Background/Purpose: Neuropsychiatric disease is common in childhood-onset Systemic Lupus Erythematosus (cSLE). Signs of cNPSLE can be subtle and difficult to ascertain in daily clinical practice.  Formal neuropsychological testing can be used to assess cognitive ability in SLE but can be expensive, time-consuming, and not always readily available. Automated neurocognitive computerized tests, such as the Pediatric Automated Neuropsychological Assessment Metrics (PedANAM), are convenient, require less training to administer, and cost-effective.  PedANAM has shown to have good reproducibility and criterion and concurrent validity in cSLE. The purpose of this study is to determine 1) if the PedANAM is feasible for monitoring cognitive status in daily clinical care and 2) its relationship with disease activity indicators. 

Methods: 10 subsets of the PedANAM were administered to cSLE patients (pts) recruited from 7 centers.  Demographic and clinically relevant information was collected.  Mean reaction time for correct responses (MNc, in msec) and accuracy (AC, % of correct responses) were measured for each subtest.  

Results:  Preliminary analysis of 113 of 200 expected pts was performed (mean age 15 years; White 29%, Black 32%, Asian 20%, Hispanic 14%).  Most pts were in high school, 10 repeated a grade, and 19 received special services.  Very low AC (<60%) scores were seen in 48 pts on at least one subtest. Only 7 pts showed low AC on 3 or more subtests. Potential reasons for the low AC scores observed were reversed key responses (1 pt) and very fast response time combined with low accuracy.  The latter can be explained by either lack of effort or poor understanding of the task.  Pts had the greatest difficulty (e.g. low AC) on subtests evaluating delayed and working memory (continuous performance (CPT), code substitution delayed (CDD), and matching to sample (M2S). Hispanic pts had the lowest AC especially on CPT (p=.014), M2S (p=.019), and subtest assessing associative learning (code substitution CDS, p=.019), suggesting greater illness burden. Pts with a history of a greater number of cNPSLE symptoms performed more slowly and less accurately, especially on the spatial processing (SPD) MNc and CPT AC (p's<.05).  Pts with active disease measured by the SLEDAI (scores >4) were consistently slower and less accurate on all the subtests with significant differences on CDS, SPD, and simple reaction time (Fig 1).  Presence of a SLEDAI-DNA binding was the only SLEDAI parameter significantly associated with PedANAM performance, especially with SPD (p=.041).

Conclusion: The PedANAM appears to be a feasible tool for the assessment of cognitive status in cSLE.  Elevated disease activity scores were significantly associated with decreased PedANAM performance. The meaning of atypical low AC scores and cSLE activity needs to be investigated.