Background/Purpose: Neuropsychiatric manifestations are frequently seen in patients with childhood-onset erythematosus (cSLE) and may include neurocognitive dysfunction. These symptoms are difficult to diagnose and classify due to their heterogeneity, especially in the early stages. In this sense, the magnetic resonance imaging (MRI) associated with cognitive dysfunction screening batteries, such as the pediatric neuropsychological assessment metrics (PedAnam), compose an investigation protocol. The objective was to explore and test approaches to calculate the Cognitive Performance Score (CSP) of the PedAnam battery based on Principal Component Analysis (PCA) and to correlate the results with the brain volume numbers obtained by MRI.

Methods: Consecutive cSLE patients followed up at the Rheumatology Unit were included in this study. They responded to PedAnam and performed MRI (Siemens 3T scanner) on the same day. Basic characteristics about the disease were collected. Sagittal MR images were T1-weighted and segmented using Freesurfer software. The PedAnam results were analyzed using PCA as the cognitive performance score. The variance-covariance matrix of the normalized precision scores of the PedANAM subtests was decomposed into a series of eigenvectors with corresponding eigenvalues. The first eigenvector was used to preserve most of the total variance of the matrix. Furthermore, the score automatically generated by Ped-ANAM is notable. The data from the PCA and effort process were correlated with volumetric results of brain volume and hippocampus.

Results: We included 54 cSLE patients [mean age 12 SD= 3;6 (86.19%) women]; age at disease onset ranged from 5 to 17 years [mean 12.85 SD=3.44]; disease duration ranged from 6-30 years [mean 14 SD=5.82]; The predominant race was white, with 41 people (75.93%), followed by Afro-descendants (24.08%). We analyzed each accuracy separately in terms of measures of centrality and dispersion. After calculating the PCA, we obtained the median between 1.13 and 3.96, which shows that for our cohort, we cannot use the cut-off point of the scientific literature (0.25). While effort, a standard metric automatically provided by the software, presented an average of 7 [ 0 – 19], the score is interpreted as ranging from 0 – 33; the lower, the better the performance, and results greater than 15 may present cognitive deficits. From the proposed cut-off point, 6 patients had cognitive deficits. As for the association with MRI, we observed that the White Brain Substance Volume correlated with CPS (r=0.41, p=0.023). While the effort did not show any statistically significant association.

Conclusion: We observed that the PCA analyzed in our cohort presented higher indices of variance distribution. Additionally, we observed that this simplified cognitive performance score was associated with white matter volume.

« Back to ACR Convergence 2022

ACR Meeting Abstracts - https://acrabstracts.org/abstract/screening-for-neurocognitive-demands-in-childhood-onset-systemic-lupus-erythematosus-patients-using-pediatric-automated-neuropsychological-assessment-metrics-and-association-with-magnetic-resonance-im/