HAI Book 2025 - Flipbook - Page 551
Tonietto, Matteo
122
Impact of amyloid removal on cerebral volume changes with
Gantenerumab
Matteo Tonietto1, Erica Silvestri1, Christopher Belder2,3, Frederik Barkhof2,4, Nick Fox2, Janice
Smith5, Gregory Klein1
1
Research and Early Development (pRED), Hoffmann-La Roche, Basel, CH
Dementia Research Centre, UCL Queen Square Institute of Neurology, London, GB
3
Adelaide Medical School, University of Adelaide, Adelaide, AU
4
Department of Radiology & Nuclear Medicine, Amsterdam UMC, Vrije Universiteit, Amsterdam, NL
5
Roche Products Ltd, Hoffmann-La Roche, Welwyn Garden City, GB
2
Introduction: Gantenerumab is a monoclonal antibody targeting aggregated forms of amyloid-beta. The phase III
GRADUATE investigational studies aimed to assess the efficacy and safety of gantenerumab in early AD.
Participants treated with Gantenerumab exhibited substantial removal of amyloid plaques and accelerated brain
volume loss compared to placebo. In this study we investigated the relationship between amyloid removal and
brain volume loss.
Methods: Volume change was estimated using the tensor-based morphometry (TBM) technique, which produced
a voxelwise map of relative volume changes for each participant. Spearman correlation was used to assess the
association between amyloid removal and cerebral volume loss in gantenerumab-treated participants. Two types
of analyses were performed: 1) Global analysis, where amyloid removal was measured in centiloids, and volume
change was assessed in the whole brain and whole cortex; 2) Voxelwise analysis, where voxelwise centiloid
changes from baseline were computed using centiloid conversion equations applied to amyloid PET SUVR images
(Figure 1A), and voxelwise volume changes from baseline were obtained from the TBM method (Figure 1B).
Results: Eighty-eight participants (aged 71.2±8.4, F/M=44/44) in the active arm had both amyloid PET and MRI at
screening and at the end of the study. Greater amyloid removal was associated with reduced brain and whole
cortex volume loss (Figure 2). Voxelwise analysis revealed a dual pattern: regions typically affected by tau
pathology, such as the temporoparietal cortex, followed the global results, while regions usually spared by tau
(e.g., thalamus) or those accumulating tau only in later stages of the disease (e.g., Braak VI) exhibited a positive
correlation between amyloid removal and volume loss (Figure 3).
Conclusion: The relationship between amyloid removal and brain volume loss with gantenerumab varies between
regions typically affected by tau pathology and those that are not. Further analyses are ongoing to better
characterize these findings.
HAI2025 - 551
