Background Early and accurate detection of erosive joint changes is important when validating and monitoring rheumatoid arthritis (RA). Magnetic resonance imaging (MRI) and ultrasound (US) have shown to be more sensitive than conventional radiography (CR) but have the drawbacks of being expensive (MRI) or difficult to learn (US). In several studies these methods have been compared to multidetector computed tomography (MDCT), which can be considered a reference method for detecting erosions but has several disadvantages, for example the exposure to ionising radiation. With the newly developed cone-beam CT (CBCT) which is dedicated for imaging of the extremities, these disadvantages can be minimized. The radiation dose is low, about 10% of the effective dose of MDCT, comparable to an extra CR exposure.
Objectives To assess the intra-and interobserver agreement of erosions detected and scored with CBCT. To compare CBCT with CR for assessment of bone erosions in patients with established rheumatoid arthritis.
Methods 30 patients (23 females, 7 males) with established RA – median disease duration (min-max) 15 (14–15) year – who had CR performed according to the Better AntiRheumatic PharmacOTherapy (BARFOT) study protocol were examined also with extremity CBCT (hands and feet) at the same occasion. Bone erosions were scored twice according to RA MRI score (RAMRIS), on digital images in three planes from the CBCT by two observers. The radiographs were scored by a third observer with Sharp van der Heijde Score (SHS). Disease activity was assessed with disease activity score of 28 joints (DAS28). Remission was defined as DAS28<2.6. Comparisons between groups was performed using Mann-Whitney U test. Correlations were performed with Spearman correlation test and intra class correlation coefficient (ICC) was calculated.
Results The median age (min-max) was 54 (27–75) and median DAS28 (min-max) 2.25 (0.63–6.49). All 30 patients showed erosions on CBCT and 26 on CR. The correlation between CBCT and CR was 0.77 (p<0.001). Intra-and interobserver reliability for erosions scored by CBCT was excellent for both hands and feet for the two observers (ICC 0.90–0.97).The CBCT erosion scores of 16 patients in remission was median RAMRIS (min-max) 9 (1–67) vs. 21 (2–224), p=0.077 for 11 patients not in remission. The results for CR were for patients in remission median SHS (min-max) 12 (0–97) and those not in remission 40 (6–151), p=0.034. The median (min-max) numbers of eroded joints were 8 (1–40) with CBCT vs. 2 (0–20) with CR for patients in remission and for those not in remission 16 (2–68) with CBCT, vs. 7 (0–34) with CR. There were significantly more eroded joints detected by CBCT, p=0.002.
Conclusions CBCT is a new low-dose 3 D imaging technique, which has high reproducibility and is more sensitive than CR in detecting erosions. CBCT has a potential to become a useful tool in detecting and follow up of erosions in patients with RA.
Disclosure of Interest None declared