In vivo hypothalamic regional volumetry across the frontotemporal dementia spectrum

Noah Lorenz Shapiro, Emily Todd, Benjamin Billot, David M. Cash, Juan Eugenio Iglesias, Jason D. Warren, Jonathan D. Rohrer, Martina BocchettaNeuroImage 2022 · DOI

TL;DR

Frontotemporal dementia (FTD) is associated with significant shrinkage in the hypothalamus, a brain region crucial for regulating bodily functions like eating. This study reveals that this shrinkage occurs across all FTD subtypes and is linked to abnormal eating behaviors, suggesting the hypothalamus could be an early target for FTD pathology and a potential biomarker.

What Problem Does This Paper Address?

Frontotemporal dementia (FTD) is a group of neurodegenerative diseases characterized by a wide range of symptoms, including behavioral changes, language difficulties, and motor impairments. While the FTD spectrum is complex and heterogeneous, involving various genetic and pathological causes, the specific role of subcortical structures like the hypothalamus in FTD symptomatology has not been fully elucidated. The hypothalamus plays a vital role in maintaining homeostasis, regulating neuroendocrine functions, and controlling behaviors such as eating. Abnormal eating behaviors, including changes in food preference and increased appetite, are common in FTD, particularly in the behavioral variant (bvFTD). However, a comprehensive investigation into how different hypothalamic regions are affected across the entire FTD spectrum and how these changes relate to eating behaviors has been lacking, partly due to the challenges in accurately measuring hypothalamic volumes from MRI scans.

Methods

The researchers utilized an automated segmentation tool, previously developed by Billot et al. (2020), to measure hypothalamic regional volumes from T1-weighted MRI scans. This tool employs a deep convolutional neural network for segmentation, allowing for efficient and accurate analysis of large cohorts. They analyzed data from 439 FTD patients, categorized by clinical diagnosis (bvFTD, FTD-MND, svPPA, nfvPPA, PPA-NOS), genetic mutations (MAPT, C9orf72, GRN), and pathological subtypes (tauopathy, TDP-43opathy, FUSopathy). These volumes were compared to those of 118 age-matched healthy controls. Additionally, they correlated hypothalamic volumes with the presence of abnormal eating behaviors assessed using the revised version of the Cambridge Behavioural Inventory (CBI-R). Statistical analyses included ANCOVA (adjusted for age, sex, scanner type, and total intracranial volume) and Spearman's rho correlation, with corrections for multiple comparisons.

Datasets

The study used a retrospective cohort of 439 FTD patients and 118 age-matched healthy controls from the UCL Dementia Research Centre FTD MRI database. The dataset included patients with various clinical, genetic, and pathological diagnoses of FTD. A subset of 130 patients also underwent assessment of behavioral symptoms using the Cambridge Behavioural Inventory-Revised version (CBI-R).

Key Insights and Ideas

  • Hypothalamic Shrinkage is Pervasive in FTD: The study provides strong evidence that hypothalamic volume reduction is a common feature across all FTD subtypes, not just specific variants.
  • Regional Specificity Matters: While the entire hypothalamus is affected, different subregions show varying degrees of atrophy, suggesting distinct roles in FTD pathology and symptomatology. The anterior and posterior regions appear to be particularly vulnerable.
  • Link to Eating Behaviors: The observed hypothalamic volume reductions, especially in anterior and posterior regions, are significantly correlated with abnormal eating behaviors, highlighting the hypothalamus's critical role in regulating appetite and food-related behaviors in FTD.
  • Potential Early Biomarker: The widespread involvement of the hypothalamus, even in early stages of the disease, suggests it could serve as a valuable imaging biomarker for FTD diagnosis and progression.
  • Differential Pathology Impacts: The study hints at distinct pathological mechanisms underlying hypothalamic involvement based on genetic mutations (e.g., MAPT vs. C9orf72) and proteinopathies (e.g., Tau vs. TDP-43), with some regions being differentially affected.

Key Results

  • FTD patients, on average, showed a 14% smaller hypothalamic volume compared to controls.
  • The FTD-MND group exhibited the most significant reduction (20%), followed by svPPA and bvFTD (15-16%).
  • Specific hypothalamic regions showed substantial volume loss: anterior regions were reduced by 26-42%, and posterior regions by 17-35% in various FTD groups compared to controls.
  • The inferior tuberal (infTub) region was largely spared, except in tauopathies and TDP-43 type C, and did not correlate with eating symptoms.
  • Abnormal eating behaviors were negatively correlated with hypothalamic volumes in most regions (except infTub) across the FTD cohort.
  • Decline in table manners was specifically linked to reduced anterior hypothalamic volumes, while preference for sweet foods correlated with reduced superior tuberal and anterior areas.
  • MAPT mutation carriers showed the largest reduction in whole hypothalamic volume (25%) among genetic groups.
  • FUS pathology showed the most pronounced hypothalamic volume loss, particularly in the anterior inferior hypothalamus (62% reduction).

Limitations and Future Work

The authors acknowledge several limitations:

  • Small Sample Sizes: Subgroups like FUS and FTD-MND had very small sample sizes, making their results exploratory and requiring caution in interpretation.
  • Limited Behavioral Data: Not all patients had complete behavioral data, limiting the scope of correlations.
  • CBI-R Limitations: The CBI-R's eating behavior assessment is limited to four items, potentially not capturing the full complexity of these behaviors.
  • Image Resolution Variability: Different MRI scanner types and resolutions could have influenced segmentation accuracy, particularly in anterior regions.

Future work should focus on:

  • Longitudinal Studies: To understand the temporal progression of hypothalamic involvement and its relationship with disease stage.
  • Neuropeptide Research: Investigating the role of specific neuropeptides within the hypothalamus as potential therapeutic targets.
  • Interactions with Other Brain Networks: Further exploring how hypothalamic dysfunction interacts with other affected brain regions in FTD.
  • Larger Cohorts: To confirm findings in smaller subgroups and investigate differential effects of pathologies and genetics more robustly.
  • More Comprehensive Eating Behavior Assessments: Utilizing more detailed questionnaires to capture the nuances of eating abnormalities.

Summary updated

Abstract

Abnormal hypothalamic volumes are present in all the FTD forms, but different hypothalamic regions might play a different role in the development of abnormal eating behavioural and metabolic symptoms. These findings might therefore help in the identification of different underlying pathological mechanisms, suggesting the potential use of hypothalamic imaging biomarkers and the research of potential therapeutic targets within the hypothalamic neuropeptides.
BibTeX
@inproceedings{shapiro2022vivo,
  author = {Noah Lorenz Shapiro and Emily Todd and Benjamin Billot and David M. Cash and Juan Eugenio Iglesias and Jason D. Warren and Jonathan D. Rohrer and Martina Bocchetta},
  title = {{In vivo hypothalamic regional volumetry across the frontotemporal dementia spectrum}},
  booktitle = {NeuroImage},
  year = {2022},
  doi = {10.1016/j.nicl.2022.103084},
}
Figure