and neurectomy of the anterior branch of obturator nerve on the side of the dislocation. The hip is exposed using the Watson–Jones lateral approach13,14 from the anterior iliac spine to the lateral aspect of the thigh. The gluteal muscles and the capsular attachment of the rectus femoris are detached and the capsule of the hip opened. The ligamentum teres, the transverse ligament, and connective tissue within the joint are removed. Shortening of the femur is defined by the distance between the upper edge of the acetabulum and the most proximally positioned part of the femoral head. The DVSO is usually performed with femoral shortening of between 1 cm and 4 cm. The hip is reduced and the Dega osteotomy is based on a semi-circular osteotomy of the ilium beginning 1 cm to 2 cm proximal to the edge of the acetabulum and passing deeper, parallel to its internal surface. The osteotomy passes through the outer and inner surfaces of the iliac cortex but does not involve the internal cortex in the region of the sciatic notch. The osteotomised portion, with its pedicle, is transposed distally to form the roof of the acetabulum. It is stabilised using bone graft taken from the femur during the shortening osteotomy (Fig. 2). The children are immobilised post-operatively in a spica for four to six weeks, with the hip in 20° to 25° of abduction, 10° to 15° of flexion, and slight internal rotation of the hip. Clinical and radiological data were collected prior to surgery, one year post-operatively and at final review. The ranges of movement (ROM) of the hip and knee joints were recorded using a goniometer. Spasticity was graded according to the Ashworth scale.9,17 Intra-operative data such as the angle of anteversion, the neck–shaft angle, and their corrections were collected using a goniometer with Kirschner wire markers. Radiological evaluation was performed using standard anterior-posterior radiographs of the pelvis. The MP,9 acetabular index (AI),10 epiphyseal shaft angle VOL. 97-B, No. 2, FEBRUARY 2015
(ESA),18 and pelvic femoral angle (PFA)19 were recorded pre-operatively and at final follow-up. The presence and grade of AVN was evaluated according to Kruczynski’s classification.3 This grade can be made no sooner than one year after surgery. Subsequent remodelling of the proximal femur can interefere with the classification. The shape of the femoral head was classified according to Miller et al:8 grade 1 involves a round femoral head, grade 2 a femoral head with medial or lateral flattening, and in grade 3 there is both medial and lateral flattening of the femoral head. Pain was assessed using the NRS-11,20 in which patients or parents grade the pain. For children who were unable to answer the question, parents completed the questionnaire on the basis of the child’s behaviour. Assessment of pain was performed on the day of examination and focused on the dislocated side in children with unilateral dislocation and on both sides in those with bilateral dislocation. The assessment of pain was undertaken at the final follow-up, when the children described the pain from the operation(s) on their hip. Statistical analysis. Both the Student's t-test and chisquared test were used to compare the two groups and analysis of variance (ANOVA) was used when comparing more than two groups. Correlation between parameters was evaluated by Pearson’s correlation coefficient (r). Differences were measured using the Wilcoxon test, and considered significant with p-values < 0.05.
Results Signs of AVN (grades I to V) were observed in 79 dislocated hips (68.7%). A total of 23 of the dislocated hips (20%) were classified between grades III and V and were thought to be the most severe. There were no signs of AVN in 36 dislocated hips (31.3%). An increase in the ROM of the hip, especially in abduction, was observed in all children. A reduced flexion contracture and popliteal angle was seen in all children. Some children had reduced flexion of the knee; there was an increased incidence of shortening of the rectus femoris muscle (positive Duncan-Ely test). The mean flexion contracture of the hip improved from 22.3º (0º to 60º) to 9.6º (0º to 40º) and the mean abduction improved from 12.5º (-25º to 40º)
272
A. KOCH, M. JOZWIAK, M. IDZIOR, M. MOLINSKA-GLURA, A. SZULC
Fig. 2 Diagram of the Dega pelvic osteotomy.
Table II. Changes in the level of pain (NRS-11 grade)
Table I. Changes in the mean spasticity according to the Ashworth scale9 (range)
Hip flexors Hip adductors Knee flexors Knee extensors
Pre-operative
At final follow up
2.32 (0 to 4) 2.87 (1 to 4) 2.67 (0 to 4) 1.08 (0 to 4)
1.33 (0 to 4) 1.95 (0 to 4) 1.98 (0 to 4) 1.53 (0 to 4)
Grade of pain
Pre-operatively
At final follow-up
No pain < Grade 2 Grade 3 to 4 > Grade 4
48 children (69 hips) 17 children (22 hips) 13 children (20 hips) 3 children (4 hips)
59 children (82 hips) 9 children (17 hips) 8 children (10 hips) 5 children (6 hips)
Table III. Changes in intra-operative measurements Parameters of the proximal femur
Mean (°)
Range (°)
SD
Neck–shaft angle, pre-operatively At final follow-up Anteversion angle, pre-operatively At final follow-up
141.60° 119.26° 63.26° 23.47°
(120° to 170°) (100° to 140°) (5° to 90°) (5° to 55°)
11.09 7.75 16.87 11.7
SD,
standard deviation
to 24º (-15º to 70º). These changes were statistically significant (all p < 0,001, Wilcoxon test). Before surgery, a positive Ely test was observed in 49 hips (42.6%). At final follow-up, shortening of the rectus femoris was found in 79 hips (68.7%). A statistically significant decrease in the mean spasticity of the flexors and adductors of the hip, and the flexors of the knee was noticed at final follow-up. (p < 0.001, Wilcoxon test) (Table I). The only increase in spasticity was observed in the rectus femoris muscle. A total of 33 children (41%) had pain pre-operatively and 23 (28.4%) had pain at final examination (Table II). These changes were statistically significant in the whole group and also in those with uni- and bilateral dislocations (p < 0.001; Wilcoxon test). The changes in the shape of the proximal femur are shown in Table III. In summary, the mean change of the
neck–shaft angle was 22º (0º to 40º) and of anteversion was 40º (0º to 80º) and the mean femoral shortening was 1.82 cm (1 cm to 4 cm). The radiological changes are shown in Table IV. During the study period, two re-dislocations (MP > 80%) and six severe subluxations (50% < MP < 80%) occurred. A total of 36 hips (31.3%) had no signs of AVN; 34 (29.5%) had grade I and 22 (19.1%) grade II, both assessments considered to be mild changes. A total of seven hips (6%) had grade III, 11 (9.5%) had grade IV, and five (4.3%) had grade V AVN. Four surgeons (MJ, MI, AK, AS) assessed post-operative radiographs on two occasions with a two-week interval and the mean inter-observer repeatability was 0.84 (95% confidence interval (CI) 0.79 to 0.87) and the intra-observer reliability was 0.87 (95% CI 0.81 to 0.91). Agreement was arrived at by consensus. An example of the surgical procedure is shown in Figure 3. THE BONE & JOINT JOURNAL
AVASCULAR NECROSIS AS A COMPLICATION OF THE TREATMENT OF DISLOCATION OF THE HIP IN CHILDREN WITH CEREBRAL PALSY
273
Table IV. Radiological parameters Radiological parameters
No. of hip joints
Pre-operative mean (range) (º)
SD
Migration percentage (%) Acetabular index (º) Epiphyseal shaft angle (º) Pelvic femoral angle (º)
115 115 115 115
98.33% (80° to 100°) 30.73º (10° to 62°) 82.83º (54° to 107°) -6.33º (-40° to10°)
4.76º 8.53º 9.17º 8.64º
SD,
(º)
Post-operative mean (range) (º)
SD
16.00% (0° to 100°) 21.25º (3° to 50°) 54.09º (21° to 90°) -2.27º (-46° to 45°)
20.8º 9.27º 16.67º 13.6º
(º)
Statistical significance (Wilcoxon) p < 0.001 p < 0.001 p < 0.001 p = 0.005
standard deviation
Fig. 3a
Fig. 3b
Pelvic radiographs showing a) a 9.5 year old girl pre-operatively with bilateral spastic hip dislocation and b) post-operatively at 3.5 years following surgery with a Dega acetabular osteotomy and proximal femoral osteotomy secured with a blade plate, revealing Kruczynski classification of avascular necrosis grade 4 bilaterally; Miller’s femoral head shape classification was grade 2 bilaterally prior to surgery and grade 3 bilaterally post-surgery.
95
ESA after surgery
85 75 65 55 45 35 25 15 0
1 2 Femoral head shape according to Miller
3
Fig. 4 Graph showing the correlation between the epiphyseal shaft angle (ESA) post-operatively and the shape of the femoral head according to Miller et al.8
Using Miller's classification, the shape of the femur was grade I in 56 (49%) and grade II in 59 hips (51%), preoperatively. At final follow-up, 45 hips (39%) were grade I, 56 (49%) were grade II and 14 (12%) were grade III. The deterioration in the shape of the femoral head was statistically significant (p < 0.001, Wilcoxon test). VOL. 97-B, No. 2, FEBRUARY 2015
The final value of ESA correlates with the severity of AVN (p < 0.0001, r = -0.347) and with the shape of the femoral head (p < 0.001, r = -0.350). In those hips with ESA values less than the normal value between 65° and 75°, AVN and deterioration in the shape of the femoral head were seen more frequently (Fig. 4). The correlations between the PFA and the incidence of AVN (p < 0.001, r = -0.328) and between the PFA and the shape of the femoral head were statistically significant (p = 0.001, r = -0.328) (Fig. 5). The age of the child at the time of surgery appears to be a risk factor for the development of changes in the shape of the femoral head. These changes were seen more frequently in older children (p < 0.001, r = 0.410). Hips requiring more shortening during surgery were at a higher risk of deterioration in the shape of the femoral head (p = 0.036, r = 0.195) (Fig. 6) and were more likely to develop AVN (p = 0.006, r = 0.251) (Fig. 7) than those requiring less shortening. There was a strong positive correlation between the incidence of AVN and the PFA of the hip and a strong negative correlation between the contralateral PFA (p < 0,001, r = -0,328) The appearance of a windblown deformity post-operatively appeared to be the main factor leading to changes in the shape of the femoral head, and is primarily responsible for a worse final outcome.
274
A. KOCH, M. JOZWIAK, M. IDZIOR, M. MOLINSKA-GLURA, A. SZULC
4
60
3.6 Femoral shortening
Final PFA value
40 20 0 -20 -40
3.2 2.8 2.4 2 1.6 1.2 0.8
-60
0.4 0
0
1 2 3 4 5 AVN type according to Kruczynski
1
2
3
4
5
AVN type according to Kruczynski
Fig. 5
Fig. 7
Graph showing the correlation between the pelvic femoral angle (PFA) post-operatively and avascular necrosis (AVN).
Graph showing the correlation between the amount of femoral shortening and the grade of avascular necrosis (AVN) according to Kruczynski.3
4.5
3.5 Neck – shaft angle (°)
Femur bone shortening
4
3 2.5 2 1.5 1 0.5 0 0
1 2 Head shape according to Miller et al
3
165 160 155 150 145 140 135 130 125 120 115 0
1 2 3 4 AVN type according to Kruczynski
5
Fig. 6
Fig. 8
Graph showing the correlation between the amount of femoral shortening and the shape of the femoral head according to Miller et al.8
Graph showing the correlation between the neck–shaft angle post-operatively and the grade of avascular necrosis (AVN) according to Kruczynski.3
Children with a flexion contracture of the hip and limited abduction post-operatively are more prone to deterioration in the shape of the femoral head (p = 0.006, r = 0.25, and p = 0.01, r = -0.28, respectively) and AVN (p = 0.012, r = 0.240, and p = 0.004, r = -0.274, respectively). Additionally, spasticity of the iliopsoas and adductor muscles is a risk factor for deterioration in the shape of the femoral head and the development of AVN (p < 0.001, r = 0.32 and p = 0.002, r = 0.28, respectively). The pre-operative NSA had a negative correlation with the incidence of AVN (p = 0.034, r = -0.197). A higher preoperative NSA is thus associated with a lower incidence of AVN (Fig. 8). There was a positive correlation between the shape of the femoral head pre-operatively and the development of AVN. A grade 2 pre-operative shape of the femoral head is a risk factor for the development of AVN (p < 0.001, r = 0.337). Post-operatively the shape of the femoral head is strongly correlated with development of AVN (p < 0.001, r = 0.812).
There was a correlation between the pre- and postoperative levels of pain and the amount of femoral shortening obtained at operation (p = 0.025, r = 0.208, and p = 0.035, r = 0.196, respectively). The relationship between the levels of pain and the incidence of AVN was statistically significant (p < 0.001, r = 0.279) (Fig. 9). There were no major complications, although two children had a femoral fracture treated by cast immobilisation.
Discussion The treatment of a child with CP and dislocation of the hip continues after reduction of the dislocation. The postoperative course depends on the stability of the hip and the appearance of well-recognised complications such as AVN, and is strongly correlated with the presence of risk factors. No previous studies have considered an analysis of the risk factors. The evaluation of the outcome after surgery must include the incidence of AVN which can be undertaken using three classifications.1-3 However, all three were iniTHE BONE & JOINT JOURNAL
AVASCULAR NECROSIS AS A COMPLICATION OF THE TREATMENT OF DISLOCATION OF THE HIP IN CHILDREN WITH CEREBRAL PALSY
Pain appearance
10 8 6 4 2 0 0
1
2
3
4
5
Type of AVN accordnig to Kruczynskio Fig. 9 Graph showing the correlation between the grade of avascular necrosis (AVN) according to Kruczynski3 and pain measured accordind to the NRS-11 grade.20
tially described for children with DDH. We found that the incidence of AVN after the surgical treatment of dislocation of the hip in children with CP was 68.7%. This high incidence may reflect the strict inclusion criteria of this study. The incidence of AVN in these children has been reported to vary between 0% and 75%.4-6,15,21,22 Mubarack et al7 reported an 8% incidence of AVN in a series of 18 children with a mean follow-up of 82 months. However, Khalife et al5 reported an incidence of about 37% in a series of 50 children with a mean follow-up of six years. Such differences in the incidence of AVN are probably due to the heterogeneity of the studies, especially the pre-operative value of the MP. Our study group consisted of an extremely homogenous group of 81 children with 115 complete dislocations of the hip. Our results cannot be easily compared with those of the previous studies because of the strict inclusion criteria, which limited our cohort to patients with GMFCS IV and V and a pre-operative MP of > 80%. The incidence of AVN in children with a GMFCS of > III and a poor prognosis was only 17%. There was no need to confirm the assessment of AVN by MRI, because Kruczynski’s classification is based only on radiographs, and concurrent validity with MRI scans has previously been reported.3 We can identify pre-operative risk factors for the development of AVN as: a shape of the femoral head of > grade I, age at the time of surgery of > eight years, and a NSA of < 140º. Explanations of the first and second factors seem to be obvious. A grade 2 shape of the femoral head may be considered as a pre-AVN stage. Pre-operative changes in the shape of the femoral head may be due to vascular abnormalities caused by a change in hip position, tension in the joint capsule and pressure within the joint. It is harder to explain the third correlation. It appears that a valgus deformity of the proximal femur should be considered as a quasi-protective factor for the prevention of AVN. An explanation for this hypothesis was proposed by Miller et al23 concerning forces generated in the hip in CP. Although surgery was undertaken, risk factors leading to dislocation remained. We have identified the following risk factors that are present after surgical treatment in these children: a windblown deformity, a low ESA, limited VOL. 97-B, No. 2, FEBRUARY 2015
275
abduction, a flexion contracture, and spasticity of the flexors and adductors of the hip. Improvement in the ROM in the sagittal and coronal planes seems to predict a good outcome. The correlation between the post-operative MP and abduction of the hip is a good explanation for how limited abduction and spasticity of the adductors are contributing factors to the development of dislocation. This assumption corresponds to the theory of dislocation of the hip in CP proposed by Miller et al.23 Additionally, correlations between AI, abduction of the hip, the flexion contracture, and spasticity of the flexors and extensors of the hip can be interpreted in the same way. The persistence of risk factors post-operatively will result in recurrent dislocation. The correlations between the development of AVN, the post-operative MP and intra-operative femoral shortening are surprising. In our opinion, the severity of dislocation correlates with the risk of AVN. The correlation between the final MP and the femoral shortening suggests that shortening of > 2 cm will protect the hip from redislocation. AVN is commonly seen in children with shortening of > 2 cm. This can be explained by the requirement of more shortening in hips with high dislocation. Proximal femoral osteotomy with great intra-operative shortening may induce injury or excessive tension in the nutritional arteries originating at the medial circumflex artery of the thigh. These vessels are mostly found in the joint capsule that has not been removed during open reduction of the hip. However, this is only a hypothesis, as no vascular studies were undertaken. Similar observations were made by Khalife et al,5 who also noticed a correlation between the magnitude of shortening and development of AVN.5 The frequency of pain in spastic hip disease remains controversial. In a series of 73 children, Jóźwiak et al24 showed that 55% of children with spastic dislocation of the hip were in pain, and the incidence reported in the literature varied between 25% and 55%.25-29 In our study, 40% of the children had pain. Our data differ slightly from those presented between 1998 to 2002,23 due to strict criteria excluding patients primarily treated with proximal femoral resection. Limited communication skills are a major difficulty in the evaluation of pain in children with CP. There are many scales used for this assessment such as the NRS-11 and Visual Analogue Scale, which need the commitment and cooperation of the child.20 Verbal scales seem not to be effective for children at GMFCS levels IV and V. The FLACC scale (face, legs, activity, cry, and consolability),30 commonly used in paediatric oncology and intensive care units, cannot be used for children with CP.25 The NRS-11 scale is easy and acceptable by children and parents. Parents can often report on the basis of idiosyncratic signs like crying, moaning and groaning.25,27-29 The operative treatment of dislocation of the hip in children with CP is effective in reducing pain. A statistically significant mean reduction of pain from 42% to 27% was noticed (p < 0.001). Similarly, Root et al31 and Sankar et al32
276
A. KOCH, M. JOZWIAK, M. IDZIOR, M. MOLINSKA-GLURA, A. SZULC
reported reduced levels of pain after open reduction of the hip. Not all children, however, have reduced levels of pain; Maslon et al33 suggested that persistent pain was due to sensitisation of the nociceptors in the joint capsule and ligamentum teres. This study has limitations. We only evaluated the results of one form of surgical treatment, performed in one department. Another limitation is the strict inclusion criteria of MP > 80% and GMFCS grades IV and V, preventing a prediction of the influence of MP values of < 80% and of the results in GMFCS grades I to III. Additionally, such a precise analysis of the results of treatment done in a very homogenous group of children with uni- or bilateral spastic hip dislocation cannot be generalised or compared with the results of other authors. Children who undergo open reduction and DVSO and a Dega osteotomy require long-term follow-up. The high incidence of AVN found in our study is due to the sensitive classification system used. In the Kruczynski classification,3 only grades III to V are considered severe with a poor prognosis; in these children there is less proximal remodelling than in grades I to II. Children with a high dislocation, aged > eight years, and those with pre-operative alteration of the shape of the femoral head are at risk of developing AVN. In such children, palliative treatment rather than reconstructive surgery should be considered. We recommend proximal femoral abduction osteotomy with subcapital resection instead of proximal femoral resection in these children. Author contributions A. Koch: Study design, Data collection, Data analysis, Performed surgeries, Writing paper, Literature search. M. Jozwiak: Study design, Data collection, Data analysis, Performed surgeries, Writing paper, Literature search. M. Idzior: Study design, Data collection, Data analysis, Performed surgeries, Writing paper, Literature search. M. Molinska-Glura: Study design, Data analysis, Writing paper, Literature search. A. Szulc: Study design, Data collection, Data analysis, Performed surgeries, Writing paper, Literature search. The authors thank Proper Medical Writing for the language assistance provided in the preparation of this paper. Our paper has a grant from the Polish government: number N N 403 364 639. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by J. Scott and first proof edited by G. Scott.
References 1. Bucholz R, Ogden J. Patterns of ischemic necrosis of the proximal femur in nonoperatively treated congenital hip disease. Proceedings of the 6th Scientific Meeting of the Hip Society; Mosby St Louis.1978:43–63. 2. Beaty JH MD, Kasser JR MD (Eds), Rockwood CA. Rockwood and Wilkins’ Fractures in Children. fifth ed. Lippincott Williams & Wilkins, 2001. 3. Kruczynski J. Avascular necrosis of the proximal femur in developmental dislocation of the hip. Incidence, risk factors, sequelae and MR imaging for diagnosis and prognosis. Acta Orthop Scand Suppl 1996;268:1–48. 4. Kalamchi A, MacEwen GD. Avascular necrosis following treatment of congenital dislocation of the hip. J Bone Joint Surg [Am] 1980;62-A:876–888. 5. Khalife R, Ghanem I, El Hage S, Dagher F, Kharrat K. Risk of recurrent dislocation and avascular necrosis after proximal femoral varus osteotomy in children with cerebral palsy. J Pediatr Orthop B 2010;19:32–37.
6. McNerney NP, Mubarak SJ, Wenger DR. One-stage correction of the dysplastic hip in cerebral palsy with the San Diego acetabuloplasty: results and complications in 104 hips. J Pediatr Orthop 2000;20:93–103. 7. Mubarak SJ, Valencia FG, Wenger DR. One-stage correction of the spastic dislocated hip. Use of pericapsular acetabuloplasty to improve coverage. J Bone Joint Surg [Am] 1992;74-A:1347–1357. 8. Miller F, Girardi H, Lipton G, et al. Reconstruction of the dysplastic spastic hip with peri-ilial pelvic and femoral osteotomy followed by immediate mobilization. J Pediatr Orthop 1997;17:592–602. 9. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther 1987;67:206–207. 10. 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. 11. Robin J, Graham HK, Baker R, et al. A classification system for hip disease in cerebral palsy. Dev Med Child Neurol 2009;51:183–192. 12. Krebs A, Strobl WM, Grill F. Neurogenic hip dislocation in cerebral palsy: quality of life and results after hip reconstruction. J Child Orthop 2008;2:125–131. 13. Dega W. Transiliac osteotomy in the treatment of congenital hip dysplasia. Chir Narzadow Ruchu Ortop Pol 1974;39:601–613. (In Polish). 14. Joziwak M, Marciniak W, Piontek T, Pietrzak S. Dega's transiliac osteotomy in the treatment of spastic hip subluxation and dislocation in cerebral palsy. J Pediatr Orthop B 2000;9:257–264. 15. Robb JE, Brunner R. A Dega-type osteotomy after closure of the triradiate cartilage in non-walking patients with severe cerebral palsy. J Bone Joint Surg [Br] 2006;88B:933–937. 16. Karlen JW, Skaggs DL, Ramachandran M, Kay RM. The Dega osteotomy: a versatile osteotomy in the treatment of developmental and neuromuscular hip pathology. J Pediatr Orthop 2009;29:676–682. 17. Palisano R, Rosenbaum P, Walter S, et al. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol 1997;39:214–223. 18. Sharp IK. Acetabular dysplasia: the acetabular angle. J Bone Joint Surg [Br] 1961;43-B:268–272. 19. Jones DA. Sub-capital coxa valga after varus osteotomy for congenital dislocation of the hip. J Bone Joint Surg [Br] 1977;59-B:152–158. 20. Jensen MP, Engel JM, McKearnan KA, Hoffman AJ. Validity of pain intensity assessment in persons with cerebral palsy: a comparison of six scales. J Pain 2003;4:56–63. 21. Abel MF, Blanco JS, Pavlovich L, Damiano DL. Asymmetric hip deformity and subluxation in cerebral palsy: an analysis of surgical treatment. J Pediatr Orthop 1999;19:479–485. 22. Al-Ghadir M, Masquijo JJ, Guerra LA, Willis B. Combined femoral and pelvic osteotomies versus femoral osteotomy alone in the treatment of hip dysplasia in children with cerebral palsy. J Pediatr Orthop 2009;29:779–783. 23. Miller F, Slomczykowski M, Cope R, Lipton GE. Computer modeling of the pathomechanics of spastic hip dislocation in children. J Pediatr Orthop 1999;19:486–492. 24. Jóźwiak M, Harasymczuk P, Koch A, Kotwicki T. Incidence and risk factors of hip joint pain in children with severe cerebral palsy. Disabil Rehabil 2011;33:1367– 1372. 25. Noonan KJ. Natural history of hip function in cerebral palsy. European Pediatric Orthopaedic Society 2005 Annual Meeting, Mallorca, Spain, 2005 Course Book; Controversies in Cerebral Palsy. 26. Noonan KJ, Jones J, Pierson J, Honkamp NJ, Leverson G. Hip function in adults with severe cerebral palsy. J Bone Joint Surg [Am] 2004;86-A:2607–2613. 27. Pritchett JW. The untreated unstable hip in severe cerebral palsy. Clin Orthop Relat Res 1983;173:169–172. 28. Schwartz L, Engel JM, Jensen MP. Pain in persons with cerebral palsy. Arch Phys Med Rehabil 1999;80:1243–1246. 29. Hodgkinson I, Jindrich ML, Duhaut P, et al. Hip pain in 234 non-ambulatory adolescents and young adults with cerebral palsy: a cross-sectional multicentre study. Dev Med Child Neurol 2001;43:806–808. 30. Willis MH, Merkel SI, Voepel-Lewis T, Malviya S. FLACC Behavioral Pain Assessment Scale: a comparison with the child's self-report. Pediatr Nurs 2003;29:195–198. 31. Root L, Laplaza FJ, Brourman SN, Angel DH. The severely unstable hip in cerebral palsy: treatment with oper reduction, pelvic osteotomy, and femoral osteotomy with shortening. J Bone Joint Surg [Am] 1995;77-A:703–712. 32. Sankar WN, Spiegel DA, Gregg JR, Sennett BJ. Long-term follow-up after onestage reconstruction of dislocated hips in patients with cerebral palsy. J Pediatr Orthop 2006;26:1–7. 33. Maslon A, Jóźwiak M, Pawlak M, Modrzewski T, Grzegorzewski A. Hip joint pain in spastic dislocation: aetiological aspects. Dev Med Child Neurol 2011;53:1019–1023. THE BONE & JOINT JOURNAL
Avascular necrosis as a complication of the treatment of dislocation of the hip in children with cerebral palsy A. Koch, M. Jozwiak, M. Idzior, M. Molinska-Glura, A. Szulc From University of Medical Sciences, Poznan, Poland
We investigated the incidence and risk factors for the development of avascular necrosis (AVN) of the femoral head in the course of treatment of children with cerebral palsy (CP) and dislocation of the hip. All underwent open reduction, proximal femoral and Dega pelvic osteotomy. The inclusion criteria were: a predominantly spastic form of CP, dislocation of the hip (migration percentage, MP > 80%), Gross Motor Function Classification System, (GMFCS) grade IV to V, a primary surgical procedure and follow-up of > one year. There were 81 consecutive children (40 girls and 41 boys) in the study. Their mean age was nine years (3.5 to 13.8) and mean follow-up was 5.5 years (1.6 to 15.1). Radiological evaluation included measurement of the MP, the acetabular index (AI), the epiphyseal shaft angle (ESA) and the pelvic femoral angle (PFA). The presence and grade of AVN were assessed radiologically according to the Kruczynski classification. Signs of AVN (grades I to V) were seen in 79 hips (68.7%). A total of 23 hips (18%) were classified between grades III and V. Although open reduction of the hip combined with femoral and Dega osteotomy is an effective form of treatment for children with CP and dislocation of the hip, there were signs of avascular necrosis in about two-thirds of the children. There was a strong correlation between post-operative pain and the severity of the grade of AVN. Cite this article: Bone Joint J 2015;97-B:270–6.
A. Koch, MD, PhD, Orthopaedic Surgeon, M. Idzior, MD, PhD, Orthopaedic Surgeon, M. Jozwiak, MD, PhD, Orthopaedic Surgeon, Professor, A. Szulc, MD, PhD, Orthopaedic Surgeon, Professor, Department of Pediatrics, Orthopedics and Traumatology University of Medical Sciences, ul. 28 czerwca 1956r 135/147 61545 Poznan, Poland. M. Molinska-Glura, PhD, Statistician, Department of Computer Science and Statistics Poznan University of Medical Sciences, ul. Dąbrowskiego 79 (second floor), 60-529 Poznan, Poland. Correspondence should be sent to Dr A. Koch; e-mail: [email protected] ©2015 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.97B2. 34280 $2.00 Bone Joint J 2015;97-B:270–6. Received 5 May 2014; Accepted after revision 27 October 2014
270
Avascular necrosis (AVN) of the epiphysis of the femoral head is a severe complication of the treatment of developmental dysplasia of the hip (DDH).1-3 There is little information about AVN as a complication of the treatment of children with cerebral palsy (CP) and dislocation of the hip.4,5 There is no detailed information about the classification of AVN in these children and therefore no data about the severity of the changes.5-8 The most common system used for the classification of AVN in children is that of Kalamachi and MacEwen,4 which was described in 1980 and modified by Kruczynski in 1996.3 According to Kruczynski,3 AVN grades I and II involve mild to moderate changes of the femoral head, and grades III to V involve severe changes (Fig. 1). The purpose of our study was to evaluate the incidence of AVN in children with CP who have been treated for dislocation of the hip and to identify risk factors associated with this complication.
Patients and Methods In this retrospective cohort study, we reviewed the medical records of 620 children with CP
who were treated surgically between 1998 and 2008. A detailed analysis of all those who had been treated for dislocation of the hip was performed. Inclusion criteria were: a predominantly spastic form of pathology assessed by the Ashworth scale (modified by Bohannon and Smith9), high subluxation or dislocation of the hip (Reimer's hip migration10 percentage (MP) > 80%) on one or both sides, Gross Motor Function Classification System (GMFCS)11 level IV or V, open reduction of the hip as primary surgical treatment, and followup of more than one year. Exclusion criteria were: athetotic or dyskinetic forms of CP, subluxation of the hip joint (MP < 80%), GMFCS levels I to III, and open reduction of the hip not as primary surgical treatment. A group of 81 consecutive children (40 girls, 41 boys) meeting the inclusion criteria were selected. All underwent open reduction of the hip combined with a derotation-varus shortening femoral osteotomy (DVSO) and a Dega transiliac osteotomy.12-16 The mean age of the children at the time of surgery was nine years (3.5 to 13.8). The mean follow-up was 5.5 years (1.6 to 15.1). The 81 children included 115 dislocated hips. There were 47 THE BONE & JOINT JOURNAL
AVASCULAR NECROSIS AS A COMPLICATION OF THE TREATMENT OF DISLOCATION OF THE HIP IN CHILDREN WITH CEREBRAL PALSY
Fig. 1a
Fig. 1b
Fig. 1c
Fig. 1d
271
Fig. 1e
Classification of avascular necrosis as described by Kruczynski. Example radiographs one year after surgery show changes to the femoral head with a) no signs of fragmentation, b) with fragmentation, c) lateral and d) medial part of epiphysis and e) whole growth cartilage.
children (47 dislocated hips, 24 girls and 23 boys) with unilateral dislocation and 34 (68 dislocated hips, 16 girls and 18 boys) with bilateral dislocation. There were 20 children (nine unilateral and 11 bilateral dislocations) with GMFCS level IV and 61 (38 unilateral and 23 bilateral dislocations) with GMFCS level V. The first stage involves bilateral release of the adductors, psoas, and hamstrings, distal lengthening, and neurectomy of the anterior branch of obturator nerve on the side of the dislocation. The hip is exposed using the Watson–Jones lateral approach13,14 from the anterior iliac spine to the lateral aspect of the thigh. The gluteal muscles and the capsular attachment of the rectus femoris are detached and the capsule of the hip opened. The ligamentum teres, the transverse ligament, and connective tissue within the joint are removed. Shortening of the femur is defined by the distance between the upper edge of the acetabulum and the most proximally positioned part of the femoral head. The DVSO is usually performed with femoral shortening of between 1 cm and 4 cm. The hip is reduced and the Dega osteotomy is based on a semi-circular osteotomy of the ilium beginning 1 cm to 2 cm proximal to the edge of the acetabulum and passing deeper, parallel to its internal surface. The osteotomy passes through the outer and inner surfaces of the iliac cortex but does not involve the internal cortex in the region of the sciatic notch. The osteotomised portion, with its pedicle, is transposed distally to form the roof of the acetabulum. It is stabilised using bone graft taken from the femur during the shortening osteotomy (Fig. 2). The children are immobilised post-operatively in a spica for four to six weeks, with the hip in 20° to 25° of abduction, 10° to 15° of flexion, and slight internal rotation of the hip. Clinical and radiological data were collected prior to surgery, one year post-operatively and at final review. The ranges of movement (ROM) of the hip and knee joints were recorded using a goniometer. Spasticity was graded according to the Ashworth scale.9,17 Intra-operative data such as the angle of anteversion, the neck–shaft angle, and their corrections were collected using a goniometer with Kirschner wire markers. Radiological evaluation was performed using standard anterior-posterior radiographs of the pelvis. The MP,9 acetabular index (AI),10 epiphyseal shaft angle VOL. 97-B, No. 2, FEBRUARY 2015
(ESA),18 and pelvic femoral angle (PFA)19 were recorded pre-operatively and at final follow-up. The presence and grade of AVN was evaluated according to Kruczynski’s classification.3 This grade can be made no sooner than one year after surgery. Subsequent remodelling of the proximal femur can interefere with the classification. The shape of the femoral head was classified according to Miller et al:8 grade 1 involves a round femoral head, grade 2 a femoral head with medial or lateral flattening, and in grade 3 there is both medial and lateral flattening of the femoral head. Pain was assessed using the NRS-11,20 in which patients or parents grade the pain. For children who were unable to answer the question, parents completed the questionnaire on the basis of the child’s behaviour. Assessment of pain was performed on the day of examination and focused on the dislocated side in children with unilateral dislocation and on both sides in those with bilateral dislocation. The assessment of pain was undertaken at the final follow-up, when the children described the pain from the operation(s) on their hip. Statistical analysis. Both the Student's t-test and chisquared test were used to compare the two groups and analysis of variance (ANOVA) was used when comparing more than two groups. Correlation between parameters was evaluated by Pearson’s correlation coefficient (r). Differences were measured using the Wilcoxon test, and considered significant with p-values < 0.05.
Results Signs of AVN (grades I to V) were observed in 79 dislocated hips (68.7%). A total of 23 of the dislocated hips (20%) were classified between grades III and V and were thought to be the most severe. There were no signs of AVN in 36 dislocated hips (31.3%). An increase in the ROM of the hip, especially in abduction, was observed in all children. A reduced flexion contracture and popliteal angle was seen in all children. Some children had reduced flexion of the knee; there was an increased incidence of shortening of the rectus femoris muscle (positive Duncan-Ely test). The mean flexion contracture of the hip improved from 22.3º (0º to 60º) to 9.6º (0º to 40º) and the mean abduction improved from 12.5º (-25º to 40º)
272
A. KOCH, M. JOZWIAK, M. IDZIOR, M. MOLINSKA-GLURA, A. SZULC
Fig. 2 Diagram of the Dega pelvic osteotomy.
Table II. Changes in the level of pain (NRS-11 grade)
Table I. Changes in the mean spasticity according to the Ashworth scale9 (range)
Hip flexors Hip adductors Knee flexors Knee extensors
Pre-operative
At final follow up
2.32 (0 to 4) 2.87 (1 to 4) 2.67 (0 to 4) 1.08 (0 to 4)
1.33 (0 to 4) 1.95 (0 to 4) 1.98 (0 to 4) 1.53 (0 to 4)
Grade of pain
Pre-operatively
At final follow-up
No pain < Grade 2 Grade 3 to 4 > Grade 4
48 children (69 hips) 17 children (22 hips) 13 children (20 hips) 3 children (4 hips)
59 children (82 hips) 9 children (17 hips) 8 children (10 hips) 5 children (6 hips)
Table III. Changes in intra-operative measurements Parameters of the proximal femur
Mean (°)
Range (°)
SD
Neck–shaft angle, pre-operatively At final follow-up Anteversion angle, pre-operatively At final follow-up
141.60° 119.26° 63.26° 23.47°
(120° to 170°) (100° to 140°) (5° to 90°) (5° to 55°)
11.09 7.75 16.87 11.7
SD,
standard deviation
to 24º (-15º to 70º). These changes were statistically significant (all p < 0,001, Wilcoxon test). Before surgery, a positive Ely test was observed in 49 hips (42.6%). At final follow-up, shortening of the rectus femoris was found in 79 hips (68.7%). A statistically significant decrease in the mean spasticity of the flexors and adductors of the hip, and the flexors of the knee was noticed at final follow-up. (p < 0.001, Wilcoxon test) (Table I). The only increase in spasticity was observed in the rectus femoris muscle. A total of 33 children (41%) had pain pre-operatively and 23 (28.4%) had pain at final examination (Table II). These changes were statistically significant in the whole group and also in those with uni- and bilateral dislocations (p < 0.001; Wilcoxon test). The changes in the shape of the proximal femur are shown in Table III. In summary, the mean change of the
neck–shaft angle was 22º (0º to 40º) and of anteversion was 40º (0º to 80º) and the mean femoral shortening was 1.82 cm (1 cm to 4 cm). The radiological changes are shown in Table IV. During the study period, two re-dislocations (MP > 80%) and six severe subluxations (50% < MP < 80%) occurred. A total of 36 hips (31.3%) had no signs of AVN; 34 (29.5%) had grade I and 22 (19.1%) grade II, both assessments considered to be mild changes. A total of seven hips (6%) had grade III, 11 (9.5%) had grade IV, and five (4.3%) had grade V AVN. Four surgeons (MJ, MI, AK, AS) assessed post-operative radiographs on two occasions with a two-week interval and the mean inter-observer repeatability was 0.84 (95% confidence interval (CI) 0.79 to 0.87) and the intra-observer reliability was 0.87 (95% CI 0.81 to 0.91). Agreement was arrived at by consensus. An example of the surgical procedure is shown in Figure 3. THE BONE & JOINT JOURNAL
AVASCULAR NECROSIS AS A COMPLICATION OF THE TREATMENT OF DISLOCATION OF THE HIP IN CHILDREN WITH CEREBRAL PALSY
273
Table IV. Radiological parameters Radiological parameters
No. of hip joints
Pre-operative mean (range) (º)
SD
Migration percentage (%) Acetabular index (º) Epiphyseal shaft angle (º) Pelvic femoral angle (º)
115 115 115 115
98.33% (80° to 100°) 30.73º (10° to 62°) 82.83º (54° to 107°) -6.33º (-40° to10°)
4.76º 8.53º 9.17º 8.64º
SD,
(º)
Post-operative mean (range) (º)
SD
16.00% (0° to 100°) 21.25º (3° to 50°) 54.09º (21° to 90°) -2.27º (-46° to 45°)
20.8º 9.27º 16.67º 13.6º
(º)
Statistical significance (Wilcoxon) p < 0.001 p < 0.001 p < 0.001 p = 0.005
standard deviation
Fig. 3a
Fig. 3b
Pelvic radiographs showing a) a 9.5 year old girl pre-operatively with bilateral spastic hip dislocation and b) post-operatively at 3.5 years following surgery with a Dega acetabular osteotomy and proximal femoral osteotomy secured with a blade plate, revealing Kruczynski classification of avascular necrosis grade 4 bilaterally; Miller’s femoral head shape classification was grade 2 bilaterally prior to surgery and grade 3 bilaterally post-surgery.
95
ESA after surgery
85 75 65 55 45 35 25 15 0
1 2 Femoral head shape according to Miller
3
Fig. 4 Graph showing the correlation between the epiphyseal shaft angle (ESA) post-operatively and the shape of the femoral head according to Miller et al.8
Using Miller's classification, the shape of the femur was grade I in 56 (49%) and grade II in 59 hips (51%), preoperatively. At final follow-up, 45 hips (39%) were grade I, 56 (49%) were grade II and 14 (12%) were grade III. The deterioration in the shape of the femoral head was statistically significant (p < 0.001, Wilcoxon test). VOL. 97-B, No. 2, FEBRUARY 2015
The final value of ESA correlates with the severity of AVN (p < 0.0001, r = -0.347) and with the shape of the femoral head (p < 0.001, r = -0.350). In those hips with ESA values less than the normal value between 65° and 75°, AVN and deterioration in the shape of the femoral head were seen more frequently (Fig. 4). The correlations between the PFA and the incidence of AVN (p < 0.001, r = -0.328) and between the PFA and the shape of the femoral head were statistically significant (p = 0.001, r = -0.328) (Fig. 5). The age of the child at the time of surgery appears to be a risk factor for the development of changes in the shape of the femoral head. These changes were seen more frequently in older children (p < 0.001, r = 0.410). Hips requiring more shortening during surgery were at a higher risk of deterioration in the shape of the femoral head (p = 0.036, r = 0.195) (Fig. 6) and were more likely to develop AVN (p = 0.006, r = 0.251) (Fig. 7) than those requiring less shortening. There was a strong positive correlation between the incidence of AVN and the PFA of the hip and a strong negative correlation between the contralateral PFA (p < 0,001, r = -0,328) The appearance of a windblown deformity post-operatively appeared to be the main factor leading to changes in the shape of the femoral head, and is primarily responsible for a worse final outcome.
274
A. KOCH, M. JOZWIAK, M. IDZIOR, M. MOLINSKA-GLURA, A. SZULC
4
60
3.6 Femoral shortening
Final PFA value
40 20 0 -20 -40
3.2 2.8 2.4 2 1.6 1.2 0.8
-60
0.4 0
0
1 2 3 4 5 AVN type according to Kruczynski
1
2
3
4
5
AVN type according to Kruczynski
Fig. 5
Fig. 7
Graph showing the correlation between the pelvic femoral angle (PFA) post-operatively and avascular necrosis (AVN).
Graph showing the correlation between the amount of femoral shortening and the grade of avascular necrosis (AVN) according to Kruczynski.3
4.5
3.5 Neck – shaft angle (°)
Femur bone shortening
4
3 2.5 2 1.5 1 0.5 0 0
1 2 Head shape according to Miller et al
3
165 160 155 150 145 140 135 130 125 120 115 0
1 2 3 4 AVN type according to Kruczynski
5
Fig. 6
Fig. 8
Graph showing the correlation between the amount of femoral shortening and the shape of the femoral head according to Miller et al.8
Graph showing the correlation between the neck–shaft angle post-operatively and the grade of avascular necrosis (AVN) according to Kruczynski.3
Children with a flexion contracture of the hip and limited abduction post-operatively are more prone to deterioration in the shape of the femoral head (p = 0.006, r = 0.25, and p = 0.01, r = -0.28, respectively) and AVN (p = 0.012, r = 0.240, and p = 0.004, r = -0.274, respectively). Additionally, spasticity of the iliopsoas and adductor muscles is a risk factor for deterioration in the shape of the femoral head and the development of AVN (p < 0.001, r = 0.32 and p = 0.002, r = 0.28, respectively). The pre-operative NSA had a negative correlation with the incidence of AVN (p = 0.034, r = -0.197). A higher preoperative NSA is thus associated with a lower incidence of AVN (Fig. 8). There was a positive correlation between the shape of the femoral head pre-operatively and the development of AVN. A grade 2 pre-operative shape of the femoral head is a risk factor for the development of AVN (p < 0.001, r = 0.337). Post-operatively the shape of the femoral head is strongly correlated with development of AVN (p < 0.001, r = 0.812).
There was a correlation between the pre- and postoperative levels of pain and the amount of femoral shortening obtained at operation (p = 0.025, r = 0.208, and p = 0.035, r = 0.196, respectively). The relationship between the levels of pain and the incidence of AVN was statistically significant (p < 0.001, r = 0.279) (Fig. 9). There were no major complications, although two children had a femoral fracture treated by cast immobilisation.
Discussion The treatment of a child with CP and dislocation of the hip continues after reduction of the dislocation. The postoperative course depends on the stability of the hip and the appearance of well-recognised complications such as AVN, and is strongly correlated with the presence of risk factors. No previous studies have considered an analysis of the risk factors. The evaluation of the outcome after surgery must include the incidence of AVN which can be undertaken using three classifications.1-3 However, all three were iniTHE BONE & JOINT JOURNAL
AVASCULAR NECROSIS AS A COMPLICATION OF THE TREATMENT OF DISLOCATION OF THE HIP IN CHILDREN WITH CEREBRAL PALSY
Pain appearance
10 8 6 4 2 0 0
1
2
3
4
5
Type of AVN accordnig to Kruczynskio Fig. 9 Graph showing the correlation between the grade of avascular necrosis (AVN) according to Kruczynski3 and pain measured accordind to the NRS-11 grade.20
tially described for children with DDH. We found that the incidence of AVN after the surgical treatment of dislocation of the hip in children with CP was 68.7%. This high incidence may reflect the strict inclusion criteria of this study. The incidence of AVN in these children has been reported to vary between 0% and 75%.4-6,15,21,22 Mubarack et al7 reported an 8% incidence of AVN in a series of 18 children with a mean follow-up of 82 months. However, Khalife et al5 reported an incidence of about 37% in a series of 50 children with a mean follow-up of six years. Such differences in the incidence of AVN are probably due to the heterogeneity of the studies, especially the pre-operative value of the MP. Our study group consisted of an extremely homogenous group of 81 children with 115 complete dislocations of the hip. Our results cannot be easily compared with those of the previous studies because of the strict inclusion criteria, which limited our cohort to patients with GMFCS IV and V and a pre-operative MP of > 80%. The incidence of AVN in children with a GMFCS of > III and a poor prognosis was only 17%. There was no need to confirm the assessment of AVN by MRI, because Kruczynski’s classification is based only on radiographs, and concurrent validity with MRI scans has previously been reported.3 We can identify pre-operative risk factors for the development of AVN as: a shape of the femoral head of > grade I, age at the time of surgery of > eight years, and a NSA of < 140º. Explanations of the first and second factors seem to be obvious. A grade 2 shape of the femoral head may be considered as a pre-AVN stage. Pre-operative changes in the shape of the femoral head may be due to vascular abnormalities caused by a change in hip position, tension in the joint capsule and pressure within the joint. It is harder to explain the third correlation. It appears that a valgus deformity of the proximal femur should be considered as a quasi-protective factor for the prevention of AVN. An explanation for this hypothesis was proposed by Miller et al23 concerning forces generated in the hip in CP. Although surgery was undertaken, risk factors leading to dislocation remained. We have identified the following risk factors that are present after surgical treatment in these children: a windblown deformity, a low ESA, limited VOL. 97-B, No. 2, FEBRUARY 2015
275
abduction, a flexion contracture, and spasticity of the flexors and adductors of the hip. Improvement in the ROM in the sagittal and coronal planes seems to predict a good outcome. The correlation between the post-operative MP and abduction of the hip is a good explanation for how limited abduction and spasticity of the adductors are contributing factors to the development of dislocation. This assumption corresponds to the theory of dislocation of the hip in CP proposed by Miller et al.23 Additionally, correlations between AI, abduction of the hip, the flexion contracture, and spasticity of the flexors and extensors of the hip can be interpreted in the same way. The persistence of risk factors post-operatively will result in recurrent dislocation. The correlations between the development of AVN, the post-operative MP and intra-operative femoral shortening are surprising. In our opinion, the severity of dislocation correlates with the risk of AVN. The correlation between the final MP and the femoral shortening suggests that shortening of > 2 cm will protect the hip from redislocation. AVN is commonly seen in children with shortening of > 2 cm. This can be explained by the requirement of more shortening in hips with high dislocation. Proximal femoral osteotomy with great intra-operative shortening may induce injury or excessive tension in the nutritional arteries originating at the medial circumflex artery of the thigh. These vessels are mostly found in the joint capsule that has not been removed during open reduction of the hip. However, this is only a hypothesis, as no vascular studies were undertaken. Similar observations were made by Khalife et al,5 who also noticed a correlation between the magnitude of shortening and development of AVN.5 The frequency of pain in spastic hip disease remains controversial. In a series of 73 children, Jóźwiak et al24 showed that 55% of children with spastic dislocation of the hip were in pain, and the incidence reported in the literature varied between 25% and 55%.25-29 In our study, 40% of the children had pain. Our data differ slightly from those presented between 1998 to 2002,23 due to strict criteria excluding patients primarily treated with proximal femoral resection. Limited communication skills are a major difficulty in the evaluation of pain in children with CP. There are many scales used for this assessment such as the NRS-11 and Visual Analogue Scale, which need the commitment and cooperation of the child.20 Verbal scales seem not to be effective for children at GMFCS levels IV and V. The FLACC scale (face, legs, activity, cry, and consolability),30 commonly used in paediatric oncology and intensive care units, cannot be used for children with CP.25 The NRS-11 scale is easy and acceptable by children and parents. Parents can often report on the basis of idiosyncratic signs like crying, moaning and groaning.25,27-29 The operative treatment of dislocation of the hip in children with CP is effective in reducing pain. A statistically significant mean reduction of pain from 42% to 27% was noticed (p < 0.001). Similarly, Root et al31 and Sankar et al32
276
A. KOCH, M. JOZWIAK, M. IDZIOR, M. MOLINSKA-GLURA, A. SZULC
reported reduced levels of pain after open reduction of the hip. Not all children, however, have reduced levels of pain; Maslon et al33 suggested that persistent pain was due to sensitisation of the nociceptors in the joint capsule and ligamentum teres. This study has limitations. We only evaluated the results of one form of surgical treatment, performed in one department. Another limitation is the strict inclusion criteria of MP > 80% and GMFCS grades IV and V, preventing a prediction of the influence of MP values of < 80% and of the results in GMFCS grades I to III. Additionally, such a precise analysis of the results of treatment done in a very homogenous group of children with uni- or bilateral spastic hip dislocation cannot be generalised or compared with the results of other authors. Children who undergo open reduction and DVSO and a Dega osteotomy require long-term follow-up. The high incidence of AVN found in our study is due to the sensitive classification system used. In the Kruczynski classification,3 only grades III to V are considered severe with a poor prognosis; in these children there is less proximal remodelling than in grades I to II. Children with a high dislocation, aged > eight years, and those with pre-operative alteration of the shape of the femoral head are at risk of developing AVN. In such children, palliative treatment rather than reconstructive surgery should be considered. We recommend proximal femoral abduction osteotomy with subcapital resection instead of proximal femoral resection in these children. Author contributions A. Koch: Study design, Data collection, Data analysis, Performed surgeries, Writing paper, Literature search. M. Jozwiak: Study design, Data collection, Data analysis, Performed surgeries, Writing paper, Literature search. M. Idzior: Study design, Data collection, Data analysis, Performed surgeries, Writing paper, Literature search. M. Molinska-Glura: Study design, Data analysis, Writing paper, Literature search. A. Szulc: Study design, Data collection, Data analysis, Performed surgeries, Writing paper, Literature search. The authors thank Proper Medical Writing for the language assistance provided in the preparation of this paper. Our paper has a grant from the Polish government: number N N 403 364 639. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by J. Scott and first proof edited by G. Scott.
References 1. Bucholz R, Ogden J. Patterns of ischemic necrosis of the proximal femur in nonoperatively treated congenital hip disease. Proceedings of the 6th Scientific Meeting of the Hip Society; Mosby St Louis.1978:43–63. 2. Beaty JH MD, Kasser JR MD (Eds), Rockwood CA. Rockwood and Wilkins’ Fractures in Children. fifth ed. Lippincott Williams & Wilkins, 2001. 3. Kruczynski J. Avascular necrosis of the proximal femur in developmental dislocation of the hip. Incidence, risk factors, sequelae and MR imaging for diagnosis and prognosis. Acta Orthop Scand Suppl 1996;268:1–48. 4. Kalamchi A, MacEwen GD. Avascular necrosis following treatment of congenital dislocation of the hip. J Bone Joint Surg [Am] 1980;62-A:876–888. 5. Khalife R, Ghanem I, El Hage S, Dagher F, Kharrat K. Risk of recurrent dislocation and avascular necrosis after proximal femoral varus osteotomy in children with cerebral palsy. J Pediatr Orthop B 2010;19:32–37.
6. McNerney NP, Mubarak SJ, Wenger DR. One-stage correction of the dysplastic hip in cerebral palsy with the San Diego acetabuloplasty: results and complications in 104 hips. J Pediatr Orthop 2000;20:93–103. 7. Mubarak SJ, Valencia FG, Wenger DR. One-stage correction of the spastic dislocated hip. Use of pericapsular acetabuloplasty to improve coverage. J Bone Joint Surg [Am] 1992;74-A:1347–1357. 8. Miller F, Girardi H, Lipton G, et al. Reconstruction of the dysplastic spastic hip with peri-ilial pelvic and femoral osteotomy followed by immediate mobilization. J Pediatr Orthop 1997;17:592–602. 9. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther 1987;67:206–207. 10. 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. 11. Robin J, Graham HK, Baker R, et al. A classification system for hip disease in cerebral palsy. Dev Med Child Neurol 2009;51:183–192. 12. Krebs A, Strobl WM, Grill F. Neurogenic hip dislocation in cerebral palsy: quality of life and results after hip reconstruction. J Child Orthop 2008;2:125–131. 13. Dega W. Transiliac osteotomy in the treatment of congenital hip dysplasia. Chir Narzadow Ruchu Ortop Pol 1974;39:601–613. (In Polish). 14. Joziwak M, Marciniak W, Piontek T, Pietrzak S. Dega's transiliac osteotomy in the treatment of spastic hip subluxation and dislocation in cerebral palsy. J Pediatr Orthop B 2000;9:257–264. 15. Robb JE, Brunner R. A Dega-type osteotomy after closure of the triradiate cartilage in non-walking patients with severe cerebral palsy. J Bone Joint Surg [Br] 2006;88B:933–937. 16. Karlen JW, Skaggs DL, Ramachandran M, Kay RM. The Dega osteotomy: a versatile osteotomy in the treatment of developmental and neuromuscular hip pathology. J Pediatr Orthop 2009;29:676–682. 17. Palisano R, Rosenbaum P, Walter S, et al. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol 1997;39:214–223. 18. Sharp IK. Acetabular dysplasia: the acetabular angle. J Bone Joint Surg [Br] 1961;43-B:268–272. 19. Jones DA. Sub-capital coxa valga after varus osteotomy for congenital dislocation of the hip. J Bone Joint Surg [Br] 1977;59-B:152–158. 20. Jensen MP, Engel JM, McKearnan KA, Hoffman AJ. Validity of pain intensity assessment in persons with cerebral palsy: a comparison of six scales. J Pain 2003;4:56–63. 21. Abel MF, Blanco JS, Pavlovich L, Damiano DL. Asymmetric hip deformity and subluxation in cerebral palsy: an analysis of surgical treatment. J Pediatr Orthop 1999;19:479–485. 22. Al-Ghadir M, Masquijo JJ, Guerra LA, Willis B. Combined femoral and pelvic osteotomies versus femoral osteotomy alone in the treatment of hip dysplasia in children with cerebral palsy. J Pediatr Orthop 2009;29:779–783. 23. Miller F, Slomczykowski M, Cope R, Lipton GE. Computer modeling of the pathomechanics of spastic hip dislocation in children. J Pediatr Orthop 1999;19:486–492. 24. Jóźwiak M, Harasymczuk P, Koch A, Kotwicki T. Incidence and risk factors of hip joint pain in children with severe cerebral palsy. Disabil Rehabil 2011;33:1367– 1372. 25. Noonan KJ. Natural history of hip function in cerebral palsy. European Pediatric Orthopaedic Society 2005 Annual Meeting, Mallorca, Spain, 2005 Course Book; Controversies in Cerebral Palsy. 26. Noonan KJ, Jones J, Pierson J, Honkamp NJ, Leverson G. Hip function in adults with severe cerebral palsy. J Bone Joint Surg [Am] 2004;86-A:2607–2613. 27. Pritchett JW. The untreated unstable hip in severe cerebral palsy. Clin Orthop Relat Res 1983;173:169–172. 28. Schwartz L, Engel JM, Jensen MP. Pain in persons with cerebral palsy. Arch Phys Med Rehabil 1999;80:1243–1246. 29. Hodgkinson I, Jindrich ML, Duhaut P, et al. Hip pain in 234 non-ambulatory adolescents and young adults with cerebral palsy: a cross-sectional multicentre study. Dev Med Child Neurol 2001;43:806–808. 30. Willis MH, Merkel SI, Voepel-Lewis T, Malviya S. FLACC Behavioral Pain Assessment Scale: a comparison with the child's self-report. Pediatr Nurs 2003;29:195–198. 31. Root L, Laplaza FJ, Brourman SN, Angel DH. The severely unstable hip in cerebral palsy: treatment with oper reduction, pelvic osteotomy, and femoral osteotomy with shortening. J Bone Joint Surg [Am] 1995;77-A:703–712. 32. Sankar WN, Spiegel DA, Gregg JR, Sennett BJ. Long-term follow-up after onestage reconstruction of dislocated hips in patients with cerebral palsy. J Pediatr Orthop 2006;26:1–7. 33. Maslon A, Jóźwiak M, Pawlak M, Modrzewski T, Grzegorzewski A. Hip joint pain in spastic dislocation: aetiological aspects. Dev Med Child Neurol 2011;53:1019–1023. THE BONE & JOINT JOURNAL
