The Diagnostic Value Of Diffusion Tensor MRI Metrics In Relation To The MND Phenotype Heterogeneity (P7.007)

OBJECTIVE: To investigate diffusion tensor (DT) MRI metrics as predictors of motor neuron disease (MND) diagnosis. BACKGROUND: Reliable markers for upper motor neuron (UMN) and extra-motor white matter (WM) damage may facilitate MND diagnosis. DESIGN/METHODS: Corticospinal tract (CST), corpus callosum (CC) and extra-motor tract DT MRI measures were obtained from 123 MND patients and 35 healthy subjects. DT MRI metric C-indices were estimated using logistic regression analyses. Support vector machine (SVM) classification algorithm assessed DT MRI metric accuracy as predictors of MND diagnosis at individual patient level. RESULTS: MND patients showed damage to CST and motor callosal fibers (C-index range: 0.65-0.74). No DT MRI abnormalities were found in pure lower motor neuron (LMN) and flail arm patients. The most severe and widespread damage was found in pure UMN/pyramidal ALS (C-index up to 0.81 for CST and 0.91 for CC). Classical, respiratory and bulbar ALS showed damage to CST and callosal motor fibers (C-index≍0.70). In bulbar patients, uncinate and cingulum damage was also found. In classical, respiratory and bulbar ALS, patterns of damage were confirmed in cases with disease duration <12 months (47% of sample). Between patient group comparisons showed that severity and extent of WM damage increased in the following order: pure UMN/pyramidal ALS>bulbar>classical/respiratory>pure LMN/flail arm. The highest C-index (up to 0.94) was found for CC DT MRI measures distinguishing pure UMN/pyramidal from pure LMN patients. SVM showed the highest diagnostic accuracy (0.93) in the comparison pure UMN/pyramidal vs. pure LMN. Disease severity and UMN involvement correlated with damage to CST, entire CC and motor callosal fibers. CONCLUSIONS: DT MRI provides sensitive objective measures of UMN and extra-motor burden at the individual level in MND patients, including those at the early phase of the disease. This study provides a roadmap for translation of MRI predictors of MND into daily practice. Study Supported by: Italian Ministry of Health (#RF-2010-2313220).