Fovea Alta on MR Images: Is It a Marker of Hip Dysplasia in Young Adults?

Transcription

1 Musculoskeletal Imaging Original Research Beltran et al. MRI of Fovea Alta Musculoskeletal Imaging Original Research Luis S. Beltran 1 Jason D. Mayo 2 Zehava S. Rosenberg 1 Maria Diaz De Tuesta 3 Olga Martin 3 Luis P. Neto, Sr. 4 Jenny T. Bencardino 1 Beltran LS, Mayo JD, Bencardino JT, et al. Keywords: developmental dysplasia of the hip, diagnostic marker, fovea alta, MRI DOI: /AJR Received November 6, 2011; accepted after revision January 10, Department of Radiology, New York University Hospital for Joint Diseases, New York University Langone Medical Center, 660 First Ave, Rm 218, New York, NY Address correspondence to L. S. Beltran (Luis.Beltran@nyumc.org). 2 New Britain Radiologic Associates, New Britain, CT. 3 Hospital Universitario Ramón y Cajal, Madrid, Spain. 4 Universidade Federal de São Paulo, São Paulo, Brazil. AJR 2012; 199: X/12/ American Roentgen Ray Society Fovea Alta on MR Images: Is It a Marker of Hip Dysplasia in Young Adults? OBJECTIVE. The objective of our study was to investigate the association between high fovea capitis (fovea alta) and hip dysplasia in young adults. MATERIALS AND METHODS. In a retrospective study, blinded observers reviewed 82 pelvic radiographic and hip MRI studies of three groups of patients: those with developmental dysplasia of the hip (DDH) (center-edge angle, 20 ), those with borderline DDH (center-edge angle, ), and control patients (center-edge angle, > 25 ). The center-edge angle and coxa valga (femoral neck-shaft angle, > 135 ) were assessed on pelvic radiographs, and fovea alta was assessed on MR images (delta angle, 10 ). The Mann-Whitney and Fisher exact tests were used to correlate fovea alta with DDH and with coxa valga, respectively. Interobserver agreement for center-edge and delta angles and the diagnostic performance of fovea alta as a marker of DDH were calculated. RESULTS. Thirty-one patients with DDH, 23 with borderline DDH, and 28 without DDH were included. Excellent interobserver agreement was found for center-edge angle (concordance correlation coefficient, 0.94) and for delta angle (concordance correlation coefficient, 0.91). Fovea alta had a significant association with DDH (p < 0.001) but no association with coxa valga (p > 0.57). A significant difference (p < 0.001) was found between patients with DDH (3.4 ) and those without DDH (21.7 ) with respect to mean delta angle measurements. Fovea alta had 69.4% sensitivity, 82.1% specificity, 67.2% positive predictive value, 81.0% negative predictive value, and 75.6% overall accuracy as an indicator of DDH. CONCLUSION. Fovea alta shows promise as a strong MRI marker of DDH. D evelopmental dysplasia of the hip (DDH) can vary greatly in morphologic features, ranging from a mildly shallow acetabulum to a completely dislocated high-riding hip. Despite routine screening during infancy, DDH, particularly when mild, can remain undetected into young adulthood, predisposing patients to early osteoarthritis [1]. This condition can have serious clinical implications because unrecognized DDH with superimposed arthrosis has been found to account for a large number of hip arthroplasties. A recent study by Engesæter et al. [2] showed that 26% of 634 total hip replacements in patients younger than 44 years were performed for osteoarthritis secondary to hip dysplasia. The cause of osteoarthritis in mild hip dysplasia may be related to high asymmetric concentration of force across the hip joint due to relative lateralization of the hip center of rotation, poor acetabular coverage of the femo- ral head, and small contact area between the femoral head and dysplastic acetabulum. All of these factors can produce irreversible articular cartilage and labral damage [3 5] with a secondary detrimental effect on the longevity of the hip joint. Thus accurate and prompt diagnosis of DDH, before the premature onset of osteoarthritis, is paramount for determining the treatment regimen and for assessing the necessity of surgical intervention, such as periacetabular osteotomy and femoral varus derotation osteotomy. The imaging diagnosis of hip dysplasia in the young adult may be subtle and has traditionally been based on radiographic evaluation. Patients with mild hip dysplasia typically have subtle uncovering of the femoral head by the acetabulum rather than a highriding dislocated hip. The most commonly used measurement is the center-edge angle of Wiberg, which is used to assess acetabular coverage of the femoral head. Other com- AJR:199, October

2 Beltran et al. Fig year-old man with normal left center-edge angle. Anteroposterior pelvic radiograph shows angle measuring 35 (normal, > 25 ). monly used radiographic markers of hip dysplasia are the acetabular index angle and the vertical-center-anterior margin angle. The femoral head has a smooth surface coated with articular cartilage, except over an ovoid depression, called the fovea capitis, which is normally situated slightly inferior and posterior to the center of the head and is the attachment site of the ligamentum teres. Nötzli et al. [6] reported that the abnormal superior position of the fovea capitis, the so-called fovea alta, is a potential additional radiographic diagnostic marker of adult hip dysplasia. They also suggested that fovea alta, assessed by measurement of the delta angle, may be a predisposing factor for early acetabular and perifoveal articular cartilage injury and for osteoarthritis by decreasing the contact area of the femoral head with the superior weight-bearing articular surface of the acetabulum. Unlike radiography, which lacks sensitivity for assessing early osteoarthritis, MRI can be used to simultaneously detect morphologic disease associated with DDH, such as center-edge angle abnormalities, and to predict the presence of predisposing factors in osteoarthritis, such as cartilage and labral injury. The utility of MRI in detecting DDH, however, may be further enhanced by assessment for additional markers of hip dysplasia, such as the fovea alta. The primary aim of our study was to confirm the association of fovea alta and hip dysplasia through the use of MRI measurement of the delta angle. A secondary aim was to assess whether fovea alta is a true morphologic abnormality in hip dysplasia or reflects apparent head abnormality related to the presence of coxa valga deformity associated with hip dysplasia. Materials and Methods Our study was assessed by the internal review board at our institution and deemed exempt from full internal review board review. We performed a retrospective search of our PACS (isite PACS, Philips Healthcare) database from July 2007 through July 2010 for all patients in the age range years who had a clinical history of hip pain in the radiology report. Inclusion criteria were availability of MR images of the hip and a wellcentered anteroposterior pelvic radiograph obtained within 12 months of each other. Exclusion criteria included previous surgery and inadequate imaging studies, including absence of a well-centered anteroposterior pelvic radiograph and MR images of the hip within 12 months of each other, suboptimal image quality, or a hip that was in external rotation. Two radiologists (a musculoskeletal fellow and an attending radiologist with 1 Fig year-old man with normal right femoral neck-shaft angle. Anteroposterior pelvic radiograph cropped to show measurements shows angle measuring 128 (normal, ). Fig year-old woman with left developmental hip dysplasia. Anteroposterior radiograph shows decreased center-edge angle measuring 11. year of experience in musculoskeletal radiology) blindly and independently measured radiographic center-edge and femoral neck-shaft angles and MRI delta angles. Hip dysplasia was defined as a center-edge angle of 20 or less and borderline hip dysplasia as a center-edge angle of A nondysplastic hip was defined as having a centeredge angle greater than 25. Coxa valga was defined as a femoral neck-shaft angle greater than 135 and coxa vara as a femoral head-neck-shaft angle less than 120 [7]. Fovea alta was defined as a delta angle of 10 or less. Imaging Technique The anteroposterior radiograph of the pelvis was obtained according to standard radiographic protocol with the patient supine and both lower extremities oriented in 15 of internal rotation [8]. Technical radiographic exclusion criteria were Fig year-old woman with coxa valga. Anteroposterior pelvic radiograph cropped to show measurements shows increased femoral neck-shaft angle measuring AJR:199, October 2012

3 MRI of Fovea Alta Fig. 5 Illustration shows delta angle (δ) measured on midcoronal MR image where there is optimal visualization of fovea capitis and sourcil, determined by cross-referencing with scout lines on axial and coronal images. Angle is formed between two lines extending from medial edge of sourcil (point A) and superior edge of fovea capitis (point B) to center of femoral head. lack of obturator foramen symmetry, distance between the tip of the coccyx and symphysis pubis decreased less than 1 cm or increased more than 3 cm, and femoral external rotation. MR images were obtained according to standard MRI protocol with the patient supine and the hips in slight internal rotation, the latter achieved by taping the feet while maintaining contact of the great toes. Coronal proton density (with or without fat suppression) or T1-weighted images of the hip were selected for measurements. Imaging was performed with a 1.5-T MRI system (Magnetom Vision or Avanto, Siemens Healthcare) or a 3-T MRI system (Verio, Fig year-old man with normal delta angle. Coronal T1-weighted MR image shows angle greater than 10. Siemens Healthcare). The slice thickness ranged from 3 to 5 mm with no interslice gap. Fig year-old woman with fovea alta manifested by decreased delta angle. Coronal proton density weighted MR image shows angle of 10 (normal, > 10 ). Negative delta angle value indicates that medial edge of sourcil is medial to superior edge of fovea capitis, resulting in overlapping. Mild uncovering of femoral head also is evident. Measurements The radiographic center-edge angle was determined by constructing a best-fit circle on each femoral head and connecting a horizontal line through the centers of both circles. The angle formed by a line drawn perpendicular to the horizontal line at the midpoint of each femoral head and a second line from the center of the femoral head to the outer margin of the acetabular roof was then measured [9] (Figs. 1 and 2). The femoral neck-shaft angle was measured on anteroposterior pelvic radiographs at the intersection of the femoral neck axis and the long axis of the femur (Figs. 3 and 4). MRI measurements of the delta angle were performed on midcoronal proton density weighted (with or without fat suppression) or T1-weighted images on which there was optimal visualization of both the fovea capitis and the acetabular sourcil, as determined by cross-referencing with the scout lines on sagittal and axial images (Figs. 5 7). The angle formed by lines drawn from the medial edge of the sourcil and the superior edge of the fovea capitis to A Fig year-old man without developmental dysplasia of hip. A, Axial proton density weighted MR image shows externally rotated hip with anterior location of fovea capitis. Dotted line indicates axis of femoral neck. B, Coronal proton density weighted MR image shows decreased delta angle of 7 (normal, > 10 ) and anterior location of fovea capitis relative to medial edge of sourcil. C, Anteroposterior pelvic radiograph shows normal center-edge angle of greater than 25, indicating absence of hip dysplasia. Presence of external rotation of femur can simulate fovea alta deformity. B C AJR:199, October

4 Beltran et al. the center of the femoral head was measured. The delta angle was defined as positive when the medial edge of the sourcil was lateral to the superior edge of the fovea capitis. The delta angle was defined as negative when the medial edge of the sourcil was medial to the superior edge of the fovea capitis. Fovea alta was defined as a delta angle of 10 or less. The femoral head was in neutral or mild internal rotation position. Images on which the femur was externally rotated as determined by anterior orientation of the axis of the femoral neck and anterior location of the fovea capitis on axial MR images (which can simulate fovea alta deformity) were excluded from the study (Fig. 8). Statistical Analysis The Mann-Whitney exact test was used to correlate dysplastic (center-edge angle, 20 ) and nondysplastic (center-edge angle, > 25 ) hips with the delta angle. The Fisher exact test was used to correlate fovea alta and coxa valga. Concordance correlation coefficient was used to assess interobserver agreement for center-edge angle and delta angle measurements. A concordance correlation coefficient greater than 0.7 was interpreted as excellent agreement. The diagnostic performance of fovea alta as a marker of hip dysplasia was assessed. Results Twelve patients did not meet the inclusion criteria because of inadequate imaging. Eightytwo consecutively registered patients (23 men, 59 women; age range, years; mean age, 29 years) met the inclusion criteria. Thirty-one of the patients had dysplastic hips, 23 patients had borderline dysplastic hips, and 28 patients had nondysplastic hips, according to radiographic center-edge angle measurements. For both readers, there was a significant association between fovea alta, measured by MRI delta angle, and hip dysplasia (p < 0.001). There was also a statistically significant difference in mean delta angle measurements between dysplastic (δ = 3.4 ) and nondysplastic (δ = 21.7 ) hips (p < 0.001). In the comparison of average delta angle measurements between normal, borderline dysplastic, and dysplastic hips (Table 1), the average delta angle was significantly different between nondysplastic hips and dysplastic hips and between borderline dysplastic and dysplastic hips for both readers (Table 2). Also for both readers, no significant difference was seen between average delta angle in nondysplastic hips and borderline dysplastic hips (Table 2). As an indicator of hip dysplasia, fovea alta had 69.4% sensitivity, 82.1% specificity, 67.2% positive predictive value, TABLE 1: Mean Delta Angle MRI Measurements Reader Normal Hip (n = 28) Borderline Dysplasia (n = 23) Dysplasia (n = 31) ± ± ± ± ± ± 12.7 Note Values are mean ± SD in degrees. Normal hip is a center-edge angle greater than 25 ; borderline dysplasia, greater than 20 and up to 25 ; dysplasia, 20 or less. TABLE 2: Results of Exact Mann-Whitney Test of MRI Delta Angle Measurements Reader Normal Hip Versus vs Borderline Dysplasia 81.0% negative predictive value, and 75.6% overall accuracy. There was no significant association between coxa valga and fovea alta evaluations for either reader (p > 0.57). Excellent agreement was seen between readers for radiographic center-edge angle (concordance correlation coefficient, 0.94) and MRI delta angle (concordance correlation coefficient, 0.91). Patient Group Normal Hip Versus Dysplasia Borderline Dysplasia Versus Dysplasia < < Note Values are p. Discussion The concept that young adults with DDH have an abnormally high position of the fovea capitis and that this finding can be used as an additional radiographic marker of hip dysplasia was first introduced by Nötzli et al. in 2001 [6]. To the best of our knowledge, there have been no further reports on this topic since then. Nevertheless, hip dysplasia continues to be a serious health issue. In a recent study, Engesæter et al. [2] found that osteoarthritis secondary to hip dysplasia in patients younger than 44 years accounted for 26% of 634 hip replacements in their database. MRI has been found reliable for assessing morphologic abnormalities of DDH and for measuring acetabular coverage of the femoral head with the center-edge angle [10]. MRI has also been found to be an accurate technique for assessing associated injuries of the articular cartilage and labrum [11 13]. Furthermore, advanced biochemical MRI techniques, such as delayed gadolinium-enhanced MRI of cartilage (dgemric), have shown great potential for earlier detection of articular cartilage changes in osteoarthritis secondary to hip dysplasia [14]. The addition of fovea alta as a diagnostic MRI marker of hip dysplasia can further enhance the utility of MRI for detecting hip dysplasia, making the modality a comprehensive examination of both morphologic abnormalities and associated cartilage and labral injuries in the hip joint. Extrapolating the radiographic delta angle measurements of fovea alta developed by Nötzli et al. [6] to MRI measurements, we were able to confirm the association between fovea alta and hip dysplasia. We found a significant association between fovea alta and hip dysplasia for both readers (p < 0.001) and a statistically significant difference in mean delta angle measurements between dysplastic (d = 3.4 ) and nondysplastic (δ 21.7 ) hips (p < 0.001). We also found a statistically significant difference in mean delta angle measurements when comparing dysplastic and borderline dysplastic hips. However, no significant difference was seen between average delta angle in nondysplastic and borderline dysplastic hips (Table 2). These findings suggest that the delta angle may be a more reliable indicator of hip dysplasia in patients with a centeredge angle of 20 or less. Thus caution should be exercised in the use of delta angle to diagnose hip dysplasia in patients with a mildly decreased center-edge angle and borderline hip dysplasia (center-edge angle, ). Although, to our knowledge, there is no evidence in the literature to prove or disprove potential causes of the presence of fovea alta in hip dysplasia, we hypothesize that the abnormal contact pressure along the superior aspect of the femoral head and the lateralized center of hip rotation that occur in hip dysplasia likely contribute to the altered development of the shape of the femoral head during skeletal maturation, including the eventual position of the fovea capitis along a more superior aspect of 882 AJR:199, October 2012

5 MRI of Fovea Alta the femoral head. Nötzli et al. [6] also suggested that the position of the fovea capitis might be a predisposing factor in early acetabular and perifoveal articular cartilage injury leading to osteoarthritis by decreasing the contact area of the chondral tissue lined femoral head with the superior, weight-bearing chondral surface of the acetabulum. Our results show no significant association between coxa valga and fovea alta (p > 0.57). This evaluation was performed to exclude the possibility that the apparently high position of the fovea capitis along the femoral head in hip dysplasia is simply a manifestation of coxa valga (increased femoral neckshaft angle), which is frequently associated with hip dysplasia. In the assessment of interobserver variability, we found excellent agreement between readers for radiographic center-edge angle (concordance correlation coefficient, 0.94) and MRI delta angle (concordance correlation coefficient, 0.91). This finding is likely a reflection of the presence of discrete anatomic points in the measurements used to calculate the center-edge and delta angles, making them reliable diagnostic tools in assessment for hip dysplasia. Limitations of our study include the retrospective design and small patient sample. Furthermore, rather than using clinical or surgical confirmation of disease, we subdivided our patients into those with normal hips, those with borderline dysplastic hips, and those with definite hip dysplasia on the basis of radiographic center-edge angle measurements. Nevertheless, radiographic center-edge angle measurements are widely acceptable diagnostic parameters of hip dysplasia. In addition, the excellent interobserver agreement on our center-edge angle measurements suggests that our margin of error was small. Similarly, there is a potential for error in MRI measurement of the delta angle related to the position of the femur at the hip joint. Ideally, the femoral head should be in neutral position or in slight external or internal rotation to allow accurate measurement of the delta angle. Extreme internal or external rotation of the hip can lead to overestimation of the medial edge of the sourcil, which can result in a false-positive diagnosis of fovea alta (Fig. 8). Conclusion Using MRI measurements, we found a significant association between fovea alta and hip dysplasia in our patient sample. Future studies with a larger patient group may help confirm the reliability of fovea alta as a marker of hip dysplasia on MR images. Awareness of the imaging appearance of fovea alta by radiologists and orthopedic surgeons coupled with recognition of other established diagnostic imaging markers may lead to improved diagnosis of mild hip dysplasia, which may facilitate treatment and prevent or delay the onset of osteoarthritis in this patient population. Future directions may include the use of advanced MRI (such as dgemric) of the articular cartilage to assess for characteristic patterns of biochemical cartilage alterations in patients with fovea alta and hip dysplasia, particularly involving the chondral surfaces in the perifoveal femoral head and the superomedial aspect of the acetabular roof. References 1. Mahan ST, Katz JN, Kim YJ. To screen or not to screen? A decision analysis of the utility of screening for developmental dysplasia of the hip. J Bone Joint Surg Am 2009; 91: Engesæter IØ, Lehmann T, Laborie LB, Lie SA, Rosendahl K, Engesæter LB. Total hip replacement in young adults with hip dysplasia: age at diagnosis, previous treatment, quality of life, and validation of diagnoses reported to the Norwegian Arthroplasty Register between 1987 and Acta Orthop 2011; 82: Armand M, Lepistö J, Tallroth K, Elias J, Chao E. Outcome of periacetabular osteotomy: joint contact pressure calculation using standing AP radiographs 12 patients followed for average 2 years. Acta Orthop 2005; 76: Hipp JA, Sugano N, Millis MB, Murphy SB. Planning acetabular redirection osteotomies based on joint contact pressures. Clin Orthop Relat Res 1999; Jul: Michaeli DA, Murphy SB, Hipp JA. Comparison of predicted and measured contact pressures in normal and dysplastic hips. Med Eng Phys 1997; 19: Nötzli HP, Müller SM, Ganz R. The relationship between fovea capitis femoris and weight bearing area in the normal and dysplastic hip in adults: a radiologic study. Z Orthop Ihre Grenzgeb 2001; 139: Reikerås O, Høiseth A, Reigstad A, Fönstelien E. Femoral neck angles: a specimen study with special regard to bilateral differences. Acta Orthop Scand 1982; 53: Clohisy JC, Carlisle JC, Beaulé PE, et al. A systematic approach to the plain radiographic evaluation of the young adult hip. J Bone Joint Surg Am 2008; 90(suppl 4): Wiberg G. Studies on dysplastic acetabula and congenital subluxation of the hip joint, with special reference to the complication of osteoarthritis. Acta Chir Scand 1939; 83(suppl 58): Chen L, Boonthathip M, Cardoso F, Clopton P, Resnick D. Acetabulum protrusio and center edge angle: new MR-imaging measurement criteria a correlative study with measurement derived from conventional radiography. Skeletal Radiol 2009; 38: James S, Miocevic M, Malara F, Pike J, Young D, Connell D. MR imaging findings of acetabular dysplasia in adults. Skeletal Radiol 2006; 35: Nishii T, Sugano N, Tanaka H, Nakanishi K, Ohzono K, Yoshikawa H. Articular cartilage abnormalities in dysplastic hips without joint space narrowing. Clin Orthop Relat Res 2001; Feb: Nishii T, Tanaka H, Nakanishi K, Sugano N, Miki H, Yoshikawa H. Fat-suppressed 3D spoiled gradient-echo MRI and MDCT arthrography of articular cartilage in patients with hip dysplasia. AJR 2005; 185: Kim YJ, Jaramillo D, Millis MB, Gray ML, Burstein D. Assessment of early osteoarthritis in hip dysplasia with delayed gadolinium-enhanced magnetic resonance imaging of cartilage. J Bone Joint Surg Am 2003; 85: AJR:199, October