MRI Findings in Painful Metal-on- Metal Hip Arthroplasty

Transcription

1 Musculoskeletal Imaging Original Research Hayter et al. MRI of Painful Metal-on-Metal Hip Prostheses Musculoskeletal Imaging Original Research Catherine L. Hayter 1 Stephanie L. Gold 1 Matthew F. Koff 1 Giorgio Perino 2 Danyal H. Nawabi 3 Theodore T. Miller 1,4 Hollis G. Potter 1,4 Hayter CL, Gold SL, Koff MF, et al. Keywords: adverse local tissue reaction, aseptic lymphocytic vasculitis associated lesions, hip resurfacing arthroplasty, metal-on-metal arthroplasty, MRI, multiacquisition variable-resonance image combination (MAVRIC), total hip arthroplasty DOI: /AJR Received November 3, 2011; accepted after revision January 19, The Hospital for Special Surgery receives research support from GE Healthcare. 1 Department of Radiology and Imaging, Hospital for Special Surgery, 535 E 70th St, New York, NY Address correspondence to H. G. Potter (potterh@hss.edu). 2 Department of Pathology and Laboratory Medicine, Hospital for Special Surgery, New York, NY. 3 Adult Reconstruction and Joint Replacement Division, Hospital for Special Surgery, New York, NY. 4 Weill Cornell Medical College of Cornell University, New York, NY. CME This article is available for CME credit. AJR 2012; 199: X/12/ American Roentgen Ray Society MRI Findings in Painful Metal-on- Metal Hip Arthroplasty OBJECTIVE. The objective of our study was to compare the frequency of osseous and soft-tissue abnormalities in patients presenting with hip pain after resurfacing arthroplasty and after total hip arthroplasty (THA), correlate the MRI findings with histologic results, and determine which MRI findings are predictive of aseptic lymphocytic vasculitis associated lesions. MATERIALS AND METHODS. The MRI examinations of patients with metal-on-metal hip prostheses placed at resurfacing arthroplasty (n = 31) or THA (n = 29) were reviewed for osteolysis, synovitis, extracapsular disease, synovial pattern, and mode of decompression into adjacent bursae. Regional muscles and tendons were assessed for tendinosis, tear, atrophy, and edema. Histologic and operative findings were reviewed in 19 patients (20 hips) who underwent revision surgery. Chi-square tests were performed to detect differences between the resurfacing arthroplasty and THA groups. The Wilcoxon rank sum test was performed to detect differences in MRI findings in patients with and those without aseptic lymphocytic vasculitis associated lesions. RESULTS. Synovitis was detected in 77.4% of resurfacing arthroplasty hips and 86.2% of THA hips. Extracapsular disease was present in 6.5% of resurfacing arthroplasty hips and 10.3% of THA hips. Osteolysis was detected in 9.7% of resurfacing arthroplasty hips and 24.1% of THA hips. There was no difference in the incidence of synovitis (p = 0.51), osteolysis (p = 0.17), or extracapsular disease (p = 0.67) between the resurfacing arthroplasty and THA groups. Patients with aseptic lymphocytic vasculitis associated lesions had higher volumes of synovitis (p = 0.04) than patients without aseptic lymphocytic vasculitis associated lesions. Extracapsular disease and muscle edema were seen only in patients with aseptic lymphocytic vasculitis associated lesions. CONCLUSION. Synovitis is common in patients with metal-on-metal hip prostheses and occurs with a similar incidence after resurfacing arthroplasty and after THA; osteolysis and extracapsular disease are uncommon. The MRI signs most suggestive of aseptic lymphocytic vasculitis associated lesions are high volumes of synovitis, extracapsular disease, and intramuscular edema. I n recent years there has been a resurgence in the use of metal-onmetal hip prostheses, which offer the potential of low implant wear and prolonged implant survival compared with conventional metal-on-polyethylene constructs [1, 2]. Despite overall good outcomes with metal-on-metal hip prostheses [1], some studies have described periprosthetic fluid collections and soft-tissue masses complicating metal-on-metal prostheses [3 6]. Many of these lesions correspond to aseptic lymphocytic vasculitis associated lesions on histology, which are characterized by perivascular or diffuse infiltrates of lymphocytes, often with extensive tissue necrosis [7]. Conventional radiographs of patients with aseptic lymphocytic vasculitis associated lesions usually show normal findings despite the findings of extensive soft-tissue necrosis at revision surgery [4, 8]. Ultrasound may be used to detect periprosthetic fluid collections but is limited in its ability to detect deep fluid collections and osseous abnormalities. MRI is, therefore, increasingly recognized as a valuable tool in the assessment of patients with hip pain after placement of a metal-on-metal prosthesis. MRI is the most accurate method with which to detect and quantify osteolysis and wear-induced synovitis [9 11]. MRI can also assess for periprosthetic soft-tissue collections or masses that may indicate aseptic lymphocytic vasculitis 884 AJR:199, October 2012

2 MRI of Painful Metal-on-Metal Hip Prostheses associated lesions [4, 6]. Although conventional MRI is limited because of susceptibility artifacts generated by the cobalt-chromium components of the metal-on-metal prosthesis, prototype sequences have shown a reduction in artifacts and improved depiction of the surrounding soft tissues [12, 13]. The aim of this retrospective observational study was to compare the incidence of osseous and soft-tissue abnormalities in patients presenting with unexplained pain after hip resurfacing arthroplasty and after total hip arthroplasty (THA). We also aimed to correlate the MRI appearance with operative and histologic findings in a subset of patients who underwent revision surgery to determine which MRI findings are most predictive of aseptic lymphocytic vasculitis associated lesions. A Fig year-old man who presented for follow-up imaging 15 months after total hip arthroplasty. A and B, Coronal fast spin-echo (FSE) (A) and multiacquisition variable resonance image combination (MAVRIC) (B) images. Synovitis (black arrows, B) is seen on MAVRIC image but is not visualized on corresponding FSE image. Also note improved depiction of prosthesis-bone interface (white arrows, B) on MAVRIC image. Materials and Methods Patient Cohort All methods were approved by the local institutional review board and the informed consent of subjects was obtained before enrollment in the study. Patients with a metal-on-metal prosthesis placed at resurfacing arthroplasty or THA who had been referred for MRI to investigate unexplained hip pain were included in this study. Exclusion criteria were patients presenting with mechanical symptoms including loosening and impingement, asymptomatic patients referred for investigation of elevated metal ion levels or a recalled implant, and patients referred from an outside institution whose clinical and radiographic data were not available. Data were collected about the demographic characteristics of patients including sex, age, and body mass index (BMI); type of implant used; and length of time since arthroplasty. MR Image Acquisition All subjects underwent MRI using standard clinical protocols optimized to minimize metallic susceptibility artifact [14]. Reduction of susceptibility artifact was primarily achieved through the use of a wide receiver bandwidth, which increases the strength of the readout gradient. A wide receiver bandwidth reduces interecho spacing, allowing longer echo-train lengths and a reduction in scanning time. A high readout matrix was applied to improve spatial resolution and to better define the interface between the implant and the surrounding structures. The number of signals acquired was increased to increase the signal-to-noise ratio [15]. Scanning was performed using 1.5-T scanners (450 or HDx, GE Healthcare) and a three-element shoulder coil or an eight-channel cardiac coil. Two-dimensional intermediate-te (proton density weighted) fast spin-echo (FSE) images were obtained in three orthogonal planes. The parameters were as follows: TR range, ms; TE range, ms; bandwidth, ± khz; FOV, cm; number of signals acquired, 4 6; acquisition matrix, 512 (frequency) (phase); and slice thickness, 3 4 mm [14]. A fast inversion recovery sequence was performed in the B coronal plane using the following parameters: TR/ TE, 4500/18; inversion time, 150 ms; bandwidth, ± khz; FOV, cm; number of signals acquired, 2; acquisition matrix, 256 (frequency) 192 (phase); and slice thickness, 5 mm. The average acquisition time was 6 minutes (range, 5 8 minutes) for each imaging plane. An additional prototype multiacquisition variable-resonance image combination (MAVRIC) scan designed to further reduce susceptibility artifact [12, 13] was acquired. The MAVRIC technique minimizes image distortions by combining multiple individual datasets that are acquired at frequency bands incrementally offset from the dominant proton frequency [13]. The MAVRIC scan was acquired in the coronal plane using the following parameters: TR range, ms; TE range, ms; bandwidth, ± 125 khz; FOV, cm; number of signals acquired, 0.5; acquisition matrix, (frequency) (phase); and slice thickness, mm (Fig. 1). The average acquisition time was 8 minutes (range, 6 9 minutes) for a single imaging plane. MR Image Analysis The MR images were evaluated by two musculoskeletal radiologists via consensus agreement. At the time of the study, one observer, a senior musculoskeletal MRI attending radiologist, had more than 10 years experience in the MRI assessment of joints after arthroplasty; the other observer was a musculoskeletal radiologist with 15 months experience in MRI of joints after arthroplasty. Fig. 2 Coronal fast spin-echo image of 63-year-old woman who presented for follow-up imaging 30 months after resurfacing arthroplasty. MR image shows extensive proximal femoral osteolysis (arrows), which is seen as intermediate-signal-intensity material replacing normal high-signal-intensity fatty marrow. AJR:199, October

3 Hayter et al. A Fig. 3 Axial fast spin-echo images show different patterns of synovitis. A, 39-year-old man who presented for follow-up imaging 23 months after resurfacing arthroplasty. MR image shows fluid signal intensity with thin pseudocapsule (arrows). B, 51-year-old woman who presented for follow-up imaging 27 months after resurfacing arthroplasty. MR image shows intermediate- to low-signal-intensity solid-appearing debris (arrows). C, 50-year-old woman who presented for follow-up imaging 34 months after resurfacing arthroplasty. MR image shows mixed fluid signal intensity (black arrows) and solid-appearing debris (white arrow). D, 57-year-old man who presented for follow-up imaging 43 months after total hip arthroplasty. MR image shows case characterized as Other ; in this case, fluid signal intensity with thickened intermediate-signalintensity pseudocapsule (arrows) is seen. B Whether osteolysis was present and, if present, the location (acetabulum, femur, or both) were recorded. Osteolysis was denoted as intermediatesignal-intensity marrow replacement that appeared in contrast to the high signal intensity of the intramedullary fatty marrow on intermediate-te sequences (Fig. 2). The presence of synovitis, which was defined as fluid signal intensity or intermediate- to lowsignal- intensity debris within or directly communicating with the pseudocapsule of the hip, was recorded. The finding no synovitis was recorded when the pseudocapsule of the hip was observed to be closely applied to the femoral neck or to the stem of the femoral component of the prosthesis. The pattern of synovitis was classified as fluid signal intensity with a thin (< 5 mm) intermediateto low-signal-intensity pseudocapsule; intermediate- to low-signal-intensity debris; a mixed pattern, with fluid signal intensity and intermediate- to lowsignal-intensity debris; or Other, not corresponding to one of the patterns described (Fig. 3). The presence of decompression of synovitis into adjacent bursae (iliopsoas bursa, trochanteric bursa, or both bursae) was recorded (Fig. 4). Visualization of a communication with the pseudocapsule of the hip joint was required to denote the presence of decompression into these bursae. The presence of extracapsular disease was recorded. Extracapsular disease was defined as fluid signal intensity or intermediate- to low-signalintensity lesion adjacent to the prosthesis that did not communicate with the pseudocapsule of the hip (Fig. 5). Extracapsular disease was classified as fluid signal intensity or intermediate- to low-signal-intensity on intermediate-te FSE and inversion recovery sequences. The volumes of osteolysis, synovitis, and extracapsular disease were calculated from either the axial FSE images or the coronal MAVRIC images using a previously validated manual segmentation method [9, 10]. Manual segmentation was performed on a dedicated PACS workstation. The area of synovitis or osteolysis was calculated on each slice; the sum of the areas across multiple slices was then multiplied by the slice thickness to obtain a volume measurement (Fig. 6). All quantitative measurements were performed by one musculoskeletal radiologist. In half of the subjects, measurements were performed by a second musculoskeletal radiologist to allow assessment of interobserver reliability. The femoral, sciatic, and obturator nerves were assessed for the presence of neurovascular impingement. Neurovascular impingement was defined as loss of the normal fat planes around the nerve with or without displacement of the nerve fascicles. Additional findings on MRI were recorded including fracture; stress reaction; marrow infiltration; or suspected component loosening, which was defined as circumferential bone resorption around the prosthesis. The gluteus medius, gluteus minimus, and iliopsoas were assessed for the presence of tendinosis, which were graded as normal, defined as showing uniformly low signal intensity without tendon thickening; mild tendinosis, increased signal intensity but no tendon thickening; moderate tendinosis, increased signal intensity with focal tendon thickening; or severe tendinosis, increased signal intensity with diffuse tendon thickening. The presence or absence of a tear was graded as follows: 1, no tear; 2, partial-thickness tear; or 3, full-thickness tear. The presence of muscle atrophy, defined as loss of volume and the presence of fatty replacement, and the presence of muscle edema, defined as the presence of high signal intensity on inversion recovery images, was recorded. The short external rotator muscles were assessed for atrophy and dehiscence of the tendons from the posterior pseudocapsule [16]. Radiographic Analysis Standard anteroposterior pelvic and lateral hip radiographs were evaluated by an experienced musculoskeletal radiologist who was blinded to the MRI findings. The presence of osteolysis, femoral neck resorption (for the resurfacing arthroplasty cases), periprosthetic lucency larger than 2 mm, fracture, component migration, a soft-tissue mass, and dense joint effusion was recorded. Patient Follow-Up and Histologic Analysis Subsequent patient follow-up was recorded from the clinical notes. The operative notes of the patients who underwent revision surgery were reviewed by one surgeon and the final clinical diagnosis for each patient was determined using the operative notes and histologic findings at revision surgery. The presence of visible gray metallic staining of the soft tissues was noted; when staining was present in combination with the histologic C D 886 AJR:199, October 2012

4 MRI of Painful Metal-on-Metal Hip Prostheses findings of metallic debris, a diagnosis of metallosis was given. The presence of macroscopic soft-tissue necrosis at surgery was recorded. The histologic results of the tissue obtained from the patients who underwent revision surgery were analyzed by one musculoskeletal pathologist. All tissue excised at surgery was submitted for histologic examination, serially cut, and extensively sampled to obtain maximum information. Tissue was routinely processed, cut, and stained Fig year-old woman who presented for follow-up imaging 34 months after resurfacing arthroplasty. Coronal fast spin-echo image shows synovitis communicating with pseudocapsule of hip (black arrows). Low-signal-intensity extracapsular disease (white arrows), which does not communicate with joint, is also present within abductor musculature. A Fig. 4 Axial fast spin-echo images show patterns of synovial decompression. A, 67-year-old man who presented for follow-up imaging 11 months after resurfacing arthroplasty. MR image shows synovitis to decompress posterolaterally into trochanteric bursa (arrows). B, 51-year-old woman who presented for follow-up imaging 23 months after resurfacing arthroplasty. MR image shows synovitis to decompress anteriorly into iliopsoas bursa (arrows). with H and E. Histologic sections were examined at light microscopy without knowledge of the MRI classification. Sections were evaluated for the presence of fibrinous exudates, necrosis, inflammatory cells (histiocytes, lymphocytes, plasma cells, eosinophils, and neutrophils), metallic debris, and aggregates of corrosion products. Histologic sections were scored using the system proposed by Campbell et al. [17] for classifying aseptic lymphocytic vasculitis associated lesions. A score B of 0 3 was given for the status of the synovial lining and for the degree of tissue organization; a score of 0 4 was given for the presence of an inflammatory infiltrate. The scores were summed to yield the final aseptic lymphocytic vasculitis associated lesion score (0 10). A diagnosis of aseptic lymphocytic vasculitis associated lesions was given if the final score was 5 or greater (i.e., moderate or severe disease). Statistical Analysis A Wilcoxon rank sum test was performed to detect differences between the resurfacing arthroplasty and THA groups in age, BMI, length of time since implantation, synovial volume, osteolysis volume, and extracapsular disease volume. Posthoc comparisons were performed when statistical significance was found. A Spearman rank correlation analysis was performed to calculate the correlation coefficient (r) between the volume of synovitis and demographic data for each group. Chi-square tests were performed to detect differences between the groups in the proportions of individuals with osteolysis, synovitis, and extracapsular disease. The Wilcoxon rank sum test was performed for the subjects who underwent revision surgery to detect differences in demographic data, synovial volume, tendinosis, and tendon tears between patients with and those without aseptic lymphocytic vasculitis associated lesions. Chi-square tests of association (2 2 contingency tables) were performed to evaluate proportions of aseptic lymphocytic vasculitis associated lesions by sex. Fig. 6 Methods of segmentation. Outlined areas show regions of interest. A, 54-year-old woman who presented for follow-up imaging 16 months after resurfacing arthroplasty. Coronal fast spin-echo (FSE) image shows method of segmentation for measuring osteolysis volume. B, 42-year-old man who presented for follow-up imaging 31 months after resurfacing arthroplasty. Axial FSE image shows method of segmentation for measuring volume of synovitis. AJR:199, October

5 Hayter et al. Statistical significance for all analyses was taken at p < Interclass correlation coefficients and coefficients of repeatability [18] were calculated between the two examiners for synovitis and osteolysis volume measurements. Results Comparison of Resurfacing Arthroplasty Group and Total Hip Arthroplasty Group The demographic, radiographic, and MRI findings in the resurfacing arthroplasty group and the THA group are summarized in Table 1. Patient Demographics and Radiographic Assessment The resurfacing arthroplasty group comprised 31 metal-on-metal hip resurfacings in 30 patients (11 men, 19 women). The Birmingham Hip Resurfacing system (Smith and Nephew) accounted for 29 of the implants and the Conserve Plus system (Wright Medical) accounted for two of the implants. The THA group comprised 29 metal-onmetal hip implants in 27 patients (14 men, 13 women). The acetabular component consisted of an Articular Surface Replacement (DePuy) in 15 hips, Birmingham Resurfacing system (Smith and Nephew) in eight hips, ReCap Acetabular Cup (Biomet Orthopedics) in two hips, Pinnacle Acetabular Cup (DePuy) in two hips, and Conserve Plus (Wright Medical) in one hip. The original implant type could not be determined in one case. There was no significant difference in age (p = 0.08) or BMI (p = 0.69) between the two groups. The length of time since implant placement was significantly (p < ) increased in the THA group compared with the resurfacing arthroplasty group. Evaluation of concurrent radiographs revealed an abnormality in four resurfacing arthroplasty hips and five THA hips (Table 1). All radiographic abnormalities were confirmed on MRI. Osteolysis Osteolysis was detected in three of the resurfacing arthroplasty hips (9.7%), with a mean volume of 10.2 ± 6.9 cm 3 (SD; range, cm 3 ). Osteolysis was detected in seven of the THA hips (24.1%), with a mean volume of 2.6 ± 1.3 cm 3 (range, cm 3 ). There was no difference in the proportion of patients with osteolysis (p = 0.17) between the resurfacing arthroplasty and THA groups. A comparison of volume could not be performed because of the low numbers of patients in each group with osteolysis. The interclass correlation coefficient between the two observers for osteolysis volume was 0.98 (p < ). The coefficient of repeatability between the two observers was 0.25 cm 3. Synovitis Synovitis was detected in 24 of the resurfacing arthroplasty hips (77.4%), with a mean volume of 31.3 ± 48.3 cm 3 (range, cm 3 ). In five of the 24 cases, synovitis was detected only on the MAVRIC images. The pattern of synovitis is outlined in Table 1. Synovitis was seen to decompress into adjacent bursae in 14 hips (58.3%): trochanteric bursa, 10 hips; iliopsoas bursa, three hips; and trochanteric and iliopsoas bursae, one hip. The volume of synovitis did not correlate to sex (p = 0.08), age (r = 0.15, p = 0. 41), BMI (r = 0.03, p = 0.87), or length of time since implant placement (r = 0.32, p = 0.08) in the resurfacing arthroplasty group. Synovitis was detected in 25 of the THA hips (86.2%), with a mean volume of 62.1 ± cm 3 (range, cm 3 ). In five of the 25 cases, synovitis was detected only on the MAVRIC images. The pattern of synovitis is outlined in Table 1. Synovitis was seen to decompress into adjacent bursae in 13 hips (52.0%): trochanteric bursa, six hips; iliopsoas bursa, three hips; and trochanteric and iliopsoas bursae, four hips. The volume of synovitis did not correlate to sex (p = 0.39), age (r = 0.32, p = 0.10), or BMI (r = 0.31, p = 0.12) in the THA group. The volume of synovitis significantly correlated to the length of time since implant placement in the THA group (r = 0.56, p = ). There was no difference in the proportion of patients with synovitis (p = 0.51) between the resurfacing arthroplasty and THA groups. The THA group had a higher mean volume of synovitis (62.1 ± mm 3 ) than the resurfacing arthroplasty group (31.3 ± 48.3 mm 3 ); however, this difference did not reach statistical significance (p = 0.18). The interclass correlation coefficient between the two observers for synovitis volume was 0.99 (p < ). The coefficient of repeatability between the two observers was 1.8 cm 3. Extracapsular Disease Extracapsular disease was detected in two of the resurfacing arthroplasty hips (6.5%), with a mean volume of 36.6 ± 38.5 cm 3 (range, cm 3 ). In each case, extracapsular disease consisted of low-signal-intensity lesions dissecting within the hip abductor musculature. Extracapsular disease was present in three of the THA hips (10.3%), with a mean volume of 18.3 ± 20.4 cm 3 (range, cm 3 ). In all cases, extracapsular disease consisted of low-signal-intensity lesions dissecting into the hip abductor muscles (two cases) or obturator internus muscle (one case). There was no difference in the proportion of patients with extracapsular disease (p = 0.67) between the resurfacing arthroplasty and THA groups. A comparison of volume could not be performed because of the low numbers of patients in each group with extracapsular disease. Other Features Neurovascular compression secondary to synovitis was detected in four of the resurfacing arthroplasty hips (12.9%) and two of the THA hips (6.9%) (Table 1). In the resurfacing arthroplasty group there was one case of metastatic disease, one avulsion fracture of the greater tuberosity, and one pathologic fracture of the pubic ramus associated with osteolysis. In the THA group there were two cases of fracture, two cases of femoral stress reaction, and one case of suspected femoral component loosening. In the resurfacing arthroplasty group, tendinosis was detected in the gluteus medius tendon in 21 hips (67.7%), gluteus minimus in 16 hips (51.6%), and iliopsoas in 13 hips (41.9%). Muscle atrophy was observed in nine cases (gluteus medius, n = 2; gluteus minimus, n = 7). Edema was observed in both the gluteus medius and minimus in two cases. In the THA group, tendinosis was detected in the gluteus medius tendon in 28 hips (96.6%), gluteus minimus in 22 hips (75.9%), and iliopsoas in 15 hips (51.7%). Muscle atrophy was observed in four cases: gluteus medius, one case; gluteus medius and minimus, one case; and iliopsoas, two cases. Edema was observed in four cases (gluteus medius, n = 1; gluteus medius and minimus, n = 1; iliopsoas, n = 2). Short external rotator muscle atrophy was observed in 28 of the resurfacing arthroplasty hips (90.3%) and 23 of the THA hips (79.3%). There was dehiscence of the short external rotator tendons from the posterior pseudocapsule in 11 of the resurfacing arthroplasty hips (35.5%) and 11 of the THA hips (37.9%). Comparison of MRI, Operative, and Histologic Findings To date, eight patients in the resurfacing arthroplasty group and 11 patients (12 hips) in the THA group have undergone THA conversion to a metal-on-polyethylene prosthesis. Four patients in the resurfacing arthroplasty 888 AJR:199, October 2012

6 MRI of Painful Metal-on-Metal Hip Prostheses TABLE 1: Demographic, Radiographic, and MRI Findings in Resurfacing Arthroplasty Versus Total Hip Arthroplasty (THA) Groups Feature Resurfacing Arthroplasty Group (n = 31 Hips) THA Group (n = 29 Hips) Difference Between Groups (p) Demographic data Age, y (mean ± SD) 51.2 ± ± BMI (mean ± SD) 26.8 ± ± Time since implant placement, mo (mean ± SD) 22.5 ± ± 15.8 < Radiographic abnormality Osteolysis 2 2 Neck resorption 3 Fracture 1 0 Soft-tissue density 0 5 Synovitis No. (%) of patients 24 (77.4) 25 (86.2) 0.51 Volume, mm 3 (mean ± SD) a 31.3 ± ± Pattern Fluid 6 2 Debris 8 7 Mixed 8 10 Other 2 6 Decompression, no. (%) of hips 14 (58.3) 13 (52.0) 0.71 Osteolysis No. (%) of patients 3 (9.7) 7 (24.1) 0.17 Volume, mm 3 (mean ± SD) a 10.2 ± ± 1.3 b Location Femur 0 5 Acetabulum 2 1 Both 1 1 Extracapsular disease No. (%) of hips 2 (6.5) 3 (10.3) 0.67 Volume, mm 3 (mean ± SD) a 36.6 ± ± 20.4 b Pattern Fluid signal 0 0 Intermediate to low signal 2 3 Neurovascular compression No. (%) of hips 4 (12.9) 2 (6.9) b Location Femoral 2 0 Sciatic 1 1 Obturator 1 1 Note BMI = body mass index. a Mean values calculated from patients with synovitis, osteolysis, or extracapsular disease. b The p value was not calculated because of the low number of patients. group and three patients in the THA group are awaiting surgery. Table 2 outlines the MRI, operative, and histologic findings and the final diagnosis in this subset of 20 patients (Fig. 7). In one case, there was insufficient tissue submitted to pathology to allow a histologic diagnosis. The 19 cases with histologic results were stratified into two groups on the basis of the final aseptic lymphocytic vasculitis associated lesion score. Using the grading system proposed by Campbell et al. [17], we diagnosed aseptic lymphocytic vasculitis associated lesions in cases with a final score of 5 or greater. A diagnosis of aseptic lymphocytic vasculitis AJR:199, October

7 Hayter et al. TABLE 2: Demographic, Radiographic, MRI, Operative, and Histologic Findings in 19 Patients Who Underwent Revision Surgery Hip No. Implant Type Sex Volume on MRI (cm 3 ) Synovitis Osteolysis Soft-Tissue Necrosis Operating Room Gray Metallic Staining Other Findings ALVAL Score a Final Diagnosis 1 RSA F ( ) ALVAL 2 RSA F Y 7 ( ) ALVAL 3 RSA F Y Osteolysis, 10 ( ) ALVAL pathologic fracture 4 RSA F Y Y 10 ( ) ALVAL and metallic debris 5 RSA M ( ) Metal hypersensitivity 6 RSA M Y 4 ( ) Metallosis 7 RSA F ( ) Metal hypersensitivity 8 RSA F Y 6 ( ) ALVAL and metallic debris 9 THA M Y Y 6 ( ) ALVAL and metallic debris 10 THA M ( ) Wear-induced disease 11 THA F Y Y 5 ( ) ALVAL and metallic debris 12 THA F ( ) Metal hypersensitivity 13 THA F Y Y Osteolysis 7 ( ) ALVAL and metallic debris 14 THA M Y 9 ( ) ALVAL 15 THA M Y 9 ( ) ALVAL 16 THA F 0 0 Y b Metallosis 17 THA F Y 2 ( ) Metallosis 18 THA M Y Osteolysis 8 ( ) ALVAL 19 THA M Y Y 5 ( ) ALVAL and metallic debris 20 THA M ( ) Metal hypersensitivity Note ALVAL = aseptic lymphocytic vasculitis associated lesions, RSA = resurfacing arthroplasty, THA = total hip arthroplasty, Y = yes. a ALVAL score according to classification system proposed by Campbell et al. [17] in b Limited tissue available for histologic analysis. associated lesions was assigned in 12 cases (five resurfacing arthroplasty hips and seven THA hips); the remaining seven cases had a final score of less than 5 (Table 3). Cases with a diagnosis of aseptic lymphocytic vasculitis associated lesions had a significantly (p = 0.04) higher mean volume of synovitis (123.8 ± cm 3 ; range, cm 3 ) than cases without that diagnosis (26.4 ± 46.4 cm 3 ; range, cm 3 ) and had a significantly (p = 0.02) higher rate of decompression of synovitis into adjacent bursae (83.3% vs 20.0%, respectively). The most common synovial pattern in patients with histologically confirmed aseptic lymphocytic vasculitis associated lesions was fluid signal intensity with a thin (< 5 mm) intermediate-signal-intensity pseudocapsule (three cases) or Other, representing fluid signal intensity with a thickened intermediate-signalintensity pseudocapsule (five cases). In four cases, intermediate- to low-signal-intensity A Fig year-old man who presented for follow-up imaging 41 months after metal-on-metal total hip arthroplasty. Images show case of synovitis classified as Other. A, Coronal fast spin-echo image shows synovitis as fluid signal intensity outlined by thickened irregular pseudocapsule (arrows). B, Photomicrograph of corresponding histologic specimen shows thickened synovium (top) with fibrinous exudate superficially (arrow), necrotic and infarcted zone (one asterisk), and deep zone with inflammatory cell infiltrate (two asterisks). Final diagnosis was aseptic lymphocytic vasculitis-associated lesions (score 9). B 890 AJR:199, October 2012

8 MRI of Painful Metal-on-Metal Hip Prostheses debris was also seen within the joint; two of these cases had a mixed pattern of aseptic lymphocytic vasculitis associated lesions and metallic debris on histology. Extracapsular disease was observed in three cases in the group with aseptic lymphocytic vasculitis associated lesions (25%) and 0 cases in the group without the lesions. The presence of tendinosis or tendon tears did not differ between the group with a diagnosis of aseptic lymphocytic vasculitis associated lesions and the group without that diagnosis. Intramuscular edema was observed only in cases with aseptic lymphocytic vasculitis associated lesions (Table 3). Discussion MRI is well suited to depict complications after placement of a metal-on-metal hip prosthesis because of its high soft-tissue contrast and direct multiplanar capabilities. Previous MRI studies have shown that periprosthetic soft-tissue masses or fluid collections may correlate to aseptic lymphocytic vasculitis associated lesions [4 6, 19]; however, the MRI appearance of aseptic lymphocytic vasculitis associated lesions has not been clearly defined. An MRI grading system for TABLE 3: Demographic and MRI Findings in Patients Who Underwent Revision Surgery Classified by Presence of Aseptic Lymphocytic Vasculitis Associated Lesions (ALVAL) Feature No ALVAL (n = 7) ALVAL (n = 12) Difference Between Groups (p) Demographics Age, y (mean ± SD) 53.0 ± ± Sex, female-to-male ratio 4:3 7: BMI (mean ± SD) 28.1 ± ± Interval between arthroplasty and MRI, mo (mean ± SD) 27.3 ± ± Synovitis No. (%) of hips 5 (71.4) 12 (100) 0.12 Volume, mm 3 (mean ± SD) 26.4 ± ± a Pattern of synovitis Fluid 0 3 Debris 1 0 Mixed 3 4 Other 1 5 Decompression of synovitis No. (%) of hips 1 (14.3) 10 (83.3) 0.02 a Osteolysis No. (%) of hips 0 (0) 4 (33.3) 0.24 Volume, mm 3 (mean ± SD) 6.3 ± 6.8 Extracapsular disease No. (%) of hips 0 (0) 3 (25.0) 0.26 Volume, mm 3 (mean ± SD) 35.4 ± 31.5 Neurovascular compression No. (%) of hips 1 (14.3) 2 (16.7) 1.0 Abductors Tendinosis score a (mean ± SD) 2.3 ± ± Tear score b (mean ± SD) 1.2 ± ± No. of hips with edema 0 1 No. of hips with atrophy 1 3 Iliopsoas Tendinosis score a (mean ± SD) 1.6 ± ± Tear score b (mean ± SD) 1.0 ± ± No. of hips with edema 0 2 No. of hips with atrophy 0 0 Note BMI = body mass index. a Mean tendinosis score calculated with no degeneration (n = 1), mild degeneration (n = 2), moderate degeneration (n = 3), and severe degeneration (n = 4). b Mean tear score calculated with no tear (n = 1), partial tear (n = 2), and full tear (n = 3). AJR:199, October

9 Hayter et al. metal-on-metal disease has been proposed [20]; however, this MRI grading system has not been correlated to patient outcome or histologic results [20]. We found that synovitis occurs in similar proportions of patients with a painful hip after resurfacing arthroplasty and after THA. This finding confirms the results of prior studies, which have shown periprosthetic fluid collections or soft-tissue lesions in % of patients with a metal-on-metal prosthesis [5, 6, 20]. The higher incidence of synovitis observed in our study when compared with some previous studies may relate to the use of the prototype MAVRIC sequence, which has been shown to improve the depiction of synovitis around a hip prosthesis [12]. In 20% of the cases in our study, synovitis would not have been detected on the FSE images alone. The most common synovial pattern in patients with histologically confirmed aseptic lymphocytic vasculitis associated lesions in both the resurfacing arthroplasty and THA groups was fluid signal intensity with a thin (< 5 mm) intermediate-signal-intensity pseudocapsule or fluid signal intensity with a thickened intermediate-signal-intensity pseudocapsule. Synovitis was observed to decompress into adjacent bursae in a similar proportion of patients in the resurfacing arthroplasty and THA groups. The observation of a fluid collection or soft-tissue mass in the iliopsoas or trochanteric bursa in the setting of a metal-on-metal hip prosthesis should therefore prompt a careful search for a communication with the pseudocapsule. A previously published study of 20 patients with metal-onmetal implants placed at resurfacing arthroplasty reported that solid disease was more common in the iliopsoas bursa and cystic disease was more common in the trochanteric bursa [19]. We found no such correlation between the location of synovial decompression and the presence of solid or cystic disease in our cohort of 60 hips. Although decompression of synovitis was common, true extracapsular disease was uncommon, occurring in two resurfacing arthroplasty and three THA hips. Three of the five patients with extracapsular disease had undergone revision surgery; histology revealed aseptic lymphocytic vasculitis associated lesions in all three cases and the additional finding of metallic debris in two of these cases. All cases of extracapsular disease appeared on MRI as low-signal-intensity lesions within the hip abductors or short external rotator muscles. The finding of extracapsular low-signalintensity lesions should therefore alert the radiologist to the possibility of aseptic lymphocytic vasculitis associated lesions with or without metallic debris. In our study, we found that osteolysis was uncommon compared with synovitis; these findings confirm the hypothesis that the primary abnormality in patients with aseptic lymphocytic vasculitis associated lesions is generated in the soft tissues. Of the 12 patients with confirmed aseptic lymphocytic vasculitis associated lesions, radiographic findings were normal in seven. This further confirms the usefulness of MRI for assessing patients with metal-on-metal hip prostheses placed at arthroplasty because extensive soft-tissue abnormalities may occur even though radiographs show normal findings. The presence of tendon abnormalities and regional muscle atrophy has previously been proposed to be a marker of severe metal-onmetal disease [6, 20]. However, we found that tendinosis is extremely common after metal-on-metal hip arthroplasty, confirming the results of studies of conventional metal-onpolyethylene hip arthroplasty [21]. We found no correlation between tendinosis or tendon tears and the presence of aseptic lymphocytic vasculitis associated lesions at revision surgery, reflecting the finding that incidental tendon abnormalities are common in patients after hip arthroplasty. Although tendon avulsion secondary to aseptic lymphocytic vasculitis associated lesions may indicate that the need for revision surgery is urgent, the finding of a tendon abnormality per se does not indicate an adverse tissue reaction in association with a metal-on-metal prosthesis. Atrophy of the short external rotator muscles was an extremely frequent finding, confirming the results of a study of asymptomatic patients after conventional metalon-polyethylene hip arthroplasty [16]. Short external rotator muscle atrophy is therefore an expected postoperative finding, particularly when a posterior approach is used for arthroplasty placement. Muscle edema was an uncommon finding, occurring in only five hips. In two cases, edema was observed in the abductor muscles due to an acute fracture of the greater trochanter. In the remaining three hips with muscle edema, aseptic lymphocytic vasculitis associated lesions were confirmed at revision surgery. There are some limitations to our study. The MRI examinations were evaluated by consensus agreement of two radiologists. This method did not allow us to assess interobserver reliability; however, that assessment was not the aim of our study. We did perform coefficient-of-repeatability analyses for osteolysis and synovitis volumes; these results showed good agreement between the two observers with respect to volume measurements. The histologic results of patients who underwent revision surgery may have been limited by the variable amounts of tissue sampled at the time of surgery. To date, we have histologic correlation in only 19 patients. Our early results suggest that a greater volume of synovitis, the presence of extracapsular disease, and the finding of intramuscular edema without a known cause such as fracture or infection may indicate aseptic lymphocytic vasculitis associated lesions; future studies with a larger number of patients are required to evaluate whether these MRI findings are reliable markers for aseptic lymphocytic vasculitis associated lesions. In conclusion, we found that MRI is a valuable tool in the assessment of the patient with a painful metal-on-metal hip prosthesis, often showing extensive soft-tissue abnormalities despite the presence of normal radiographic findings. MRI can detect synovitis, osteolysis, and extracapsular disease as well as secondary complications such as neurovascular compression due to synovitis. Acknowledgments We thank Friedrich Boettner, Robert Buly, and Edwin Su for contributing patients to to this study and Parina Shah for assistance in performing the MAVRIC scans. References 1. Amstutz HC, Le Duff MJ, Campbell PA, Wisk LE, Takamura KM. Complications after metalon-metal hip resurfacing arthroplasty. Orthop Clin North Am 2011; 42: , viii 2. Callaghan JJ, Cuckler JM, Huddleston JI, Galante JO. How have alternative bearings (such as metalon-metal, highly cross-linked polyethylene, and ceramic-on-ceramic) affected the prevention and treatment of osteolysis? J Am Acad Orthop Surg 2008; 16(suppl 1):S33 S38 3. Hart AJ, Sabah S, Henckel J, et al. The painful metal-on-metal hip resurfacing. J Bone Joint Surg Br 2009; 91: Pandit H, Glyn-Jones S, McLardy-Smith P, et al. Pseudotumours associated with metal-on-metal hip resurfacings. J Bone Joint Surg Br 2008; 90: Sabah SA, Mitchell AW, Henckel J, Sandison A, Skinner JA, Hart AJ. Magnetic resonance imaging findings in painful metal-on-metal hips: a prospective study. J Arthroplasty 2011; 26: AJR:199, October 2012

10 MRI of Painful Metal-on-Metal Hip Prostheses 6. Toms AP, Marshall TJ, Cahir J, et al. MRI of early symptomatic metal-on-metal total hip arthroplasty: a retrospective review of radiological findings in 20 hips. Clin Radiol 2008; 63: Willert HG, Buchhorn GH, Fayyazi A, et al. Metal-on-metal bearings and hypersensitivity in patients with artificial hip joints: a clinical and histomorphological study. J Bone Joint Surg Am 2005; 87: Ollivere B, Darrah C, Barker T, Nolan J, Porteous MJ. Early clinical failure of the Birmingham metal-on-metal hip resurfacing is associated with metallosis and soft-tissue necrosis. J Bone Joint Surg Br 2009; 91: Potter HG, Nestor BJ, Sofka CM, Ho ST, Peters LE, Salvati EA. Magnetic resonance imaging after total hip arthroplasty: evaluation of periprosthetic soft tissue. J Bone Joint Surg Am 2004; 86: Walde TA, Weiland DE, Leung SB, et al. Comparison of CT, MRI, and radiographs in assessing pelvic osteolysis: a cadaveric study. Clin Orthop Relat Res 2005; 437: Weiland DE, Walde TA, Leung SB, et al. Magnetic resonance imaging in the evaluation of periprosthetic acetabular osteolysis: a cadaveric study. J Orthop Res 2005; 23: Hayter C, Koff M, Shah P, Koch K, Miller T, Potter H. MRI after arthroplasty: comparison of MA- VRIC and conventional fast spin-echo techniques. AJR 2011; 197:592; [web]w405 W Koch KM, Lorbiecki JE, Hinks RS, King KF. A multispectral three-dimensional acquisition technique for imaging near metal implants. Magn Reson Med 2009; 61: Potter HG, Foo LF. Magnetic resonance imaging of joint arthroplasty. Orthop Clin North Am 2006; 37: , vi vii 15. Sofka CM, Potter HG. MR imaging of joint arthroplasty. Semin Musculoskelet Radiol 2002; 6: Pellicci PM, Potter HG, Foo LF, Boettner F. MRI shows biologic restoration of posterior soft tissue repairs after THA. Clin Orthop Relat Res 2009; 467: Campbell P, Ebramzadeh E, Nelson S, Takamura K, De Smet K, Amstutz HC. Histological features of pseudotumor-like tissues from metal-on-metal hips. Clin Orthop Relat Res 2010; 468: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: Fang CS, Harvie P, Gibbons CL, Whitwell D, Athanasou NA, Ostlere S. The imaging spectrum of peri-articular inflammatory masses following metal-on-metal hip resurfacing. Skeletal Radiol 2008; 37: Anderson H, Toms AP, Cahir JG, Goodwin RW, Wimhurst J, Nolan JF. Grading the severity of soft tissue changes associated with metal-on-metal hip replacements: reliability of an MR grading system. Skeletal Radiol 2011; 40: Pfirrmann CW, Notzli HP, Dora C, Hodler J, Zanetti M. Abductor tendons and muscles assessed at MR imaging after total hip arthroplasty in asymptomatic and symptomatic patients. Radiology 2005; 235: FOR YOUR INFORMATION This article is available for CME credit. Log onto click on AJR (in the blue Publications box); click on the article name; add the article to the cart; proceed through the checkout process. AJR:199, October