Amyotrophic lateral sclerosis: diffusion-tensor and chemical shift MR imaging at 3.0 T

To prospectively determine whether diffusion-tensor magnetic resonance (MR) imaging in conjunction with two-dimensional chemical shift imaging can assist in identifying upper motor neuron involvement and whether disease severity and duration can be predicted based on imaging parameters in patients with amyotrophic lateral sclerosis (ALS).

Institutional review board approval and informed consent were obtained for this HIPAA-compliant study. Fifteen patients with ALS (12 men, three women; mean age, 57.3 years) with clinical evidence of upper motor neuron involvement and 10 healthy control subjects (five men and five women; mean age, 49.4 years) were studied. Fractional anisotropy (FA) and apparent diffusion coefficient (ADC) were measured from the corticospinal tracts at the level of the internal capsule. Average N-acetylaspartate (NAA)/creatine-phosphocreatine (Cr) and NAA/choline-containing compounds (Cho) ratios were calculated from the precentral gyrus. Student t test, multiple linear regression analysis, and Spearman correlation coefficients were employed to quantify relationships between imaging and clinical parameters.

Patients with ALS exhibited significantly reduced FA values and NAA/Cr and NAA/Cho ratios compared with values in control subjects (P<.05) for both affected and nonaffected sides of the brain. ADC was elevated significantly in the affected side (P<.05) and was an independent predictor of disease duration after adjusting for age; however, FA values and NAA/Cr ratios for the affected side were even stronger predictors of disease duration. Moderate but statistically significant correlation was found between the FA values for the affected side and the ALS Functional Rating Scale Revised (ALSFRS-R) score (r=0.51, P<.05). The NAA/Cr ratio also correlated with both the ALSFRS-R and upper motor neuron scores (r=0.50 and 0.54, respectively; P<.05).

Diffusion-tensor and two-dimensional chemical shift MR imaging spectroscopy can be used to identify upper motor neuron involvement and predict disease duration in patients with ALS.