Differential diagnosis of inflammatory arthropathies by musculoskeletal ultrasonography: A systematic literature review

Background: Differential diagnosis in early arthritis is challenging, especially early after symptom onset. Several studies applied musculoskeletal ultrasound in this setting, however, its role in helping diagnosis has yet to be clearly defined. The purpose of this work is to systematically assess the diagnostic applications of ultrasonography in early arthritis in order to summarize the available evidence and highlight possible gaps in knowledge. Methods: In December 2017, existing systematic literature reviews (SLR) on rheumatoid arthritis (RA), osteoarthritis (OA), psoriatic arthritis (PsA), polymyalgia rheumatica (PMR), calcium pyrophosphate deposition disease (CPPD), and gout were retrieved. Studies on ultrasound to diagnose the target conditions and detecting elementary lesions (such as synovitis, tenosynovitis, enthesitis, bone erosions, osteophytes) were extracted from the SLRs. The searches of the previous reviews were updated and data from new studies fulfilling the inclusion criteria extracted. Groups of reviewers worked separately for each disease, when possible diagnostic accuracy (sensitivities, specificities) was calculated from primary studies. When available, the reliability of ultrasound to detect elementary lesions was extracted. Results: For all the examined disease, recent SLRs were available. The new searches identified 27 eligible articles, with 87 articles included from the previous SLRs. The diagnostic performance of ultrasound in identifying diseases was addressed by 75 studies; in most of them, a single elementary lesion was used to define diagnosis, except for PMR. Only studies on RA included consecutive patients with new onset of arthritis, while studies on gout and CPPD often focused on subjects with mono-arthritis. Most of the remaining studies enrolled patients with a defined diagnosis. Synovitis was the most frequently detected lesion; clinical diagnosis was the most common reference standard. The diagnostic performance of ultrasound across different conditions was extremely variable. Ultrasound to identify elementary lesions was assessed in 38 studies in OA, gout and CPPD. Its performance in OA was very variable, with better results in CPPD and gout. The reliability of ultrasound was moderate to good for most lesions. Conclusions: Although a consistent amount of literature investigated the diagnostic application of ultrasound, in only a minority of cases its additional value over clinical diagnosis was tested. This SLR underlines the need for studies with a pragmatic design to identify the placement of ultrasound in the diagnostic pathway of new-onset arthritis.