Individual assessment of chronic brain tissue changes in MRI – the role of focal lesions for brain atrophy development. A Voxel-Guided Morphometry study

Abstract

Despite considerable progress in the treatment of MS, it is still the most frequent cause of neurological disability in young and middle-aged adults. The known neuropathological mechanisms leading to progressive disability are immune-mediated demyelination and axonal loss. Several MRI studies demonstrated brain volume reduction, consisting of white and grey matter loss as an in vivo marker of MS. As this brain atrophy is positively correlated with disability, atrophy measures have been adopted as secondary outcomes in a number of trials of disease-modifying drugs. As neurodegeneration is irreversible, brain atrophy measurements may be viewed as an important biomarker of MS brain pathology. In our study we used a new method, Voxel-Guided Morphometry (VGM) for detailed individual and focal assessment of brain structural changes in 100 MS patients. The method combines high accurracy resulting in the detection of structural changes <1mm with high reproducibility due to a fully automated analysis procedure and the possibility of individual morphometrical assessments. Corresponding 3D MRI scans of each individual were segmented for removement of non-brain structures and aligned to each other. Using a four-step procedure, exact comparisons of local brain volume were performed (IEEE TMI, 22: 62–74, 2003). In short, the 4 steps are: 1. coarse linear alignment by use of the extended principle axes theory (ePAT) generalized to affine movements; 2. a cross-correlation-based technique using a matrix-norm for fine linear alignment; 3. application of a high-dimensional multiresolution full multigrid method for determination of nonlinear deformations, thereby achieving complete exploitation of information and effective processing; and 4. determination of volume alterations between two brains imaged 12 months apart and visualization using color scale encoding. To assess brain volume changes, 2 time points 12 months apart were analysed for each patient. Focal as well as widespread brain volume changes were detected in the consecutive MRI scans of the 100 individuals. Preliminary results of the analysis of patients that developed a new lesion during the 1 year interscan period demonstrate, that new lesions are detected with high sensitivity by VGM. It was also shown, that large parts of the brain beyond the lesion site appear morphologically stable. CSF structures close to the lesion were enlarged in size, corresponding to focal brain awas affected and showed a slight otherwise invisible volume decrease. These changes were related to the location of the lesion but extended up to adjacent white matter and cortical areas, demonstrating the remote effects of the focal lesion on brain morpho-logy. In summary, VGM demonstrated focal brain structure alterations after the development of new lesions in MS patients. These structural changes included detection of the lesion itself with high sensitivity as well as morphological alterations exceeding the lesion itself that could not be detected otherwise. Thus VGM provides some information on the mechanisms of brain atrophy development and may be a suitable method for individual high accuracy detection of focal brain volume changes in MS patients over time.