Association between DBM-derived Atrophy Patterns and Cognition in Frontotemporal Dementia Variants
Poster No:
1464
Submission Type:
Abstract Submission
Authors:
Amelie Metz1, Yashar Zeighami2, Simon Ducharme3, Sylvia Villeneuve4, Mahsa Dadar5
Institutions:
1McGill University, Montreal, Quebec, 2Douglas Research Centre, Montreal, Quebec, 3McGill University, Montreal, ., 4Brain Imaging Centre, Douglas Institute Research Centre, Montreal, Qc, 5McGill University, Montreal, QC
First Author:
Co-Author(s):
Introduction:
Frontotemporal Dementia (FTD) is a prevalent form of early-onset dementia characterized by progressive neurodegeneration (1,2). It encompasses a group of heterogeneous neuropathological disorders, including behavioral variant frontotemporal dementia (bvFTD), nonfluent variant primary progressive aphasia (nfvPPA), and semantic variant primary progressive aphasia (svPPA), each exhibiting distinct symptoms. BvFTD initially presents with abnormal behavior, changes in personality, and reduced executive control while in primary progressive aphasias, cognitive deficits predominantly manifest themselves in distinct skills within the language domain (3). Previous magnetic-resonance imaging (MRI) studies have highlighted atrophy patterns in specific brain regions corresponding to symptom manifestations (2) but few have validated these findings using deformation-based morphometry (DBM) (4-7) which offers increased sensitivity to subtle local differences in structural image contrasts compared to traditional methods. Here, we tested whether DBM-derived brain atrophy patterns in the core variants of FTD are associated with severity of cognitive impairment and whether this relationship differs between the phenotypic subtypes.
Methods:
A total of 136 patients (70 bvFTD, 36 svPPA, 30 nfvPPA) and 133 cognitively unimpaired controls from the frontotemporal lobar degeneration neuroimaging initiative (FTLDNI) database underwent high-resolution brain MRI and clinical, neuropsychological examination. Cognitive measures included global cognition, language, memory and executive functions. DBM values (4) were calculated as an estimation of regional cortical and subcortical atrophy. Atlas-based associations between DBM values and performance across different cognitive tests were assessed using partial least squares (PLS)(8,9). We then applied linear regression models to discern the group differences regarding the relationship between atrophy and cognitive decline in the three FTD variants. Lastly, we assessed whether the combination of neural and behavioral patterns in the latent variables identified in the PLS analysis could be used as features in a machine learning model to predict FTD subtypes in patients.
Results:
PLS revealed three significant latent variables (LV) that combined accounted for over 85% (42.98%, 35.60%, and 6.97% respectively, permuted p-values<0.01) of the shared covariance between cognitive and brain atrophy measures. It identified neural networks whose collective atrophy was associated with the clinical phenotype (Fig.1). The atrophy pattern included left-hemispheric subcortical structures (LV-I), left temporal and bilateral subcortical areas (LV-II), and frontal and right subcortical regions (LV-III). Brain scores were significantly related to behavioral scores in all LVs in that subjects with greater expression of the respective brain pattern had greater impairment in most cognitive measures included in the LV (LV-I: R2=0.37, p=0.0, LV-II: R2=0.77, p=0.0, LV-III: R2=0.43, p=0.0). In LV-II, the brain pattern had a higher impact on cognition in bvFTD whereby the atrophy pattern was related to higher performance in language tests and lower scores in executive function (bvFTD vs nfvPPA tStat=-2.48 p=0.01, bvFTD vs svPPA tStat=3.52 p=0.0). In LV-II, subjects with svPPA showed higher performance in the included cognitive scores (vs bvFTD tStat=-5.20 p=0.0, vs nfvPPA tStat=-6.21 p=0.0), whereas subjects with nfvPPA had higher scores in LV-III compared to bvFTD (tStat=3.06 p=0.0, Fig.2). Individual variation in the atrophy and behavioral patterns predicted classification of patients into FTD subtypes with an accuracy of 75.09%.
·Latent variables (LV-I - III) obtained from the PLS analysis
·Individual patients’ brain versus behavioral PLS score and group differences between FTD variants in LV-I (top), LV-II (middle), LV-III (bottom)
Conclusions:
Findings in this study demonstrate a robust mapping between neurodegeneration as estimated by DBM values and the cognitive manifestations of FTD variants. The combination of DBM and multivariate statistical methods could potentially serve as an imaging biomarker for early disease severity assessment and phenotyping in FTD.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 2
Modeling and Analysis Methods:
Classification and Predictive Modeling 1
Multivariate Approaches
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Cortical Anatomy and Brain Mapping
Novel Imaging Acquisition Methods:
Anatomical MRI
Keywords:
Aging
Cognition
Degenerative Disease
Language
Machine Learning
Modeling
MRI
Multivariate
Sub-Cortical
Other - Deformation-based morphometry
1|2Indicates the priority used for review
Provide references using author date format
4) Cardenas, V. A. (2007). Deformation-Based Morphometry Reveals Brain Atrophy in Frontotemporal Dementia.
3) Gorno-Tempini, M. L. (2011). Classification of primary progressive aphasia and its variants. Neurology, 76(11), 1006–1014.
2) Olney, N. T. (2017). Frontotemporal Dementia. Neurologic Clinics, 35(2), 339–374.
5) Manera, A. L. (2019). Deformation based morphometry study of longitudinal MRI changes in behavioral variant frontotemporal dementia. NeuroImage: Clinical, 24, 102079.
8) McIntosh, A. R. (2004). Partial least squares analysis of neuroimaging data: Applications and advances. NeuroImage, 23 Suppl 1, S250-263.
6) Shafiei, G. (2023). Network structure and transcriptomic vulnerability shape atrophy in frontotemporal dementia. Brain, 146(1), 321–336.
7) Wisse, L. E. M. (2021). Cross-sectional and longitudinal medial temporal lobe subregional atrophy patterns in semantic variant primary progressive aphasia. Neurobiology of Aging, 98, 231–241.
9) Zeighami, Y. (2019). A clinical-anatomical signature of Parkinson’s disease identified with partial least squares and magnetic resonance imaging. NeuroImage, 190, 69–78.
