Ischiofemoral Impingement in Children: Imaging With Clinical Correlation

Transcription

1 Pediatric Imaging Original Research Stenhouse et al. MRI of Pediatric Ischiofemoral Impingement Pediatric Imaging Original Research Gregor Stenhouse 1,2 Scott Kaiser 1,3 Simon P. Kelley 4 Jennifer Stimec 1 Stenhouse G, Kaiser S, Kelley SP, Stimec J Keywords: ischiofemoral impingement, ischiofemoral space, quadratus femoris edema DOI: /AJR Received April 20, 2015; accepted after revision July 6, Supported by a research grant from The Hospital for Sick Children for financial assistance with statistical support. 1 Department of Radiology, The Hospital for Sick Children, Toronto, ON, Canada. 2 Present address: Department of Radiology, Royal United Hospital Bath, Combe Park, Bath, Somerset, BA1 3NG, UK. Address correspondence to G. Stenhouse (gregorstenhouse@nhs.net). 3 Present address: Department of Orthopaedic Surgery, University of California, Benioff Children s Hospitals, San Francisco, CA. 4 Department of Surgery, The Hospital for Sick Children, Toronto, ON, Canada. AJR 2016; 206: X/16/ American Roentgen Ray Society Ischiofemoral Impingement in Children: Imaging With Clinical Correlation OBJECTIVE. The purpose of this study is to correlate the MRI finding of quadratus femoris edema in children with established measures of ischiofemoral impingement in adults and to determine their clinical significance. MATERIALS AND METHODS. A case-control retrospective review was performed of MR images of 12 hips of nine children (mean age, 10 years) that showed abnormal signal intensity in the quadratus femoris muscle. The findings were compared with those for 13 hips in 13 control subjects (mean age, 9.8 years). Two musculoskeletal radiologists independently measured the ischiofemoral space, quadratus femoris space, inclination angle, and hamstring tendon area. Quadratus femoris muscle edema, tears, or fatty replacement were assessed visually. Statistical analysis determined inter- and intraobserver variability and statistical differences between the two subject groups. RESULTS. Subjects with abnormal signal in the quadratus femoris muscle had a statistically significantly narrower ischiofemoral space (mean, 11.5 vs 20.7 mm; p < 0.05) and quadratus femoris space (mean, 7.2 vs 14.3 mm, p < 0.05) compared with control subjects. Inclination angles were statistically significantly larger in affected subjects (145.5 vs ; p < 0.05). No statistically significant difference was found in hamstring tendon area. Four of 12 patient hips had no symptoms or symptoms unrelated to the abnormality. Inter- and intraobserver variability scores were strong for all continuous variables. CONCLUSION. Ischiofemoral impingement is a cause of hip pain in children and should be considered with the combination of quadratus femoris muscle edema and narrowing of ischiofemoral space and quadratus femoris space. Coxa valga may contribute to narrowing of these parameters and predispose to ischiofemoral impingement. The MRI features of ischiofemoral impingement, however, are nonspecific and may be seen in patients with pain localized to that hip or in patients with symptoms unrelated to the abnormality. I schiofemoral impingement is defined by hip pain associated with narrowing of the ischiofemoral space. This space lies between the ischial tuberosity of the pelvis and the lesser trochanter of the femur. It can lead to abnormalities within the quadratus femoris muscle, including edema, tears, or fatty atrophy. The condition was first described by Johnson in 1977 [1]. He hypothesized that superior and medial migration of the lesser trochanter causes a decrease in the distance from the lesser trochanter to the ischium, thus leading to painful impingement. His report included three adult patients with hip pain after hip surgery (two after total hip arthroplasty and one after proximal femoral osteotomy). Complete pain relief was achieved after resection of the lesser tro- chanter. Since then, ischiofemoral impingement has been described almost exclusively in the adult literature, occurring most frequently in middle-aged women. Ischiofemoral impingement in the pediatric population has rarely been described. A case report by Truong et al. [2] diagnosed ischiofemoral impingement in a 14-year-old girl who presented with hip pain and clunking of the hip. MRI showed edema in the quadratus femoris muscle, as well as reduced space between the ischium and lesser trochanter. Dynamic examination under anesthesia confirmed impingement by reproducing the clunking phenomenon, as well as proximity of the lesser trochanter to the ischium. This patient was successfully treated with ischioplasty. To our knowledge, only one other case report exists describing ischiofemoral 426 AJR:206, February 2016

2 MRI of Pediatric Ischiofemoral Impingement impingement in an 11-year-old child [3], although in that case, there is a history of trauma to the hip in question. Bilateral hip involvement is recognized in 25 40% of patients, and pain is often elicited using a combination of hip extension, adduction, and external rotation [3, 4]. Despite this, no specific clinical test exists for this disorder. Imaging can therefore play an important role in its diagnosis. Radiograph findings are usually normal in adults with ischiofemoral impingement but may show proximity of the lesser trochanter to the ischium [5]. MRI is the imaging modality of choice and shows edema within the quadratus femoris muscle, as well as narrowing of the ischiofemoral and quadratus femoris spaces [6]. The purpose of this case-control study is to correlate the findings of quadratus femoris edema in a larger series of children with the established measures of ischiofemoral impingement from the adult population and to determine their clinical significance. Materials and Methods After we received approval from research ethics board of The Hospital for Sick Children, we performed a retrospective database search on all consecutive hip or pelvic MRI examinations performed between 2008 and 2014 using ISYS Search Software (version 10.2, Lexmark). Affected subjects were identified by searching body fields for the term quadratus femoris edema. Patients with a history of recent trauma (< 30 days), tumor, surgery, infection, or inflammatory arthritis were excluded from the study, as were those with inadequate MRI or clinical data. A control group was established from the same database using examinations performed during We selected consecutive hip or pelvic MRI examinations performed to rule out fracture after a fall with no fracture, bone marrow edema, or soft-tissue abnormality identified. This was achieved by searching history fields with terms including fall and fracture. Data recorded included subject age and sex and the duration, character, and side of symptoms. All MRI examinations were performed with a 1.5- or 3-T MRI scanner, with a slice thickness of 4 mm, FOV of 220 mm, matrix of , and TR/TE of 5251/80, using a phased-array coil. Contrastenhanced axial and coronal T1-weighted images were obtained in selected cases at the discretion of the radiologist involved. During all examinations, hips were positioned in internal rotation with toes pointing together and heels 5 cm apart, as per the local departmental protocol. Patient MR images were retrospectively evaluated on our PACS (Centricity WS RA1000 workstation, GE Healthcare) by two fellowship-trained musculoskeletal radiologists (radiologists A and B). Radiologist A performed two separate readings with an interval of 8 weeks. Both readers were blinded to clinical information and the other radiologist s score. Quadratus femoris muscle edema was characterized by T2-weighted hyperintense signal within the quadratus femoris muscle. The MR images from both patient and control groups were assessed for the following parameters of ischiofemoral impingement: ischiofemoral space (i.e., the smallest distance between the lateral cortex of the ischial tuberosity and medial cortex of the lesser trochanter on axial T1-weighted images), quadratus femoris space (i.e., the smallest space for passage of the quadratus femoris muscle bordered by the superolateral surface of the hamstring tendons and the posteromedial surface of the iliopsoas tendon or lesser trochanter on axial T1-weighted images), the inclination angle (i.e., the angle between the long axis of the femoral neck and the long axis of the femoral shaft on coronal T1-weighted images), and the hamstring tendon area (traced at the level where the quadratus femoris space was measured, including the semimembranosus, semitendinosis, and biceps femoris tendons). Quadratus femoris muscle edema and fatty replacement were graded visually as per the criteria set out in the article by Tosun et al. [4]. Descriptive statistical analysis was used to describe the characteristics of the participating patients. Continuous variables were described using means and SDs. Two-sample independent t tests were used to test the difference in continuous measurements between the impingement and control groups. Chi-square tests were used to test the differences in categoric measurements. Both inter- and intrarater reliability for continuous outcomes was performed by using the random-effects mixed model. Values of p < 0.05 were considered to indicate statistically significant differences. All measurements are expressed in millimeters (mean ± SD). Results Twelve hips in nine patients (seven girls and two boys) with quadratus femoris edema were identified and assigned to the patient group (mean age, 10 years; age range, 4 16 years). The control group was established from 13 subjects (10 girls and three boys; mean age, 9.8 years; age range, 3 16 years). There was no statistically significant difference in mean age or sex between the patient and control groups (p > 0.05). Bilateral involvement was found in three of nine patients (33%) with impingement. Four patients among the 12 hips with MRI signs of quadratus femoris edema were either asymptomatic or had symptoms related to the contralateral hip (Table 1). A pediatric orthopedic surgeon or rheumatologist examined all patients with hip pain. Pain was exacerbated by hip extension, adduction, and external rotation in five of the eight patients with TABLE 1: Demographic and Clinical Characteristics of Nine Patients With MRI Findings of Quadratus Femoris Edema Patient No. Age (y) Sex Quadratus Duration of Precipitating Femoris Edema Symptomatic? Pain (mo) Event 1 15 Female Right Yes 36 No Clicking and hip pain; ischiofemoral impingement confirmed by examination under anesthesia 2 12 Female Right Yes 6 No Clicking and hip pain; previous left developmental dysplasia of the hip with leg length discrepancy 3 10 Female Right No None No Right leg hemihypertrophy with leg length discrepancy 4 5 Female Left Yes 24 No Clicking and hip pain; leg length discrepancy (right shorter) 5 16 Female Bilateral Left hip only 48 No Hip pain only 6 4 Female Bilateral Right hip only 22 No Clicking and hip pain 7 10 Male Right Yes 5 No Hip pain only 8 6 Male Bilateral Left hip only 13 No Hip pain only; ongoing pain in right foot (Köhler disease) 9 16 Female Left Yes 12 No Hip pain only Clinical AJR:206, February

3 Stenhouse et al. TABLE 2: Measurements of Ischiofemoral Space, Quadratus Femoris Space, Inclination Angle, and Hamstring Tendon Area in Hips of Patients and Control Subjects Measurement Patient Hips (n = 12) hip pain and quadratus edema (62%). In three patients, pain was exacerbated by flexion, adduction, and internal rotation of the hip. In one patient, pain radiated down the ipsilateral thigh to the knee. Four patients had clicking associated with their hip pain (50%). The patient group had statistically significantly narrower ischiofemoral spaces (mean, 11.5 vs 20.7 mm) and quadratus femoris spaces (mean, 7.2 vs 14.3 mm) compared with the control group (p < 0.05). The inclination angles were statistically significantly larger in the patient group compared with the control group (mean, vs ; p < 0.05). No statistically significant difference was found in hamstring tendon area between the two groups (mean, vs ; p = 0.87) (Table 2). Quadratus femoris edema scores showed grade 1 edema in five patients and grade 2 edema in seven patients. No fatty atrophy was seen in quadratus femoris muscle. Interobserver reliability scores (kappa values ± 95% CIs) showed almost perfect agreement for ischiofemoral space (κ = ± 0.015), quadratus femoris space (κ = ± 0.017), and hamstring tendon area (intraclass correlation coefficient, > 0.8) and strong agreement for inclination angle (κ = ± 0.13). Intraobserver reliability scores showed Control Subject Hips (n = 13) p Ischiofemoral space (mm) (3.97) (6.18) < (4.17) (5.95) < (3.97) (5.90) < 0.05 Quadratus femoris space (mm) (2.45) (3.03) < (2.41) (2.73) < (2.77) (3.63) < 0.05 Inclination angle ( ) (5.37) (5.59) < (4.97) (6.43) < (5.8) (5.68) < 0.05 Hamstring tendon area ( ) (62.1) (53.17) (57.7) (54.25) (56.2) (45.40) Note Except for p values, data are mean (SD). Radiologist A recorded two separate measurements for the same subjects (measurements 1 and 3), with an intervening interval of 8 weeks. Radiologist B recorded one measurement (measurement 2). almost perfect agreement for all continuous variables (ischiofemoral space, κ = ± 0.015; quadratus femoris space, κ = ± 0.007; inclination angle, κ = ± 0.062; and hamstring tendon area, κ = ± 0.016). There was almost perfect interobserver agreement for quadratus femoris muscle edema (κ = 0.944). Discussion Hip pain with mechanical snapping is a common symptom with a wide range of differential diagnoses, including intra- and extraarticular causes. Ischiofemoral impingement should also be considered because many cases of refractory hip pain may be misdiagnosed cases of this form of impingement. Only fairly recently has ischiofemoral impingement been rediscovered as a source of hip, groin, or posterior thigh pain [5, 6]. Previously, edema within the quadratus femoris muscle had been attributed to a muscle strain. It has since been recognized that a narrowed ischiofemoral space may cause impingement of the quadratus femoris muscle as a result of abnormal contact between the lesser trochanter of the femur and ischium. Radiologically, this manifests as signal abnormality in the quadratus femoris muscle and, in adults, narrowing of the quadratus femoris space (< 8 mm) and ischiofemoral space (< 17 mm) [6]. This has been hypothesized to occur as a result of degenerative changes, trauma, or prior surgery. The condition has rarely been described in children and, to our knowledge, has been described in only two case reports [2, 3]. The diagnosis is often difficult, with no specific clinical test for this disorder. Indeed, we found that over one third of cases in our study did not have exacerbating hip pain with the typical aggravating hip movements of extension, adduction, and external rotation. For this reason, imaging can play an important role in the diagnosis of ischiofemoral impingement. To our knowledge, our study is the first to show that quadratus femoris muscle edema in children is associated with significant narrowing of both the ischiofemoral and quadratus femoris space spaces when compared with healthy subjects (Fig. 1). This parallels what has been described in the adult literature [4, 6]. A study by Tosun et al. [4] retrospectively evaluated 70 hips in 50 patients and found significantly lower values for the ischiofemoral and quadratus femoris spaces compared with those of control subjects. These findings (in the absence of an acute inciting event) would suggest that edema in the quadratus femoris muscle is not related to a muscle strain or tear but is a well-defined anatomic abnormality occurring secondary to local impingement between the ischial tuberosity and lesser trochanter. Our study concurs with this concept because all patients with hip pain had long-standing symptoms but no obvious precipitating event. The quadratus femoris muscle edema occurred within the muscle belly in all patients, rather than at the myotendinous junction, as one would expect with acute injury. Over three quarters of the children in our study with MRI features of ischiofemoral impingement were female, which is similar to findings in the adult literature [4, 6]. This has been postulated to be due to the female pelvis having a greater width and smaller depth with a greater distance between the ischial tuberosities [6]. The bilaterality of quadratus femoris muscle edema (one third of patients) would support a congenital cause, and again, this is similar to the figures reported in the adult literature [4, 6]. We postulate that coxa valga narrows the ischiofemoral space and could be an important risk factor for the development of ischiofemoral impingement in children. Inclination angles in healthy adults averages about 125, 428 AJR:206, February 2016

4 MRI of Pediatric Ischiofemoral Impingement A B Fig year-old girl with right hip pain and clicking. Ischiofemoral impingement was confirmed with examination under anesthesia. A, Axial T2-weighted fat-saturation MR image of pelvis shows quadratus femoris muscle signal change (arrow), consistent with edema. B, Axial T1-weighted MR image shows narrowing of right quadratus femoris space (line 1, 7.3 mm) and ischiofemoral space (line 2, 10.4 mm) when compared with left quadratus femoris space (line 3, 11.4 mm) and left ischiofemoral space (line 4, 17.5 mm) on contralateral side. Torriani et al. [6] defined impingement in adults as quadratus femoris space < 8 mm and ischiofemoral space < 17 mm. with this value being greater in newborns (150 ) and smaller in the elderly population (120 ) [7]. The inclination angle in the patient group was statistically significantly higher than that in control subjects (mean, vs ; p < 0.05). Few data exist on the normal inclination angle in children, although mean values for 2-year-old children have been quoted as being 137 in boys and 135 in girls [8]. Increased inclination angle in adults has been recognized as a risk factor for the development of ischiofemoral impingement, as highlighted by Tosun et al. [3, 4]. We did not find a statistically significant difference in hamstring tendon area between the two groups. The MRI features of ischiofemoral impingement can be seen in children who have pain localized to that hip (8/12 hips; 66%) as well as those with symptoms unrelated to the imaging abnormality (4/12; 33%). Similar results have been recognized in adults who have Fig. 2 6-year-old boy with left hip pain and history of Köhler disease in right foot (navicular bone avascular necrosis). T2-weighted fat-saturated MR image of pelvis shows faint bilateral quadratus femoris muscle signal change (arrows), consistent with edema. imaging signs of ischiofemoral impingement but no symptoms attributable to that hip [9]. Adductor and hamstring dysfunction were raised as possible risk factors in that adult population. In the present study, half of the children presenting with hip pain had clicking. Of the children who had symptoms on the same side as their imaging abnormality, two had a leg length discrepancy, with quadratus femoris muscle edema occurring on the same side as their longer limb (one patient had prior developmental dysplasia of the hip; the other case was idiopathic). In the group of patients with symptoms unrelated to the imaging abnormality, one patient had a leg length discrepancy secondary to limb hemihypertrophy (with quadratus femoris muscle edema on the same side). One patient with bilateral quadratus femoris muscle edema had symptomatic Köhler disease of the foot (avascular necrosis of navicular bone) on the opposite side of the hip pain (Fig. 2). We therefore postulate that patients with an underlying gait abnormality and biomechanical disturbance may be at a greater risk of developing ischiofemoral impingement. Our study has a number of limitations. We have a relatively low number of patients in each group. Despite this, we identified statistically significant differences between the patient and control groups. Given its retrospective nature, our search identified subjects with quadratus femoris muscle edema, all of which had narrowing of the ischiofemoral space and quadratus femoris space. Cases therefore with narrowing of these parameters but normal quadratus femoris muscle are therefore likely to have been missed. It remains to be seen whether these asymptomatic children identified as having MRI signs of ischiofemoral impingement will go on to develop pain at a later stage, or if indeed the quadratus femoris muscle edema disappears over time. It was not possible to confirm ischiofemoral impingement as the source of pain given the lack of surgical correlation; however, to our knowledge, no specific clinical test or surgical procedure is recognized in the diagnosis of this condition. This risk was minimized because all patients were examined by a subspecialist in orthopedics or rheumatology. The case study by Truong et al. [2] was interesting because they dynamically examined the hip under fluoroscopy and observed direct abutment of the lesser trochanter against the ischium. This was not performed in the present study. The effect of hip positioning will lead to differ- AJR:206, February

5 Stenhouse et al. ences in ischiofemoral space and quadratus femoris space distances measured on MRI. As per departmental protocol, all patients were imaged with the hips internally rotated and heels 5 cm apart. Given the retrospective nature of this study, however, this could not be confirmed at the time of the scan. Conclusion Ischiofemoral impingement is a cause of hip pain with or without clicking in children and should be considered when quadratus femoris muscle edema is found in combination with narrowing of ischiofemoral space and quadratus femoris space. Coxa valga may contribute to narrowing of these parameters and predispose to ischiofemoral impingement. The MRI findings of ischiofemoral impingement, however, are nonspecific and may be seen in patients with pain localized to that hip and in patients with symptoms unrelated to the abnormality. References 1. Johnson KA. Impingement of the lesser trochanter on the ischial ramus after total hip arthroplasty: report of three cases. J Bone Joint Surg Am 1977; 59: Truong WH, Murnaghan M, Hopyan S, Kelley SP. Ischioplasty for femoroischial impingement. JBJS Case Connect 2012; 26:2 3. Tosun Ö, Çay N, Bozkurt M, Arslan H. Ischiofemoral impingement in an 11-year-old girl. Diagn Interv Radiol 2012; 18: Tosun Ö, Algin O, Yalcin N, Cay N, Ocakoglu G, Karaoglanoglu M. Ischiofemoral impingement: evaluation with new MRI parameters and assessment of their reliability. Skeletal Radiol 2012; 41: Patti JW, Ouellette H, Bredella MA, Torriani M. Impingement of lesser trochanter on ischium as a potential cause for hip pain. Skeletal Radiol 2008; 37: Torriani M, Souto S, Thomas B, Ouellette H, Bredella M. Ischiofemoral impingement syndrome: an entity with hip pain and abnormalities of the quadratus femoris muscle. AJR 2009; 193: Chaitow L, DeLany JW. Clinical applications of neuromuscular techniques: the lower body. Edinburgh, UK: Churchill Livingstone Elsevier Science, 2002: Zippel H. Normal development of the structural elements of the hip joint in adolescence (in German). Beitr Orthop Traumatol 1971; 18: Ali AM, Teh J, Whitwell D, Ostlere S. Ischiofemoral impingement: a retrospective analysis of cases in a specialist orthopaedic centre over a four-year period. Hip Int 2013; 23: AJR:206, February 2016