MRT-basierte Morphometrie zur Darstellung der Gehirnatrophie in frühen Stadien des Morbus Huntington - Vergleich einer manuellen und einer automatisierten Methode

Abstract

Objectives: To find a non-invasive method to analyse the grey matter changes in early stages of Huntington"s disease (HD) and to determine statistically how these changes correlate to clinical features and the CAG (cytosine-adenosine-guanosine) -repeat length. Methods: High resolution volume rendering MRI Scans (MP-RAGE) of 44 patients in early stages of HD and 28 healthy volunteers were acquired. The imaging software Analyze® was used to determine manually the absolute volume of nucleus caudatus and putamen in 12 HD patients compared to 6 healthy controls. Without predefined cerebral regions the data of 44 patients was analysed by the automatic voxel-wise comparison technique of voxel based morphometry (VBM) and compared to an age-matched database, using statistical parametric mapping (SPM99). To visualise the influence of clinical severity on the degree of the cerebral atrophy an additional VBM and covariance analysis (ANCOVA) was done. Results: With Analyze® a decreased volume of the striatum in affected patients and an atrophy progress with more severe clinical impairment was demonstrated. With VBM a significant decrease in grey matter density was found in HD patients compared to the control population. The changes were most consistent bilaterally in the striatal regions. A covariance analysis showed that scores of clinical severity correlate strongly with the atrophy. Higher CAG-repeat numbers and advanced clinical stages showed more extensive striatal involvement. Conclusions: Without preselection of specific cerebral regions, VBM is able to show even small changes in grey matter density associated with HD and its covariation with clinical severity and CAG repeat length. Analyze® is a complementary method to obtain absolute volume but is time consuming and investigator dependent. VBM may have the potential to monitor the progressive changes of HD in future longitudinal intervention studies or preclinical investigations.