Distinct patterns of cerebellar atrophy relate to wider patterns of disease-specific brain network degeneration, according to two recent studies. The findings reveal details of cerebellar atrophy in Alzheimer disease (AD) and frontotemporal dementia (FTD), with implications for future research and therapy.

Cerebellar degeneration has largely been disregarded in dementia owing to its association with movement disorders. However, the cerebellum is involved in cognition and emotion, and evidence indicates that cerebellar changes are important in a range of neurological disorders. The two new studies, one led by Michael Hornberger and the other by Jonathan Rohrer, add to this evidence.

Mutations in C9orf72 were associated with atrophy in lobule VIIa–Crus I (arrow) of the cerebellum, which is involved in cognition. Image courtesy of J. Rohrer.

Hornberger and colleagues continued from their previous characterization of cerebellar atrophy in AD, FTD and motor neuron disease. “The question that remained was whether cerebellar changes were associated with cortical changes, or just occurred concomitantly,” says Hornberger.

The team used MRI to visualize cerebellar degeneration in 217 patients with AD or one of three FTD subtypes: behavioural variant FTD, nonfluent variant primary progressive aphasia (nfvPPA), and semantic variant PPA (svPPA). They then compared the atrophy maps with an atlas of cerebral and cerebellar connectivity.

Patterns of cerebellar atrophy differed between AD and bvFTD, and the areas that degenerated had connections to cortical networks that exhibit disease-specific degeneration. In AD, atrophy occurred in the Crus I and II subregions, part of the default mode network. In bvFTD, atrophy occurred in lobule VI of the cerebellum, part of the salience network. “In our view, this is the first evidence that discrete cortical changes are linked to cerebellar atrophy,” says Hornberger.

Differences in cerebellar atrophy between FTD subtypes were less clear, although the patterns seemed to be distinct, suggesting subtype-specific degeneration. The findings of Rohrer and colleagues, who focused on cerebellar atrophy in genetic subtypes of FTD, lend weight to this hypothesis.

Patterns of cerebellar atrophy differed between AD and bvFTD

Rohrer and co-workers expanded on their previous demonstrations of global presymptomatic changes in cerebellar volume in genetic FTD. “We decided to investigate the volume of cerebellar subregions to determine whether specific areas are associated with mutations in the key FTD genes,” explains Rohrer.

Using MRI, the team determined the volumes of cerebellar subregions in 44 patients with mutations in C9orf72, MAPT or GRN. GRN mutations were not associated with cerebellar atrophy, whereas C9orf72 mutations were specifically associated with atrophy in lobule VIIa–Crus I, and MAPT mutations were specifically associated with atrophy in the vermis.

“C9orf72-associated atrophy reflects degeneration of a cortico-thalamo-cerebellar network important in cognition, and the vermis is involved in emotion processing,” explains Rohrer.

Both teams now intend to further investigate the implications of their findings. “The next question is whether cerebellar or cerebral changes occur first,” says Hornberger. He also points out that the cerebellum is often used as a reference area in PET studies of neurodegeneration, and that these findings cast doubt on this approach.

Rohrer and colleagues hope to learn more about the clinical potential of their findings. “Identification of particular network involvement in FTD could help to identify specific targets for therapies,” he concludes.