[Neuropsychological tests, MRI and SPECT in the evaluation of CNS involvement in SLE]

To assess the agreement between presence and location of central nervous system (CNS) structural damage and neuropsychological performance.

21 unselected SLE patients underwent a 3 hours-long battery of neuropsychological tests sampling 15 cognitive functions. A neuropsychologist hypothesized for each SLE patient the most likely site of possible involvement, according to the neuropsychological performance. Patients underwent MRI scans within 6 months (3 months before or after) from neuropsychological tests and SPECT.

14 of the 21 SLE patients (66.6%) were impaired in at least 1 function; among these, 7 patients (33.3%) were mildly impaired and 7 (33.3%) had more than 3 functions impaired. 10 patients (48%) had abnormal MRI scan. Negative and positive agreements between neuropsychological performance and MRI scan were detected in 15 patients (71%; ?2 with Yates' correction 6.09, p 0.007, measure K for concordance 0.44, p 0.03). All the 6 negative agreements had no records of major neurologic or psychiatric events; among the 9 positive agreements, 6 presented previous major neurologic events and 2 minor neuropsychiatric symptoms. Among the subjects with disagreement, 1 was unimpaired with positive MRI and without neuropsychiatric symptoms, 5 were mildly impaired with negative MRI.

A detailed neuropsychological evaluation along with MRI arose as a valid method to exclude or to identify, localize and clinically interpret CNS involvement in SLE. Disagreement between MRI and neuropsychological tests was detected mainly for mild cognitive impairment that might be the expression of functional (inflammatory or ischemic) processes.