Comparing Sensitivities of Childhood-Onset SLE Classification: SLICC vs ACR Criteria

For childhood-onset systemic lupus erythematosus (cSLE), the Systemic Lupus International Collaborating Clinics (SLICC) classification criteria were significantly more sensitive than the American College of Rheumatology (ACR) criteria, particularly shortly after diagnosis, according to findings published in The Journal of Rheumatology.

Although ACR and SLICC criteria are well validated for diagnosis and classification of adult SLE, there has been relatively little research into their use in cSLE. Investigators sought to compare the 2 accepted systems regarding their clinical utility in cSLE using a large and diverse cohort of children.

A retrospective analysis was performed on prospectively collected data from 722 pediatric patients (mean age at diagnosis, 12.8 years; 81.7% women) diagnosed with cSLE between 1984 and 2017. Scoring for all criteria was derived from cumulative symptoms through the most recent clinic visit.

Quantitative analysis compared sensitivities of the 2 systems using McNemar test, while qualitative analysis employed descriptive statistics to identify unique ACR- or SLICC-specific disease features as well as any autoantibodies excluded from both sets of criteria. Patients were divided into 4 groups according to whether they fulfilled one, both, or neither sets of criteria: 89.6% of patients met both sets of criteria, 6.6% met only SLICC criteria, 2.8% met only ACR criteria, and 1% had an expert clinical diagnosis only.

With the requirement of meeting ≥4 criteria, the sensitivity of SLICC was significantly higher than that of ACR overall and during all early periods following diagnosis. Overall sensitivities were 96.3% for SLICC and 92.4% for ACR (P =.001), with sensitivities of 91.3% and 82.5% for patients scored ≤1 year after diagnosis (P =.01), and 97.9% and 92.7% (P =.02) for those scored 2 to 3 years following diagnosis. An upward trend continued until both systems reached 100% when scored >10 years from diagnosis. Concordance between the 2 systems was excellent at 89.6%. Stratified by sex, the respective sensitivities were 96.1% and 92.9% for women, and 97.0% and 90.2% for men (P =.02 for both).

Among the 55 individuals who did not fulfill ACR criteria, 48 (87.3%) met SLICC criteria. Of the 27 participants who did not meet SLICC criteria, 20 (74.1%) fulfilled ACR criteria through satisfaction of either photosensitivity (73.9%) or lymphopenia (26.1%). Inclusion of both photosensitivity and lymphopenia in SLICC criteria would increase SLICC sensitivity to 99.0% (P = .0001). Additionally, of the 7 patients who did not meet either criteria but who had a clinical diagnosis, 6 (85.7%) were positive for anti-Ro antibodies. This suggested that despite its usual exclusivity in primary Sjogren syndrome, this autoantibody might be useful for cSLE classification, given the rarity of pediatric primary Sjogren syndrome.

Study limitations included a lack of specificity analysis as a consequence of not having a control group, the location at a single center, and reliance on consensus diagnosis by 3 experts instead of a gold standard, which does not exist.

“The use of SLICC criteria may improve clinical care and allow us to study patients earlier in their disease course,” noted the authors, as they looked ahead to anticipated criteria updates in terms of their potential effect on cSLE diagnosis and classification.

Reference

Tao JJ, Hiraki LT, Levy DM, Silverman ED. Comparison of sensitivities of American College of Rheumatology and Systemic Lupus International Collaborating Clinics classification criteria in childhood-onset systemic lupus erythematosus [published online February 15, 2019]. J Rheumatol. doi:10.3899/jrheum.180337