Comment on: Aringer M., Costenbader K., Daikh D., et al. “2019 European League against Rheumatism/American College of rheumatology classification criteria for systemic lupus erythematosus” Ann Rheum Dis 2019;78:1151–9. Papers 1-11.
Commented by: Matthias Schneider, Heinrich-Heine- Universität Düsseldorf, HHU, Policlinic and Hiller Research Unit Rheumatology. E-mail: Matthias.Schneider@med.uni-duesseldorf.de
The new EULAR/ACR are published and will change the perspective on lupus in many aspects (1,2). This new concept of preselection by positive ANA is challenging the ANA determination (3). Although the criteria are validated during the developing process with high values for both sensitivity and specificity, it is the duty of the scientific lupus community to analyse their value in various aspects of care and especially trials. The first are data already:
Chi et al. Raise in Annals of Rheumatic Diseases the question whether the 2019 EULAR/ACR SLE classification criteria close the door on certain groups of SLE patients (4). Especially they address the importance of rare lupus manifestations and malignancies and criticize that the clinical items are mostly unchanged. As someone, who accompanied several steps of the development, I can say that nearly every possible manifestation of SLE was on the table and discussed. But it also clear that these 2019 EULAR/ACR criteria are for classification and not for diagnosis. This means SLE can be diagnosed with just one rare manifestation, e.g. hepatitis, and malignancies, especially lymphoma, need to be excluded by the diagnosing physician.
Florence Assan and colleagues addressed the differential classification of SLE and primary Sjögren’s syndrome (pSS) (5), which was already focus of the classification process (6). They confirmed a sensitivity of 98% and a specificity of 96% for the EULAR/ACR criteria when directly tested against a pSS population. In addition, they could for the first time show a 85% sensitivity for the EULAR/ACR criteria in 26 patients with overlap.
Dahlström and Sjöwall tested the diagnostic accuracies of the new criteria in 56 patients with SLE and 55 controls (7). All controls were autoantibody positive, which somehow overcomes the first hurdle of the new classification criteria to be ANA positive. The result was a 93% sensitivity, but just a 74% specificity. The data were comparable to the SLICC criteria, but the 1982/1997 ACR criteria performed better with 83% specificity. These data point out the importance that only manifestations can be defined as valid if another reason is excluded.
During the validation of the new classification criteria the importance of ethnicity was raised. Therefore, Pons-Estel and colleagues compared the new EULAR/ACR to the 1982/1997 ACR criteria in two multiethnic lupus cohorts from a US cohort and from a Latin America cohort were included (8). Only 7.3% respectively 7.4% of patients achieved the 2018 EULAR/ACR criteria earlier in the US and Latin American cohorts; about one third of the patients in both cohorts achieved them later. Patients who accrued the 2018 EULAR/ACR earlier were more likely to have high anti-dsDNA titers and later earlier classification less likely to have mucocutaneous and joint manifestations. These data indicate that the disease expression is important for the new classification – a concept that was intended by the weighting of the classification criteria.
In the same direction are data from London (9). David Isenberg and colleagues analysed the predictive value of the new classification criteria and detected a statistically significant association between the EULAR/ACR set and renal damage, but not with CV events or death. This association persisted after the renal domain (all 3 criterion) was removed from the final score, but was neither detectable with the 1982/1997 ACR criteria nor with the SLICC criteria. After adjustment for age and sex higher EULAR/ACR scores were associated with higher SDI after 10 years. The effect was stronger in patients negative for rheumatoid factor and antibodies to RNP, Ro and La.
Dutch colleagues analyzed the performance of the 2019 ACR–EULAR classification criteria for SLE in a cohort of patients with SLE with neuropsychiatric symptoms (10). They investigated 294 (82%) with SLE and 66 patients without the clinical diagnosis of SLE and showed a sensitivity of 87% (95% CI 83% to 91%) and specificity of 74% (95% CI 62% to 84%), which is low comparable to validation process. Comparable results were found for the SLICC criteria, the 1997 ACR criteria had a sensitivity of 89% (95% CI 85% to 92%)and a higher specificity of 89% (95% CI 80% to 96%). These data may indicated some weakness for patients with neuropsychiatric symptoms due to limited manifestations as items or it may be driven by the selection of patients in this cohort due the use of 1997 ACR criteria in primary diagnosis. In any case these data need be challenged in other cohorts.
Anne Eberhard and colleagues validated the 2017 weighted criteria in comparison to the 1997 ACR and the 2012 SLICC in pediatric systemic lupus erythematosus (11). The sensitivity for the 2017 weighted criteria was 0.974 and specificity 0.984. The findings were comparable to the 2012 SLICC criteria and indicate that both sets may be applied also to pediatric SLE.
- Aringer M , Costenbader K, Daikh D , et al. 2019 European League against Rheumatism/American College of rheumatology classification criteria for systemic lupus erythematosus. Ann Rheum Dis 2019;78:1151–9.
- Aringer M, Costenbader K, Daikh D , et al. 2019 European League against Rheumatism/American College of rheumatology classification criteria for systemic lupus erythematosus. Arthritis Rheumatol 2019
- Pisetsky DS, Spencer DM, Rovin B, Lipsky PE. Role of ANA testing in the classification of patients with systemic lupus erythematosus. Ann Rheum Dis. 2019 Sep 20. pii: annrheumdis-2019-216259. doi: 10.1136/annrheumdis-2019-216259. [Epub ahead of print] PubMed PMID: 31540932.
- Chi H, Teng J, Wang Z, Yang C, Su Y. Do the 2019 EULAR/ACR SLE classification criteria close the door on certain groups of SLE patients? Ann Rheum Dis. 2019 Sep 16. pii: annrheumdis-2019-216253. doi: 10.1136/annrheumdis-2019-216253. [Epub ahead of print] PubMed PMID: 31527019.