Abstract #1553

Reddick W, Dirksen C, Glass J, Pui C, Cheng C

This project focuses on detecting leukoencephalopathy early during therapy for ALL in 70 subjects and its association with neurocognitive performance 2.5 years later. Conventional MR imaging was performed (T1, T2, PD, and FLAIR) and analyzed with an automated computer-aided detection algorithm for leukoencephalopathy. Neurocognitive assessments were acquired, stratified by MR examination, and then assessed between groups using a two-way ANOVA. All intelligence and some verbal learning measures were significantly lower in patients with leukoencephalopathy. None of the academic achievement measures were significantly different between patient groups or normative test averages. This study establishes a significant relationship between MR changes seen early in therapy and later neurocognitive impact in children treated for ALL.