Quantified Brain MR Volumes Contribute to Diagnostic Delineation of Behavioral Variant Frontotemporal Dementia from Early Onset Alzheimer’s Disease (2489)

Objective: We tested the hypothesis that MR volumes alone can diagnostically delineate behavioral variant frontotemporal dementia (bvFTD) from Early Onset Alzheimer’s Disease (EOAD). Background: As bvFTD must often be distinguished from AD, assessment of volumetric quantification of brain MRI may identify distinct atrophy patterns to address this clinical need. Design/Methods: Participants from IRB approved studies of bvFTD and EOAD (IRB#16-000496, IRB#16-001491, and IRB#10-001097) with availability of brain MRI with 3D T1 sequences were studied. This resulted in 65 bvFTD (n = 20) and EOAD (n = 45) scans quantified using two software programs. The first is an FDA cleared software program called Neuroreader that produced hippocampal, subcortical, and lobar brain volumes. The second was Volbrain used to quantify hippocampal subfields. For each type of brain volume generated, the following analyses were done: i) Discriminant analysis with leave one out cross validation ii) Input of predicted probabilities from this analysis into an ROC analysis iii) Automated linear regression to identify specific regions of interest that maximize diagnostic delineation between bvFTD and EOAD. Results: No statistically significant differences between bvFTD and EOAD with age or gender were noted. EOAD on average had lower MMSE (t (63) = 2, p = .04) compared to the bvFTD group. bvFTD had lower brain volume percentiles than EOAD in multiple regions including frontal lobes, thalamus and putamen. Individuals with EOAD had lower brain volumes than bvFTD in the parietal lobes. ROC analyses showed separation accuracy of 99.3% with the Neuroreader regional volume percentiles and 80.2% with hippocampal subfields volumes from Volbrain. Automated linear regression analyses identified the parietal lobe Neuroreader percentile as the most predictive feature and left CA2-CA3 as the most predictive Volbrain feature. Conclusions: Percentiles from an MR based volumetric quantification software program can delineate between bvFTD from EOAD. Hippocampal subfield volumes also attained reasonable diagnostic separation. Disclosure: Dr. Meysami has nothing to disclose. Prof. Raji has received personal compensation in the range of $50,000-$99,999 for serving as a Consultant for Brainreader ApS. Prof. Raji has received personal compensation in the range of $10,000-$49,999 for serving on a Scientific Advisory or Data Safety Monitoring board for Apollo Health . Prof. Raji has received personal compensation in the range of $50,000-$99,999 for serving as an Expert Witness for Neurevolution Medicine. Dr. Mendez has received personal compensation in the range of $500-$4,999 for serving on a Speakers Bureau for Medical Education Speakers’ Bureau. Dr. Mendez has received personal compensation in the range of $500-$4,999 for serving as an Editor, Associate Editor, or Editorial Advisory Board Member for UpToDate. The institution of Dr. Mendez has received research support from NIH. Dr. Mendez has received publishing royalties from a publication relating to health care.