Background/Purpose: Through the clinical
experience, we noted that there seems to be differences not in the kind or the
frequency but in the severity of synovial inflammation between polymyalgia rheumatica (PMR) and elderly-onset rheumatoid arthritis with
PMR-like onset (polymyalgic-EORA). We compared
the severity of the shoulder syonvitis between PMR
and polymyalgic-EORA using semi-quantitative US scoring of the synovial components, and also explored the application
of the scoring system to the differential diagnosis between the two
diseases.

Methods: We analyzed consecutive records of
15 PMR patients and 15 polymyalgic-EORA patients. All of them were examined by US before treatment-start. The
severity of tenosynovitis of the long-head of the biceps (LHB), bursitis of
shoulder including subdeltoid, subacrominal
and subcoracoid bursitis, and joint synovitis of grenohumeral joint
(GHJ) were subjectively scored for grey-scale (GS) and power Doppler (PD) on a
four-point scale: 0 = absent, 1 = mild, 2 = moderate or 3 = severe. Sum of the
scores of one shoulder was defined as “shoulder synovitis
score (SSS)”, and sum of the scores of both shoulders was defined as “patient-shoulder
synovitis score (PSSS)”. The correlation between PSSS
and serum markers was assessed. Indices contributing to discriminating between the two
diseases were explored.

Results: There
was a tendency that synovitis in each synovial component in the shoulder in PMR
was milder than that in polymyalgic-EORA, while there was no difference in the frequencies of synovitis (Table). There
were statistically significant differences in the PD-grade of LHB, GS-grade of LHB,
bursitis and GHJ. There was also a significant difference in
GS-score of bursitis (0.40 ± 0.62 vs 1.2 ± 1.2,
p=0.003), PD-score of bursitis (0.37 ± 0.72 vs 1.0 ± 1.2,
p=0.015) and SSS (4.0 ± 2.2 vs 6.2 ± 4.6,
p=0.023) (mean ± SD, PMR vs polymyalgic-EORA).
PSSS in PMR tended to be lower than that in polymyalgic-EORA (7.7 ± 3.8 vs
12.3 ± 8.6, p=0.072).

PSSS were
positively correlated with serum MMP3 (|R|=0.707, p< 0.0001). Both PSSS (|R|=0.602,
p= 0.0175) and MMP3 (|R|=0.463, p= 0.095) were positively correlated with serum
CRP in polymyalgic-EORA but not in PMR. The ratio of
PSSS to CRP (mg/dL) (PSSS/CRP) was significantly lower
in PMR than in polymyalgic-EORA (1.35 ± 1.63 vs 5.55 ± 4.46, p=0.003). The ratio of MMP3 (ng/mL) x10⁴ to CRP (MMP3/CRP) was significantly
lower in PMR than in polymyalgic-EORA (32.2 ± 35.9 vs 101 ± 87.1, p=0.006). These two ratios may be useful for
discriminating between the two diseases, because ROC analysis for PSSS/CRP and MMP3/CRP demonstrated areas under the
curve of 0.902 and 0.852, respectively.

Conclusion: Semi-quantitative
US scoring of the synovial components in the shoulder revealed
that shoulder synovitis in PMR is milder than
that in EORA patients with PMR-like onset. The application of the scoring
system may be useful for discriminating between the two
diseases.

a, p<0.05; b, p<0.01