Background

Brain atrophy in frontotemporal dementia (FTD) has been shown to follow the topology of large-scale brain networks1,2 and partially co-localize with neurotransmitter receptor distributions.3 We use network spreading models4-6 (Fig. 1) to test how well connectivity-driven pathology spread and receptor-dependent regional vulnerability can explain FTD atrophy.

 

Methods

MRI data from bvFTD, nfvPPA, and svPPA patients (Table 1) were used to derive atrophy maps (Fig. 2) by group-averaging W-scores of brain volumes relative to controls.
Data were obtained from the Frontotemporal Lobar Degeneration Neuroimaging Initiative.

PET-derived neurotransmitter receptor density maps7 (Fig. 3) were incorporated into weighted models as regional modulators of pathology production or spread.

Models tested amplifying or protective effects of receptor densities on pathology accumulation or propagation. 

 

Results

bvFTD & nfvPPA: Best models had local production down-weighted by 5-HT1B.

svPPA: The unweighted model   predicted regional atrophy   well. The best weighted   model provided minimal   improvement.

Table 2 shows model   comparison across   syndromes.

 

Conclusions

svPPA atrophy pattern is explained well by connectivity-driven pathology spread alone.

Using 5-HT1B densities to down-weight production improved model performance in predicting atrophy patterns in bvFTD & nfvPPA.

Neurotransmitter receptor densities may differentially modulate regional pathology accumulation across clinical FTD phenotypes.

 

References

1. Shafiei et al., Brain, 2022; 2. Agosta et al., Brain, 2024; 3. Premi et al., Neurobiol. Dis., 2023; 4. Raj et al., Neuron, 2012; 5. Weickenmeier et al., Phys Rev Lett, 2018; 6. He et al. MICCAI, 2023; 7. Hansen et al., Nat Neurosci, 2022