Ultrasonographic criteria for the diagnosis of erosive rheumatoid arthritis using osteoarthritic patients as controls compared to validated radiographic criteria.

ABSTRACT

OBJECTIVES: The aims of this study were to compare characteristics of radiography (RX) and ultrasound (US) erosive lesions in rheumatoid arthritis (RA) and osteoarthritis (OA) patients (prevalence, topography and severity), to determine thresholds for the diagnosis of erosive RA based on US and to evaluate the performance of US and RX to establish a diagnosis of erosive RA differentiated from hand OA. METHODS: Patients fulfilling ACR 1987 and/or ACR/EULAR 2010 criteria for RA or ACR hand OA criteria were prospectively included. A modified Sharp erosion score was assessed by two blinded readers and one adjudicator for discordant cases (number of eroded joints ≤ three). Erosions in US were scored on six bilateral joints (MCP2-3, 5; MTP2-3, 5) with a four-grade scale to calculate total US score for erosions (USSe). RESULTS: A total of 168 patients were included 122 RA (32 early RA < 2 years; 90 late RA ≥ 2 years); 46 OA patients. On RX 42 RA patients (6 early; 36 late) and 5 OA patients were eroded according to EULAR 2013 definition criteria with sensitivity at 34.4% and specificity at 89.1%. On US, 95 RA patients (21 early; 74 late) and 12 OA patients were eroded. Considering at least two joint facets eroded or at least one joint facet eroded at grade 2 on US, sensitivities were good (68-72.1%) and specificities excellent (89.1-100%). Agreement between RX and US was excellent (90-92%). The positive and negative likehood ratios were respectively 3.16 and 0.73 for radiography and 6.64 and 0.31 for US (for two facets eroded). CONCLUSION: USSe can differentiate RA from OA in erosive disease and detect two times more patients with erosive RA than RX with excellent specificity and agreement.