THU0533 A SINGLE MRI DIXON SEQUENCE TO ASSESS DISEASE ACTIVITY AND CARTILAGE IN EARLY RHEUMATOID HANDS: ONE SEQUENCE TO ASSESS THEM ALL?

Abstract

Background:OMERACT recommends three “core set” MRI sequences with an optional cartilage-dedicated sequence to perform Rheumatoid Arthritis (RA) MRI scoring (RAMRIS) (1). Dixon method allows the production of four different images from a single MRI sequence.Objectives:To test a short MRI protocol based on a single Dixon sequence to assess disease activity and cartilage in hands of patients with early RA.Methods:Twenty-four patients (16 women, mean age 45.7 years old) with early DMARD-naive RA meeting the ACR/EULAR 2010 criteria were prospectively included. Both hands of each patient were imaged with MRI including contrast-enhanced T1-weighted Dixon and OMERACT “core set” MRI sequences and with conventional radiography.Two musculoskeletal radiologists (R1 and R2) separately assessed disease activity according to RAMRIS scoring system based on the Dixon images (contrast-enhanced T1-weighted Dixon water-only images to score synovitis, tenosynovitis and osteitis and contrast-enhanced T1-weighted Dixon fat-only images to score erosions) and the three OMERACT “core set” MRI sequences (contrast-enhanced fat-saturated T1-weighted images to score synovitis and tenosynovitis, fat-saturated T2-weighted images to score osteitis and T1-weighted images without contrast-material injection to score erosions).One radiologist (R1) separately measured the thickness of the cartilage in the joints corresponding to those assessed by the Sharp/van der Heijde modified scoring method on contrast-enhanced T1-weighted Dixon out-of-phase images and radiographs (2).RAMRIS scoring and measurement of the cartilage thickness were repeated by R1 to assess intra-observer agreement. Statistical analysis was based on intra-class correlation coefficients (ICC) with 95% confidence interval to assess inter-technique, intra-observer and inter-observer agreement. The strength of agreement was interpreted as follows: ≤0, poor; 0.01-0.20, slight; 0.21-0.40, fair; 0.41-0.60, moderate; 0.61-0.80, substantial and ≥0.81, excellent.Results:Agreement between total RAMRIS scores obtained with the Dixon water- and fat-only images and total RAMRIS scores obtained with the OMERACT sequences was excellent for R1 (0.94; 0.86-0.97) and R2 (0.91; 0.81-0.96). Intra-observer agreement was excellent with Dixon images (0.97; 0.92-0.98) and OMERACT sequences (0.96; 0.90-0.98). Inter-observer agreement was excellent with Dixon images (0.92; 0.82-0.96) and OMERACT sequences (0.93; 0.85-0.97).Agreement between the measures of cartilage thickness on the Dixon out-of-phase images and the measures of cartilage thickness on radiographs was substantial (0.71; 0.66-0.75). Intra-observer agreement was excellent with Dixon out-of-phase images (0.94; 0.93-0.95) and radiographs (0.93; 0.92-0.94).Conclusion:An MRI protocol based on a single contrast-enhanced T1-weighted Dixon sequence allows reproducible RAMRIS scoring and measurement of the cartilage thickness. Further studies should be performed to evaluate the value of a short MRI protocol based on the Dixon method to monitor disease activity including cartilage loss in treated RA patients.