Background Systemic Lupus Erythematosus (SLE) often poses challenges to diagnosis in clinical practice and classification in research. The 2012 SLICC criteria have been recently introduced, with validation studies suggesting greater sensitivity yet equal or lower specificity compared to the 1997 ACR criteria. However, the prognostic significance of classifying SLE patients with the SLICC2012 versus the ACR1997 criteria is not known.
Objectives To evaluate the impact of SLICC2012 versus ACR1997 SLE classification in terms of clinical characteristics and outcome.
Methods Both the SLICC2012 and the ACR1997 classification criteria were applied to adult SLE patients enrolled in the community-based Cretan Lupus Registry over the period 1999–2013. Classified cases were assessed at the end of 2013 in terms of disease severity (determined by the severity of manifestations and the use of lupus treatments) and organ damage (assessed by the SLICC/ACR damage index). Cases who fulfilled both criteria during the observation period were categorized according to which set of criteria was satisfied first and then compared for the frequency of individual criteria.
Results At the end of the observation period (year 2013), fewer SLE patients had been classified with the SLICC2012 (n=602) as compared with the ACR1997 (n=750) criteria. The female-to-male ratio (12.6:1 versus 13:1) and the mean (±SD) age at the time of diagnosis (42±15 versus 43±15 years) were comparable between SLICC2012 and ACR1997-classified cases, respectively. Lupus was classified as mild, moderate and severe in 50%, 33% and 17% of the ACR1997- as compared to 42%, 34% and 23% of the SLICC2012-classified patients (p<0.001). Damage occurred in 30.5% of the ACR1997 versus 36% of the SLICC2012 cohort (p=0.01) despite comparable disease duration. Patients who fulfilled both sets of criteria (n=512) were categorized into three groups based on which criteria were fulfilled first, i.e. group 0 (concurrently: 87% of cases), group 1 (SLICC2012 after ACR1997: 5%), and group 2 (SLICC2012 before ACR1997: 8%). We found that malar rash and neurologic disorder were significantly more prevalent in group 1 (malar rash: 54.3%, 80.8% and 48.8% in groups 0, 1 and 2, respectively, p=0.02; neurologic disorder: 3.7%, 15.4% and 0%, respectively, p=0.05). Conversely, group 1 patients had significantly lower frequency of non-scarring alopecia (51.3%, 23.1% and 68.3%, respectively, p=0.02) and synovitis (23.9%, 0% and 29.3%, respectively, p=0.012).
Conclusions Application of the SLICC2012 criteria may result in classification of SLE patients with more severe disease compared to those who fulfill the ACR1997 criteria, which may have important implications in terms of trial design; however, these data need to be validated. Lack of inclusion of malar rash in the SLICC2012 criteria and of non-scarring alopecia in the ACR1997 criteria may delay the classification of SLE patients.
Disclosure of Interest None declared