Purpose: To determine brain involvement in Lafora disease by means of 3-T MRI volumetry and 1H magnetic resonance (MR) spectroscopy.
Methods: Ten patients with Lafora disease and 10 healthy controls were included in the study. The diagnosis of Lafora disease was proven genetically by the presence of mutations in the EPM2A gene in all patients, and their evolution was staged in three groups according to their functional state. MRI volumetry was performed by means of AX3DT1 images with assessment of the cerebellum and the brainstem, by using the program Stereonauta, and all the brain structures, by using voxel-based morphometry. [1H]MR spectroscopy was performed by using an Eclipse PRESS sequence probe system with 8-cc voxels positioned in the occipital and frontal cortexes, basal ganglia, pons, and cerebellar hemispheres. Spectral peak areas corresponding to NAA (N-acetylaspartate), creatine, and choline were obtained.
Results: MRI volumetry showed no statistically significant differences in patients compared with healthy controls in any of the analyzed structures. Analysis of [1H]MR spectroscopy data showed a statistically significant reduction in the NAA/creatine ratio in patients compared with controls in the frontal (p = 0.001) and occipital cortex (p = 0.043), basal ganglia (p = 0.002), and cerebellar hemispheres (p = 0.007). The NAA/choline and choline/creatine ratios were statistically significantly different in the frontal cortex (p = 0.005). No correlation was observed between the disease-evolution stage and MRI-measured volumes (range, -0.92 to 0.44) or [1H]MR spectroscopy values (range, -0.29 to 0.50).
Conclusions: In our series of Lafora disease patients, [1H]MR spectroscopy was more sensitive than structural MRI to detect brain involvement. The brain cortex, especially frontal cortex, cerebellum, and basal ganglia, showed the greatest metabolic changes.