Radiographic erosions are a clear hallmark of rheumatoid arthritis (RA). The EULAR definition of radiographic erosive disease has a high specificity and its fulfilment alone is sufficient to classify RA.[1] In recent years, novel imaging techniques, such as magnetic resonance imaging (MRI), have emerged and are recommended by a EULAR imaging taskforce because MRI is more sensitive to detect erosions than radiography.[2] To determine the specificity of MRI-detected erosions, we recently compared hand and foot MRI [3]of patients presenting with RA with those of symptom-free persons and patients presenting with arthritides other than RA.[4] MRI-detected erosions were present in all groups. By evaluating different erosion features a few features were identified as specific for RA; these were severe erosions (grade≥2, defined as >10% of bone eroded), erosions in MTP5, and erosions in MTP1 in persons aged <40. A subsequent and clinically relevant question is whether MRI-detected erosions in patients presenting with undifferentiated arthritis (UA) are valuable in predicting future progression to RA. This was explored to a limited extend in our previous study but as the number of UA patients was limited (n=192), the predictive value of the different ‘RA specific erosions’ could not be studied. In addition, the outcome was fulfilment of classification criteria but start of disease modifying antirheumatic drugs (DMARDs) was not considered while DMARD-treatment might have hampered progression to fulfilment of RA classification criteria. Since the previous study mainly performed cross-sectional comparisons, the predictive accuracy of MRI-detected erosions in UA was not evaluated satisfactorily. Therefore we continued with a larger longitudinal study.
286 patients with UA (using the 2010-criteria to classify RA) consecutively included in the Leiden EAC between 2010 and 2016 were studied. 94% of the 2010-UA patients was ACPA-negative (Supplementary table), which is in line with other descriptions of the population of 2010-UA patients.[5,6] At baseline 1.5T MRI of the 2nd-5th MCP- and 1st-5th MTP-joints was performed as described.[4] Erosions were scored on a scale 0-10 according to the RAMRIS system.[6] First the presence of any MRI-detected erosions, defined as score ≥1, was evaluated. Then the presence of RA-specific erosions (either a grade ≥2 erosion, erosion in MTP5 and/or erosion in MTP1 in persons aged<40) was studied.[4] Patients were followed for 1-year on RA development, defined as fulfilling the 2010-criteria, or the start of DMARDs because of a clinical diagnosis of RA. The latter was added as ACPA-negative patients need >10 involved joints to fulfil the 2010-criteria which could be hampered by DMARD-treatment. 128 (45%) UA-patients developed the outcome of which 111 had a clinical diagnosis of RA and started DMARDs and 17 fulfilled the 2010-criteria.
Any MRI-detected erosion was present in 53% of the 286 UA patients and was not predictive for RA development (OR 1.2, 95%CI 0.7-1.9, PPV 47%, Table). RA-specific erosions were present in only 7% of the 2010-UA patients and were also not associated with development of RA (OR 0.6, 95%CI 0.2-1.5, PPV 33%). Similar findings were obtained for the individual ‘RA-specific erosions’ (Table). Sensitivity analyses stratified for the outcome (DMARD-start or only 2010-criteria positive) revealed similar results (data not shown).
Table. Predictive values of MRI-detected erosions within 2010-UA patients for the development of RA.
| UA patients with erosion feature n (%) | PPV (95% CI) | NPV (95% CI) | OR (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) | |
|---|---|---|---|---|---|---|
| Any MRI-detected erosion | 152 (53) | 47% (39-55) | 57% (49-66) | 1.2 (0.7-1.9) | 55% (47-64) | 49% (41-56) |
| Any ‘RA-specific erosion’ | 21 (7) | 33% (17-55) | 54% (48-60) | 0.6 (0.2-1.5) | 5% (3-11) | 91% (86-95) |
| Grade ≥2erosion | 5 (2) | 40% (12-77) | 55% (49-61) | 0.8 (0.1-5.0) | 2% (0-6) | 98% (95-99) |
| Erosion in MTP5 | 16 (6) | 44% (23-67) | 55% (49-61) | 1.0 (0.3-2.6) | 5% (3-11) | 94% (90-97) |
| Erosion in MTP1 and aged<40 | 2 (1) | 0% (0-66) | 55% (49-61) | undefined | 0% (0-3) | 99% (96-100) |
The prior risk for development of RA and/or DMARD-use within 1 year was 45%. Any MRI-detected erosion was defined as score ≥1 according to the Rheumatoid Arthritis MRI Scoring System. Any RA-specific erosion was defined on MRI as the presence of a grade ≥2 erosion, an erosion in MTP5 and/or an erosion in MTP1 in the age group <40 years, as described earlier.[4]. RA, rheumatoid arthritis; MRI, magnetic resonance imaging; DMARD, disease-modifying anti-rheumatic drug; PPV, positive predictive value; NPV, negative predictive value; OR, odds ratio; 95% CI, 95% confidence interval.
This is the largest longitudinal dataset on MRI-detected erosions in hand and foot joints in UA to date. In all analyses MRI-detected erosions were not associated with an increased risk on RA. Although MRI is sensitive, the present data suggest that evaluation of MRI-detected erosions in UA is not relevant for the early detection of RA.
Supplementary Material
Key Message.
-
-
MRI-detected erosions have no value in differentiating undifferentiated arthritis patients who will develop RA from those who will not.
Acknowledgment
Not applicable.
Funding
This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Starting grant, agreement No 714312).
Footnotes
Competing Interests
The authors have no conflicts of interests.
Ethical Approval Information
The study on early arthritis patients was approved by the local medical ethics committee, which is named ‘Commissie Medische Ethiek’. All patients signed informed consent.
Data Sharing Statement
Data can be requested from the corresponding author.
References
- 1.van der Heijde D, van der Helm-van Mil AHM, Aletaha D, et al. EULAR definition of erosive disease in light of the 2010 ACR/EULAR rheumatoid arthritis classification criteria. Ann Rheum Dis. 2013;72:479–81. doi: 10.1136/annrheumdis-2012-202779. [DOI] [PubMed] [Google Scholar]
- 2.Colebatch AN, Edwards CJ, Østergaard M, et al. EULAR recommendations for the use of imaging of the joints in the clinical management of rheumatoid arthritis. Ann Rheum Dis. 2013;72:804–14. doi: 10.1136/annrheumdis-2012-203158. [DOI] [PubMed] [Google Scholar]
- 3.Østergaard M, Peterfy C, Conaghan P, et al. OMERACT Rheumatoid Arthritis Magnetic Resonance Imaging Studies. Core set of MRI acquisitions, joint pathology definitions, and the OMERACT RA-MRI scoring system. J Rheumatol. 2003;30:1385–6. [PubMed] [Google Scholar]
- 4.Boeters DM, Nieuwenhuis WP, van Steenbergen HW, et al. Are MRI-detected erosions specific for RA? A large explorative cross-sectional study. Ann Rheum Dis. 2018;77:861–8. doi: 10.1136/annrheumdis-2017-212252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Mil AHM van der H, Zink A. What is rheumatoid arthritis? Considering consequences of changed classification criteria. Ann Rheum Dis. 2017;76:315–7. doi: 10.1136/annrheumdis-2016-209629. [DOI] [PubMed] [Google Scholar]
- 6.Krabben A, Abhishek A, Britsemmer K, et al. Risk of rheumatoid arthritis development in patients with unclassified arthritis according to the 2010 ACR/EULAR criteria for rheumatoid arthritis. Rheumatology. 2013;52:1265–70. doi: 10.1093/rheumatology/ket011. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.