Background MRI is superior to clinical examination in the detection of joint inflammation. According to an EULAR imaging taskforce the presence of inflammation seen with MRI can be used in the diagnostic process of Rheumatoid Arthritis (RA). These recommendations are largely based on the sensitivity of MRI to detect inflammation. The specificity of MRI-detected inflammation is not known as the prevalence of MRI-features in symptom-free persons living in the general population is unknown. This information is required when determining the diagnostic value of MRI-detected inflammation. We assessed the prevalence of MRI-detected inflammation (synovitis, bone marrow edema and tenosynovitis) in symptom-free persons and aimed to generate reference values.
Methods From November 2013 to December 2014 symptom-free persons were recruited via newspapers and websites. All persons were physically examined prior to inclusion. Inclusion criteria were: no history of inflammatory arthritis, no joint symptoms during the last month and no clinically detectable arthritis. A contrast enhanced 1.5T MRI of the dominant MCP, wrist and MTP joints was made and scored semi-quantitatively according to RAMRIS. Two readers scored the blinded MRIs that were mixed with scans of RA-patients to exclude observer bias. Prevalences of features were assessed.
Results 196 persons were screened; 193 persons fulfilled the criteria. The age ranged from 19 to 89 years (mean 49.8). No single inflammatory feature was seen in 28% of persons. Any BME (score≥1) was depicted in 58% of persons, any synovitis (score≥1) in 48% and tenosynovitis (score≥1) in 17%. Total RAMRIS inflammation scores were not different between men and women (p=0.36) but correlated with age (r=0.57 p<0.001). Correlations with age were also found for the separate inflammation features (BME r=0.51; synovitis r=0.55 and tenosynovitis r=0.28; all p<0.001). When evaluating the different inflammation features at different locations and for different age-categories the most prevalent observations were the following: in persons aged ≥40 synovitis (score 1) was present in 8-35% of the radio-ulnar, radio-carpal and intercapal-joint and in 8-19% of MCP-2 and MCP-3 joints but absent in those aged <40. Persons aged ≥60 had synovitis (score 1) of MCP4, MCP 5 and MTP-1 in 4%, 6% and 13% respectively. BME (score 1) is present in the lunate in 18%; this occurrence was not related to age. Tenosynovitis (score 1) was very rare at the MCPs and wrists, except for extensor carpi ulnaris (frequency of 0%, 9% and 12% in persons aged age <40, 40-59 and ≥60years). Synovitis and tenosynovitis with a score of ≥2 was very rare, the same was observed for BME, except for MTP-1 (8% in persons aged ≥60). Synovitis (score 1) preferentially occurred at locations that also showed BME or tenosynovitis. Based on these data normograms with reference values were constructed.
Conclusions MRI-detected inflammation in symptom-free persons is prevalent. Observed findings had low scores. The prevalence depended on the type of inflammation, the anatomic location and the age. RAMRIS-based and age, location and MRI-feature dependent reference values for MRI-detected inflammation were identified.
Colebatch AN, et al. Ann Rheum Dis 2013:804–14.
Disclosure of Interest None declared