White matter hyperintensities precede other biomarkers in GRN frontotemporal dementia

Abstract

Introduction: Increased white matter hyperintensities (WMHs) have been reported in genetic frontotemporal dementia (FTD) in small studies, but the sequence of WMH abnormalities relative to other biomarkers is unclear.

Methods: Using a large dataset (n = 763 GENFI2 participants), we measured WMHs and examined them across genetic FTD variants and stages. Cortical and subcortical volumes were parcellated, and serum neurofilament light chain (NfL) levels were measured. Biomarker progression was assessed with discriminative event-based and regression modeling.

Results: Symptomatic GRN carriers showed elevated WMHs, primarily in the frontal lobe, while no significant increase was observed in symptomatic C9orf72 or MAPT carriers. WMH abnormalities preceded NfL elevation, ventricular enlargement, and cortical atrophy. Longitudinally, baseline WMHs predicted subcortical changes, while subcortical volumes did not predict WMH changes, suggesting WMHs may precede neurodegeneration.

Discussion: WMHs are elevated in a subset of GRN-associated FTD. When present, they appear early and should be considered in disease progression models.

Highlights: Elevated WMH volumes are found predominantly in symptomatic GRN. WMH accumulation is mostly observed in the frontal lobe. WMH abnormalities appear early in GRN-associated FTD, before NfL, atrophy, and ventriculomegaly. Longitudinally, WMH volumes can predict subcortical changes, but not vice versa. WMHs are key early markers in GRN-associated FTD and should be included in progression models.

Keywords: C9orf72; FTD; GRN; MAPT; biomarker sequence; dementia; disease progression; early marker; event‐based modeling; magnetic resonance imaging; neurodegeneration; neurofilament light chain; neuroimaging; progranulin; white matter.

FIGURE 1

Flowchart of inclusion and exclusion criteria in study. QC, quality control; WMH, white matter hyperintensity.

FIGURE 2

Voxel‐wise distribution of white matter hyperintensity (WMH) prevalence in mutation groups. The color bar represents the proportion of participants within each cohort exhibiting WMHs at specific voxel locations.

FIGURE 3

Regional and total white matter hyperintensity (WMH) volume across genetic groups. (A) Heatmap showing mean age‐ and sex‐adjusted WMH volume (log‐transformed) across brain regions and clinical/genetic subgroups. Values represent group‐level averages of residualized WMH volumes. (B) Violin plots of total age‐ and sex‐adjusted WMH volume (log‐transformed) across mutation groups and controls, stratified by clinical status. Black horizontal bars indicate group means. Among symptomatic individuals, GRN carriers exhibited significantly higher WMH volumes compared to controls (P _Bonferroni < 0.001); no other between‐group differences reached statistical significance. Full statistical results are reported in Tables S4 and S5.

FIGURE 4

Sequence of biomarker abnormalities. The positional variance diagram for the GRN cohort illustrates the most probable sequence of biomarker abnormalities along with their corresponding uncertainties. The y‐axis (from top to bottom) orders the biomarkers by the most likely sequence as estimated by the Discriminative Event‐Based Modeling model, while the x‐axis indicates the position of each biomarker in the sequence, ranging from one to the total number of biomarkers. The color intensity of each square represents the frequency with which a biomarker was placed at a specific position during bootstrap resampling. The spread from bootstrap resampling reflects the standard error of the distribution, representing the uncertainty in the estimated ordering. GFAP, glial fibrillary acidic protein; NfL, neurofilament light chain; WMH, white matter hyperintensity.

FIGURE 5

Associations between white matter hyperintensity (WMH) and subcortical biomarkers and predictability of longitudinal variations. Diagrams illustrate the associations between predictor biomarkers (left side) and the rate of change of response biomarkers (right side). The width of the connecting lines represents the t‐statistic, indicating the strength of the predictive association. (A) Chord diagram showing t‐statistics for all tests, including both significant and non‐significant associations. (B) Diagram displaying only the significant associations (FDR‐corrected p value < 0.05). Higher baseline WMH volumes are associated with faster volume decline in the amygdala, hippocampus, and cingulate cortex.

FIGURE 6

Relationship of baseline WMH volume to subsequent subcortical atrophy progression. Each scatter plot corresponds to one of the significant associations identified in Figure 5B, with baseline WMH volume (log‐transformed and z‐scored across the full cohort) on the x‐axis and the annualized rate of change in subcortical volume (total intracranial volume‐normalized and z‐scored prior to slope estimation) on the y‐axis. Linear regression lines with 95% confidence intervals illustrate the relationships.