Premium
[P4–517]: APOE ε4 ALLELE EFFECT ON WHITE AND GRAY MATTER PERFUSION IN COGNITIVELY NORMAL AND MCI GROUPS
Author(s) -
Jung Youngkyoo,
Johnston Megan E.,
Kim Jeongchul,
Whitlow Christopher T.,
Hughes Timothy M.,
Baker Laura D.,
Craft Suzanne
Publication year - 2017
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2017.07.679
Subject(s) - white matter , cerebral blood flow , voxel , psychology , perfusion , cardiology , magnetic resonance imaging , medicine , nuclear medicine , neuroscience , radiology
only be established in longitudinal follow-up studies. In the present study, we investigate 6 year follow-up changes in white matter (WM) integrity and grey matter (GM) volume in a large cohort of presymptomatic FTD due to MAPT and GRN mutations, and examine whether these measures can serve as prediction biomarkers for symptom onset. Methods:Since 2010, we track healthy 50% at-risk subjects within our prospective FTD Risk Cohort. Within this study window, 8 mutation carriers converted to clinical FTD (6 bvFTD; 5 MAPT, 1 GRN; 2 non-fluent PPA (nf-PPA); 2 GRN). We compared these converters with non-converters (n1⁄435) and healthy controls from the same families (n1⁄430) by means of 3T diffusion tensor and GM volumetric MRI at 4 and 2 years prior to symptom onset. We assessed the prognostic performance of WM and GM regions-of-interest (ROI) by means of ROC analyses in converters. Results:Two years in advance, symptom onset could be accurately predicted by means of theWM integrity values (FA) of the bilateral uncinate fasciculus (UF), genu of corpus callosum, fornix, forceps minor and right inferior longitudinal fasciculus (AUC 0.74-0.78); and GM volumes of the bilateral frontal and temporal lobes, right anterior temporal lobe, parietal lobe, insula, and (anterior) cingulate (AUC 0.74-0.83). Additionally, the left UF had absolute classification accuracy for the clinical phenotypes bvFTD and nf-PPA (AUC 1.00; sensitivity/specificity 100%; optimal FA cut-off 0.47) (Figure 1). Four years before, WM integrity and GM volumes could not predict symptom onset. At none of the time points we could separateMAPT fromGRN converters. Conclusions: We demonstrate that two years in advance, symptom onset can be accurately predicted byWM and GM neuroimaging measures within regions known to be affected in early stage to symptomatic FTD. Our findings point to the UF as the key pathological frontotemporal tract when moving from the presymptomatic into the symptomatic stage. These neuroimaging biomarkers can be of crucial importance for e.g. the design of diseasemodifying trials and/or clinical purposes.