The cause of musculoskeletal symptoms in the proximal extremities is not completely understood. Evidence of joint synovitis has been revealed through scanning, arthroscopy, and synovial biopsy [5-8]. In a recent immunohistochemical study [8], researchers observed mild synovitis characterized by infiltration of macrophages and CD4 T cells. Diffuse and severe musculoskeletal discomfort of the proximal extremities can only be partially explained by this mild joint synovitis. In addition to involvement of the pelvic girdle, some patients have symptoms in the distal extremities that are caused by inflammation of the joints, inflammation of the tenosynovial membrane, or inflammation of both. Clinical evidence of peripheral synovitis was observed in 31% to 38% of patients who had polymyalgia rheumatica [2, 9]. In a recent study [10] conducted by the Mayo Clinic in 245 patients who had polymyalgia rheumatica, 19 patients (8%) had diffuse swelling and pitting edema in the distal extremities. The swelling and edema were similar to those observed in patients who had the remitting, seronegative, symmetrical synovitis with pitting edema syndrome (described by McCarty and colleagues in 1985 [11]). The authors of the study concluded that these clinical findings were most likely the result of vigorous tenosynovitis in the distal extremities and represent a symptom of polymyalgia rheumatica that had previously been poorly recognized.

We recently observed a patient who satisfied diagnostic criteria of polymyalgia rheumatica and had bilateral diffuse swelling with pitting edema on the dorsum of the hand [12]. Magnetic resonance imaging (MRI) showed extensor tenosynovitis of the hand and synovitis of the glenohumeral joints together with marked inflammation of subacromial and subdeltoid bursae and tenosynovitis of the biceps in both shoulders. Impressed by the severe involvement of proximal periarticular synovial structures, we decided to use MRI to study the involvement of the shoulders and pelvic girdle in a series of consecutive patients who had symptoms of active polymyalgia rheumatica.

Methods

Top Methods Results Discussion Author & Article Info References

Consecutive patients who were seen at the Prato and Reggio Emilia rheumatology centers during a 6-month period and satisfied the Healey criteria for polymyalgia rheumatica [2] were considered suitable candidates for the study. Table 1 shows the demographic characteristics, clinical findings, and MRI results in the 13 case-patients. None of the case-patients had clinical or histologic evidence of giant cell arteritis. Bilateral MRI of the shoulders of the first three case-patients who entered the study revealed that the lesions were symmetrical and their severity was identical. We therefore decided to perform only monolateral MRI in the subsequent 10 case-patients. Three of the 8 case-patients in whom the hip girdle was involved also had pelvic scanning.

Two control groups were considered. The first group consisted of nine control-patients who had elderly-onset rheumatoid arthritis (median age at onset, 71 years [range, 66 to 76 years]) and met the American Rheumatism Association 1987 modified criteria for rheumatoid arthritis [13]. The control-patients were seen at the Prato and Reggio Emilia rheumatology centers during the same 6-month period as the case-patients. The control-patients had early active disease (median disease duration, 3 months [range, 3 to 6 months]) and clinical evidence of shoulder involvement. The results of serologic examination were positive in three control-patients and negative in six control-patients. The median erythrocyte sedimentation rate at diagnosis was 65 mm/h (range, 56 to 88 mm/h). Bilateral MRI was done on two control-patients who had clinical involvement of both shoulders. Because only one shoulder was clinically involved in the other seven control-patients, monolateral MRI of only the affected shoulder was done. Second-line drug and corticosteroid therapies were not started until MRI had been completed.

The second control group consisted of 10 age-matched healthy controls who did not have any clinical problems with their shoulders. The healthy controls were relatives of medical staff at both rheumatology centers. Monolateral MRI of the shoulder was done in all 10 healthy controls.

Scanning of the shoulders of three case-patients was repeated after they began corticosteroid therapy (median, 2 months [range, 2 to 3 months]) and were in clinical remission.

Scanning was done with a 0.5-T superconductive magnet system (MR Max Plus, GE Medical System, Milwaukee, Wisconsin) and a 17-cm extremity bore transmit-receive coil. A body coil was used to evaluate hip regions. Pulse sequences included coronal T1-weighted sequences (240-ms repetition time, 25-ms echo time, and four excitations), axial proton density sequences (2000-ms repetition time, 25-ms echo time, and two excitations), and T2-weighted sequences (2000-ms repetition time, 90-ms echo time, and two excitations). The coronal section was 5 mm thick, and the axial section was 7 mm thick; both had an intersection gap of 1 mm. The field of view was 20 cm; the matrix size was 160 cm x 224 cm or 128 cm x 192 cm.

Scans were examined by a radiologist who was blinded to clinical findings and the diagnosis. The joint space, subacromial and subdeltoid bursae, and synovial sheaths of the long head of the biceps of the shoulder were evaluated for fluid collection. In addition, the joint space and the ileopectineal bursa in the hip region were evaluated. As shown in Figure 1, measurement of fluid accumulation was graded by using a semiquantitative scale (0 = no accumulation; 1 = sufficient accumulation to allow visualization of the articular shoulder structure, periarticular shoulder structure, or both; 2 = moderate accumulation; and 3 = sufficient quantity to stretch the walls of structures).

View larger version (138K): [in this window] [in a new window]   Figure 1. Magnetic resonance images of patients with polymyalgia rheumatica. Coronal T1-weighted (A) and axial T2-weighted (B) images of the left shoulder show severe subacromial and subdeltoid bursitis (arrows). Coronal T1-weighted (C) image of the left shoulder shows mild subacromial bursitis (arrows). Axial T2-weighted (D) image of the left shoulder shows moderate subdeltoid bursitis (solid white arrows), mild joint effusion (solid black arrow), and severe tenosynovitis of the long head of the biceps (open black arrow).

Statistical analysis was done by using the SPS program (SPS Inc., Chicago, Illinois). The Fisher exact test was used to compare the frequencies.

Results

Top Methods Results Discussion Author & Article Info References

All 13 case-patients (who had active polymyalgia rheumatica) showed bilateral fluid accumulation in the subacromial and subdeltoid bursae, thereby suggesting bursitis (Figure 1 and Table 1). Ten of the 13 case-patients had synovitis of the shoulder joint. Tenosynovitis of the long head of the biceps was found in 7 case-patients. No erosions were observed.

Mild (grade 1) synovitis of the hip was seen in all three case-patients who had scans of the hip girdle. One of the three case-patients also had mild (grade 1) ileopectineal bursitis.

Fluid accumulated in the subacromial and subdeltoid bursae of only two of the nine control-patients (that is, patients who had early symptoms of elderly-onset rheumatoid arthritis). Fluid accumulated in the joint space of five control-patients, and tenosynovitis of the long head of the biceps was observed in three control-patients. Joints were eroded in two control-patients. None of the 10 age-matched healthy controls showed evidence of fluid collection in joints, bursae, or sheaths of the long head of the biceps.

Inflammation in the subacromial and subdeltoid bursae occurred significantly more frequently in case-patients than in control-patients (100% compared with 22%; P < 0.001). The frequencies of joint synovitis and tenosynovitis of the biceps did not significantly differ between case-patients and control-patients (77% compared with 55% and 54% compared with 33%, respectively). However, the frequency of both disorders was higher in the case-patients.

Two of the three case-patients who had repeated MRI of the shoulder during treatment showed complete resolution of bursitis, tenosynovitis, and joint synovitis. The third case-patient showed only an improvement (from grade 3 to grade 2) of bursitis and joint synovitis. All three of these case-patients received a starting dosage of 12.5 mg of prednisone per day. When the second scan was obtained, the patients were asymptomatic, the erythrocyte sedimentation rate was normal (median, 18 mm/h [range, 15 to 20 mm/h]), and the median daily dosage of prednisone was 10 mg/d.

Discussion

Top Methods Results Discussion Author & Article Info References

All 13 case-patients showed evidence of subacromial and subdeltoid bursitis. This finding occurred significantly more frequently in the 13 case-patients than in the nine control-patients. The frequency of joint synovitis and bicipital tenosynovitis did not differ significantly between case-patients and control-patients. Joint erosions have never been observed in patients with polymyalgia rheumatica; in our study, however, erosions were seen in two control-patients (who had elderly-onset rheumatoid arthritis).

According to MRI of normal shoulders [14], no pathologic findings were observed in the age-matched healthy controls. Our MRI study of pelvic girdles was limited because of the small number of study participants. However, one of the three patients had evidence of ileopectineal bursitis in addition to mild joint synovitis.

No published studies have examined the results of MRI of the shoulders of patients with polymyalgia rheumatica. One ultrasonographic study [15] that examined hip and glenohumeral joints focused on an effusion of the two joints, which was found in 68% of the patients examined. A study [16] of MRI of the shoulder of patients who had rheumatoid arthritis did not report evidence of prominent bursitis or tenosynovitis. However, this study was not designed to investigate periarticular synovial structures.

Possible limitations of our study are the small number of patients and the inclusion of case-patients who had polymyalgia rheumatica associated with peripheral synovitis, distal swelling, and pitting edema of the extremities. Despite the limited number of patients studied, our results are strengthened by the inclusion of two control groups (one composed of control-patients and one composed of healthy controls). A more uniform group of patients with polymyalgia rheumatica would have permitted a study of a more pure form of the disease but would not have represented the full spectrum of disease.

Our MRI study shows that subacromial and subdeltoid bursitis is the predominant and most frequently observed lesion in patients with active polymyalgia rheumatica. Inflammation of these two anatomically extended bursae in association with synovitis of the glenohumeral joints and tenosynovitis of the biceps may explain the diffuse discomfort observed in the shoulder girdle of patients with polymyalgia rheumatica.

Three of the 13 case-patients had a second MRI examination after 2 months of adequate corticosteroid treatment. In 2 case-patients, MRI showed complete resolution of disease in the articular and periarticular structures that had previously been involved. In the third case-patient, bursitis and synovitis persisted (although less intensely) despite clinical remission of symptoms. In a recent immunohistochemical study [8], researchers showed the persistence of active synovitis of the shoulder in a patient who had received corticosteroids and was clinically asymptomatic (three relapses of polymyalgia rheumatica were successively observed in this patient during follow-up). In addition, longitudinal studies [17-19] have demonstrated the persistence of altered immunologic factors associated with disease activity in a subgroup of patients with clinically inactive polymyalgia rheumatica. These data suggest that although steroid treatment rapidly controls clinical symptoms of polymyalgia rheumatica, active synovitis persists for a much longer period in some patients.

In conclusion, the prominent involvement of extraarticular synovial structures in association with articular synovitis is a likely basis for much of the discomfort in the proximal extremities of patients with polymyalgia rheumatica.

Dr. Olivieri: Servizio di Reumatologia, Azienda Ospedaliera S. Orsola-Malpighi, Bologna, Italy.

Drs. Barozzi, De Matteis, and Pavlica: Servizio di Radiologia, Azienda Ospedaliera S. Orsola-Malpighi, Bologna, Italy.