P1‐419: USING A BRAIN NETWORK APPROACH TO PREDICT GENETIC MUTATION IN INDIVIDUAL PATIENTS WITH FAMILIAL FRONTOTEMPORAL DEMENTIA

Improving clinicians’ ability to predict genetic mutations causing Frontotemporal dementia (FTD) is useful for diagnostic evaluation and potentially referral to appropriate studies including therapeutic trials. Previous studies have shown neuroanatomical differences between the three main genetic mutations in familial FTD, microtubule-associated protein tau (MAPT), chromosome 9 open reading frame 72 (c9orf72) and progranulin (GRN) mutations. MAPT has been associated with atrophy predominantly in the anteromedial temporal lobes; c9orf72 with widespread atrophy in frontoparietal regions; and GRN with asymmetrical atrophy in frontotemporoparietal regions (Whitwell et al., 2012; Rohrer et al., 2010). Using a brain network approach, the main goal of this study was to assess whether different patterns of atrophy in large-scale brain networks identified in prior work (Seeley et al., 2012; Touroutoglou et al., 2012; Vincent et al., 2008) predict genetic mutations in FTD patients spanning different clinical diagnoses. We hypothesized that (1) relatively greater atrophy in the salience and semantic appraisal networks, including frontoinsula and anterior temporal regions would be present in MAPT, (2) relatively greater atrophy in the frontoparietal network would be associated with c9orf72 and (3) relatively greater asymmetrical atrophy in frontoparietal and salience networks would be associated with GRN. The analysis included symptomatic participants recruited from the Longitudinal Evaluation of Familial Frontotemporal Dementia Study (LEFFTDS). To determine network-based patterns of brain atrophy at the individual level, we used single-subject general linear model analysis and compared structural MRI data from healthy controls (N = 166) versus MAPT (N = 18), c9orf72 (N = 27), and GRN (N= 10) mutation patients. Cortical thickness measurements were calculated using FreeSurfer v6.0. An untrained rater predicted gene mutations based on the patterns of brain network atrophy. Prediction accuracy was measured with the percentage of correct predictions. The results showed 74% prediction accuracy for MAPT, 72% prediction accuracy for c9orf72 and 60% prediction accuracy for GRN. Our findings provide preliminary evidence suggesting that a brain network approach can be useful in individual patients to predict the specific genetic mutations in the three major genetic variants of familial FTD.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.