Background Conventional radiology (X-rays) is still the gold standard in hand osteoarthritis (HOA), as sparse literature data seems to reflect mediocre interest in magnetic resonance (MR) imaging.

Objectives Since no scoring method has been yet standardized for HOA, the Oslo HOA MR imaging score was proposed to define the disease’s key features which are osteophytes, joint space narrowing (JSN), erosions, cysts, malalignment, normally assessed by plain X-rays. Other aspects such as synovitis, bone marrow edema, flexor tenosynovitis, collateral ligaments can be evaluated only by MR. In this study we enrolled patients, referring at the Clinic Rheumatology Unit, University of Padua, with hand pain and early disease onset (mean disease duration<3 years))

Methods Eleven patients with early severe erosive HOA (EHOA) were studied. All patients were female, (average age 55±9.5 years; disease onset 45.4±9.48 years). HOA diagnosis was defined using American College of Rheumatology (ACR) criteria, EHOA was diagnosed in presence of proximal (PIP) or distal (DIP) interphalangeal erosions with central subcondral collapse in the same area at X-rays. All the patients underwent hand clinical examination and were evaluated with VAS, questionnaires, the grip strength test and traditional X-rays (Tab. 1). MR with contrast enhancement was performed on the dominant hand. Two radiologists used the Oslo HOA MR imaging score to assess the MR images of the 2nd-5th IP joints and the Kellgren-Lawrence (KL) score to assess the X-rays (Tab.2). The dominant hand was considered for the correlations.

Results The presence of nodes correlated with KL score (r= 0.61, p=0.04), and with DIP erosion (r=0.62; p=0.038) and DIP osteophytes (r=0.69; p=0.01) by RM. Comparing clinical features and MR the tender joints correlated with DIP erosions (r=0.70; p=0.01), and with sagittal malalignment (r=0.63; p=0.03). Disease duration correlated with JSN (r=0.63, p=0.03), with DIP (r= 0.64; p=0.03) and PIP erosions (r=0.78; p=0.003), likewise VAS correlated with DIP (r=0.65; p=0.02) and PIP erosions (r=0.73; p=0.009). Comparing radiographs and MR, KL scores and bone edema showed a correlation (r=0.65; p=0.028). Considering the total number of joints scored by MR, synovitis was found in 39% (35/88), and it was more frequently detected in PIP compared to DIP joints (25/44 vs 10/44) similarly bone edema was more frequent in PIP vs DIP joints (12/44 and 14/44vs6/44 and 7/44).

Conclusions The results confirm the correlation between nodes and KL, and between tenderness and erosions by MR. MR appear to be more sensitive with respect to X-rays, as evidenced by its relationship with early symptoms (terderness). MR could therefore be employed to differentiate EHOA from the HOA and to make early diagnosis of the erosive one.

Disclosure of Interest None Declared