Inflammatory changes of hip synovial structures in polymyalgia rheumatica F. Cantini1, L. Niccoli1, C. Nannini1, A. Padula2, I. Olivieri2, L. Boiardi3, C. Salvarani3 1II Divisione di Medicina, Unità Reumatologica, Ospedale Misericordia e Dolce di Prato; 2Dipartimento di Reumatologia, Ospedale S. Carlo di Potenza ed Ospedale Madonna delle Grazie di Matera; 3Divisione di Reumatologia, Arcispedale S. Maria Nuova, Reggio Emilia, Italy. Abstract Objective To investigate the hip inflammatory lesions and to evaluate the accuracy of clinical examination compared to magnetic resonance imaging (MRI) in patients with polymyalgia rheumatica (PMR) with pelvic girdle symptoms. Secondary end-point was to evaluate the sensitivity and specificity of ultrasonography (US) compared to MRI in the assessment of hip lesions. Methods Case-control study of 20 consecutive PMR patients and 40 controls with different rheumatic conditions. Both groups were clinically assessed for the presence of hip synovitis, trochanteric, iliopsoas and ischiogluteal bursitis. Hip MRI was performed in all case-patients and in 10 controls. Both groups were examined by US. An additional group of 10 healthy controls was examined by hip US. Results Both MRI and US detected trochanteric bursitis in 100% of PMR patients, bilateral in 18/20 (90%), and in 12/40 (30%) controls (p < 0.001). Hip synovitis was detected in 17/20 (85%) by MRI and in 9/20 (45%) by US (p < 0.02) in case- patients and in 18/40 (45%) controls. In PMR, MRI and US showed iliopsoas bursitis in 10/20 (50%) and 6/20 (30%) and ischiogluteal bursitis in 5/20 (25%) and 4/20 (20%) with no differences compared to controls. Clinical examination showed a good accuracy for hip synovitis, trochanteric and ischiogluteal bursitis, while it overestimated the presence of iliopsoas bursitis. US was less sensitive than MRI for the detection of hip synovitis and iliopsoas bursitis (53% and 60%). Conclusion Trochanteric bursitis represents the most frequent hip lesion in PMR. Acareful physical examination allows to detect all inflammatory lesions excluding iliopsoas bursitis. US is less sensitive than MRI in the assessment of hip synovitis and iliopsoas bursitis. Key words Polymyalgia rheumatica, MRI, ultrasonography, bursitis, synovitis. Clinical and Experimental Rheumatology 2005; 23: 462-468. Hip inflammatory changes in PMR /F. Cantini et al. Fabrizio Cantini, MD; Laura Niccoli, MD; Introduction maximum muscle and the posterolater- Carlotta Nannini, MD; Angela Padula, Proximal musculoskeletal pain and al surface of the greater trochanter. It is MD; Ignazio Olivieri, MD; Luigi Boiardi, stiffness in the neck, shoulder and pel- usually a multilocular, collapsed bursa MD; Carlo Salvarani, MD. vic girdles in association with elevated whose inner surface is covered by syn- Please address correspondence and acute-phase reactants is the clinical ovial tissue (10). reprint requests to: Dr. Fabrizio Cantini, hallmark of polymyalgia rheumatica Iliopsoas bursa. The iliopsoas (or ileo- II Divisione di Medicina, Unità Reuma- tologica, Ospedale Misericordia e Dolce (PMR). Pelvic girdle is involved in pectineal) bursa, which is present and di Prato, Piazza Ospedale no. 1, 59100 about 60% of the cases causing pain bilateral in 98% of adults, lies between Prato, Italy. and stiffness in the back and hips (1). the deep surface of the ileopsoas mus- E-mail: [email protected] or From proximal areas the discomfort cle and the anterior capsule of the hip. [email protected] radiates distally along the thighs It is the largest bursa in the body and is Received on October 5, 2004; accepted toward the knees. Pain is usually felt on flanked by the femoral artery, vein and in revised form on April 1, 2005. the anterior and lateral aspect of the nerve. The iliopsoas bursa communi- © Copyright CLINICALAND EXPERIMEN- thigh and less frequently posteriorly. It cates with the hip joint cavity in 14% of TAL RHEUMATOLOGY 2005. is more pronounced in the morning and adult cadavers and it is collapsed un- may become severe enough to cause less distended by fluid collection sec- difficulty getting up from bed or from a ondary to inflammatory involvement chair (2). (11). Hip joint synovitis has been considered Ischiogluteal bursa (or ischiatic bur - responsible for the proximal pelvic sa). This bursa is situated over the is- symptoms experienced by PMR pa- chiatic tuberosity and overlies the scia- tients (3,4). Recent magnetic resonance tic and the posterior femoral cutaneous imaging (MRI) and ultrasonography nerves (9, 12). (US) studies have shown that PMR mainly affects the extra-articular syn- Patients and methods ovial structures of the shoulders (5-7). Disease and control patients Similarly to that observed in the shoul- All consecutive, untreated new patients ders, other synovial structures such as with PMR according to the Healey cri- bursae of the pelvic girdle may be teria (13) who had pelvic girdle involve- involved by the inflammatory process ment observed in the outpatient clinic of PMR (8). of 3 secondary rheumatology centers We designed a case-control, magnetic (Hospitals of Prato, Reggio Emilia and resonance imaging (MRI) and ultra- Potenza) during a 6-month period were sonography (US) study with two pri- included in the study. mary end-point: to investigate the hip Patients treated with corticosteroids lesions in a clinical series of patients prior to the clinical evaluation, those with untreated PMR and to evaluate the with positivity of rheumatoid factor accuracy of clinical examination in the (Rose-Waaler titer ³ 40 or nephelomet- assessment of different articular and ric determination 20 IU/ml on 2 or soft tissue inflammatory changes com- more occasions) and/or fluorescent an- pared to MRI. As secondary end-point tinuclear antibody testing (titer ³ 1:32) we assessed the sensitivity and specifi- were excluded from the study. Tempo- city of US with respect to MRI to detect ral artery biopsies were performed only the hip inflammatory lesions in PMR. in patients with cranial signs and/or symptoms suggestive of giant cell ar- Essential anatomy of the synovial teritis (GCA). At diagnosis and over structures of the hip the follow up, all patients were evaluat- Coxo-femoral joint (hip joint). The sy- ed for fulfillment of American College novial membrane lines the deep surface of Rheumatology (ACR; formerly, the of the articular capsule. Proximally, it American Rheumatism A s s o c i a t i o n ) covers the surface of the cartilagineous 1987 revised criteria for rheumatoid rim of acetabolum and encloses the lig- arthritis (RA) (14). amentum teres. Distally, the synovial The two consecutive outpatients with membrane extends up to the cartilagi- active rheumatic diseases, older than neous surface of the femoral head (9). 50, with bilateral hip aching who were Trochanteric bursa. The trochanteric observed after the patient with PMR bursa is situated between the gluteus served as disease controls. An addition- 463 Hip inflammatory changes in PMR / F. Cantini et al. al group of 10 healthy controls consti- formed with a median interval from diagnosed when a distance of 7 mm or tuted by 6 females and 4 males with a US evaluation of 7 days (range: 1-10 more between the joint capsule and the mean age of 61 ± 8.1 years was evalu- days). The median interval between femur or a difference of more than 1 ated. clinical examination and imaging as- mm between the two hips were mea- se s s ment was 6 days (range: 2-9 days). sured and the side with the greater dis- Clinical assessment Healthy controls were examined only tance was considered as indicating syn- All case and control patients were clin- by US. ovitis (18). Hip synovitis was graded at ically assessed and prospectively fol- The equipment used for US was the US as done for MRI scans (see below). lowed up by the same rheumatologists Toshiba SSA340A(Tokyo, Japan) with Bursae inflammation was diagnosed by who recorded medical informations on a 7.5 MHz linear transducer. US if an hypohecoic area suggesting a standardized collection form at every All US examinations were performed fluid collection was detected in the ana- visit. Specifically, patients were assess- together by 2 radiologists with special tomical sites of trochanteric, ileopsoas ed for the presence of clinical signs of training in musculoskeletal sonogra- and ischiatic bursae (20-24). The hypo- hip synovitis (coxo-femural joint syn- p h y. They were blind to the clinical echoic area was measured in mm2. Flu- ovitis), trochanteric bursitis, iliopsoas diagnosis and agreed on the US find- id collection was graded 0, 1, 2 or 3 if bursitis, ischiogluteal bursitis defined ings. Hip sonograms were obtained ac- the hypoechoic area was absent, less as follows: cording to previously standardized than 50, between 50 and 100 and more - hip synovitis: pain in the groin even- techniques (18). than 100 mm2, respectively. tually radiating to the anterior and MRI scans were also evaluated by 2 ra- As previously reported (5, 6), for MRI medial thigh toward the knee wors- diologists. To compare the findings, the scans, measurement of fluid accumula- ened by passive and active move- radiologists evaluated the US and MRI tion was graded by using a semiquanti- ments and associated with positivity scans on an independent basis and tative scale (0 = no accumulation; 1 = of Fabere test (9); blind to the clinical diagnosis and to the sufficient accumulation to allow visu- - trochanteric bursitis: deep, aching reciprocal results. To reach a consen- alization of the articular shoulder struc- pain on the lateral aspect of the hip sus, the radiologists evaluated all scans, ture, periarticular shoulder structure, or and thigh increased by external rota- and after the blinded phase compared both; 2 = moderate accumulation; 3 = tion and abduction of the hip associ- the reciprocal evaluations. As regards sufficient quantity to stretch the walls ated with localized tenderness on the US examinations there were no dis- of the structures).