Diagnostic usefulness of synovial vascular morphology in chronic arthritis. A systematic survey of 100 cases.

OBJECTIVES

To assess the diagnostic usefulness of the systematic analysis of synovial vascular morphology in various inflammatory, early, and longstanding arthropathies, and to examine the validity of the vascular patterns in predicting the evolution of a group of patients with undifferentiated arthritis (UA).

METHODS

One hundred patients who underwent rheumatologic arthroscopy of a symptomatic joint (85 knees, 11 wrists, 3 elbows, 1 metacarpophalangeal joint) were evaluated. The same observer, blinded to patient diagnosis, analyzed the video recordings of the arthroscopies. Vascular morphology was classified into 3 patterns: straight, tortuous, and mixed.

RESULTS

Eighty-one patients had inflammatory arthritis: 35 rheumatoid arthritis (RA), 16 psoriatic arthritis (PsA), 13 spondyloarthropathies (SpA), and 17 UA. Forty-nine percent of patients with RA had a straight pattern, 28% a mixed, and 23% a tortuous one. The sensitivity rate of the straight pattern for RA was 77% and the specificity rate was 70%. Seventy-six percent of RA patients with a straight pattern were rheumatoid factor positive (RF+) against 25% of RA patients with a tortuous pattern. The odds ratio for RA associated to straight compared with tortuous pattern was 57.3 (95% confidence interval, 6.6 to 499.5; P <.001). Patients with PsA and SpA shared the same pattern and were analyzed as 1 group. Ninety-three percent of patients with PsA/SpA had a tortuous pattern, 4% a straight pattern, and 3% a mixed pattern. The sensitivity rate of the tortuous pattern for PsA/SpA was 61% and the specificity rate was 95%. During 2 years of follow-up, 6 of 17 patients with UA were definitely diagnosed: 4 RA (2 RF+ and straight pattern; 2 with a tortuous pattern, 1 with RF+ and HLA-B27+); 1 SpA and 1 PsA, both with a tortuous pattern. No differences in vascular patterns were observed according to disease duration. Our results indicate that vascular patterns are not modified by disease modifying antirheumatic drug (DMARD) treatment. The other 19 patients had osteoarthritis (n = 8) and calcium pyrophosphate dihydrate crystal deposition disease (n = 11) and their predominant vascular pattern was tortuous-like.

CONCLUSIONS

Arthroscopic assessment of synovial vascular changes in chronic arthritis may be of diagnostic and pathogenetic interest, although differences between published studies suggest a need for consensus in evaluating vascular patterns. A straight pattern is strongly associated with RF + RA whereas a tortuous pattern is generally associated with PsA or SpA; these associations are independent of disease duration. The vascular pattern likely does not change qualitatively with DMARD therapy. The application of this technique to the diagnosis or prognosis of patients with UA may be a complementary tool for the treatment of these patients, but larger, prospective studies are necessary to confirm this hypothesis.