P2-384: Serial MRI measurement of temporal lobe and global atrophy progression in semantic dementia

Abstract

Frontotemporal lobar degeneration (FTLD) is the third commonest cause of young–onset dementia. Semantic dementia, characterized by asymmetric left greater than right temporal lobe atrophy, is one of three cardinal FTLD subtypes. Measurements of atrophy may help in understanding disease progression and effects of therapy. To evaluate temporal lobe and brain atrophy in semantic dementia. 15 patients with semantic dementia and 9 age–matched controls each underwent two volumetric T1–weighted MRI scans (mean interval 20 months). Volume changes in brain, temporal lobes and ventricles were calculated from manually outlined regions on registered scans and using the boundary shift integral [BSI]). Sample sizes required to power 1 year placebo–controlled treatment trials were estimated. As a percentage of brain volume, the semantic dementia group had reduced temporal lobe volumes: baseline mean (SD) left temporal lobe (LTL) 2.87 (0.64)% vs 5.11 (0.18)% in controls; right temporal lobe (RTL) 4.48 (0.67)% vs 5.34 (0.17)%. Mean (SD) atrophy rate was 3.03 (1.83) mL/yr on the left and 3.98 (2.39) mL/yr on the right while in the control group it was 0.45 (0.38) mL/yr on the left, 0.40 (0.74) mL/yr on the right. The semantic dementia group had increased rates of brain atrophy, 2.10 (1.13) %/yr vs 0.57 (0.28) %/yr, and ventricular expansion (as a percentage of brain) 0.68% (0.27%) vs 0.14% (0.10%). For 90% power to detect a 20% reduction in atrophy rate (controlling for control atrophy rate), fewer patients per arm are required using ventricular volumes (133) compared to RTL (235), LTL (265) or brain atrophy (287 using BSI). The semantic dementia patients had significantly more left temporal lobe atrophy (∼55% of control mean) than right (∼85% of control mean). However, rates of temporal atrophy were greater on the right than left (p<0.05) suggesting acceleration of RTL atrophy after an initial phase of faster LTL atrophy. For trials of potentially disease–modifying therapies either temporal lobe has marginally greater power than whole brain in terms of tracking progression. Ventricular measures may offer reduced sample sizes but this must be balanced against the possibility of ventricular change being affected by factors other than neurodegeneration.