White Matter Hyperintensities in Behavioral Variant Frontotemporal Dementia and Semantic Variant Primary Progressive Aphasia
Neurol Open Access. 2026 Mar;2(1):e000064. doi: 10.1212/wn9.0000000000000064. Epub 2026 Jan 23.
ABSTRACT
BACKGROUND AND OBJECTIVES: White matter hyperintensities (WMH) in patients with cerebrovascular risk factors (CVRF), are often linked to cerebral vascular changes, but can be caused by genetic variants selectively targeting white matter. In addition, WMH can be present in neurodegenerative disorders such as frontotemporal lobar degeneration (FTLD) and are linked to some FTLD genetic variants. This study aims to investigate WMH burden in patients with behavioral variant frontotemporal dementia (bvFTD) and semantic variant primary progressive aphasia (svPPA) versus controls and to evaluates the influence of CVRF.
METHODS: This cross-sectional retrospective analysis examined individuals meeting research diagnostic criteria for bvFTD and svPPA with high-quality structural MRI at the UCSF Memory and Aging Center between September 2008 and December 2021. WMH burden and spatial distribution were assessed by disease group compared to age- and sex-matched controls and associations with CVRF evaluated.
RESULTS: We included 109 individuals with bvFTD [mean age (SD) 62.9 (8.6), 40% female], 47 with svPPA [mean (SD) age 65.4 (7.5), 51% female], and matched controls. After adjusting for age, apolipoprotein E4 (APOE-ε4) status and intracranial volume (ICV), both disease groups had higher WMH burden compared to controls (bvFTD, R 2 =0.184, p=0.001 and svPPA, R 2 =0.323, p=<0.001). Compared to controls, bvFTD group had more prevalent WMH in the frontal lobe (β=0.403 ; 95% CI 0.27 to 0.54 , p=<0.001), while those with svPPA had more prevalent WMH in the frontal (β=0.462 ; 95% CI 0.26 to 0.66, p <0.001), parietal (β=0.772 ; 95% CI 0.50 to 1.04, p <0.001), temporal (β=0.674 ; 95% CI 0.44 to 0.91, p <0.001), occipital lobes (β=0.364 ; 95% CI 0.14 to 0.59, p=0.002), and corpus callosum (β=0.342 ; 95% CI 0.13 to 0.55, p=0.002). In disease groups, WMH were not significantly associated with CVRF (F=0.468, df=2, p=0.641) suggesting a potential role of non-vascular mechanisms. We did not identify associations between the pathogenic C9orf72 hexanucleotide repeat expansions (HRE) and WMH in bvFTD patients.
DISCUSSION: bvFTD and svPPA are associated with elevated WMH burden independent of CVRF. In bvFTD, WMH are primarily distributed within the frontal lobes, while svPPA shows widespread distribution across lobes. Study limitations include its retrospective, single-center design and limited power for genetic subgroup analyses.
PMID:41799019 | PMC:PMC12962591 | DOI:10.1212/wn9.0000000000000064
