By Makoto Ishii, MD, PhD
Assistant Professor of Neuroscience and Neurology, Feil Family Brain and Mind Research Institute, Department of Neurology, Weill Cornell Medical College
Dr. Ishii reports no financial relationships relevant to this field of study.
In this cross-sectional study, researchers identified distinct patterns of cerebellar atrophy and its association with cognitive dysfunction in the frontotemporal dementia syndromes.
Chen Y, Kumfor F, Landin-Romero R, et al. Cerebellar atrophy and its contribution to cognition in frontotemporal dementias. Ann Neurol 2018;84:98-109.
Frontotemporal dementias (FTDs) are neurodegenerative disorders consisting of three primary clinical syndromes: behavioral-variant (bvFTD), semantic dementia (SD), and primary progressive nonfluent aphasia (PNFA). bvFTD is characterized by personality and behavioral changes, while SD and PNFA primarily affect language. FTDs lead to significant atrophy in the frontal and temporal lobes; however, atrophy in other brain regions, including the cerebellum, has been reported. The cerebellum is well known for its role in motor coordination and planning, but accumulating evidence suggests that the cerebellum also may be involved in cognition. Therefore, Chen et al sought to establish the changes in cerebellar gray matter integrity in bvFTD, SD, and PNFA subjects and to determine the association of cerebellar atrophy with the main cognitive domains.
Study participants included patients diagnosed with FTD (45 bvFTD, 29 SD, and 23 PNFA) from the Frontotemporal Dementia Research clinic in Sydney, Australia. Control subjects included 35 age-, sex-, and education-matched healthy elders without known psychiatric or neurodegenerative disorders or genetic mutations for FTD. All participants underwent extensive testing in the major cognitive domains (attention and processing speed, working memory, language-motor, language-semantics, visuospatial function, episodic memory, executive function, and emotional processing). The researchers obtained whole-brain 3T MRI and analyzed three-dimensional T1-weighted sequences with voxel-based morphometry (VBM). They also obtained mean intensity values of cerebellar grey matter, which then were correlated with cognitive performance across the main cognitive domains.
For all study subjects, there were no differences in demographic characteristics except disease duration, with SD patients having a longer duration than bvFTD or PNFA. On general cognitive testing, as expected, FTD patients scored worse than controls, with SD patients doing worse than bvFTD and PNFA. Disease severity based on the Frontotemporal Dementia Rating Scale was the worst in bvFTD compared to SD and PNFA groups. Neuropsychological testing revealed the typical patterns expected for the FTD syndromes, with bvFTD impaired on all measures except language-motor. The SD group was impaired on attention and processing speed, working memory, language-semantics, and emotion processing. The PNFA group was impaired on attention and processing speed, working memory, language-motor, and executive function.
A VBM analysis of the whole-brain MRI revealed typical canonical patterns of atrophy specific to each FTD syndrome. In bvFTD, there was widespread bilateral intensity reduction primarily in the frontal and temporal lobes. In SD, there was a left greater than right anterior temporal lobe atrophy. In PNFA, there was a left greater than right inferior frontoinsular region atrophy. All three FTD syndromes showed reduced cerebellar grey matter intensity compared to controls. When each FTD syndrome was compared separately, specific atrophy patterns were noted. In bvFTD, there was widespread bilateral intensity reduction involving most lobules and the vermis. In SD, there were more focal changes bilaterally in lobules VI, Crus I, and Crus II. In PNFA, a significant intensity decrease was found bilaterally in Crus I, Crus II, and lobule VIIb and in the right lobule VI. Across all FTD patients, there was overlapping atrophy in Crus I, Crus II, and lobule VI bilaterally.
Interestingly, there were significant correlations between cerebellar atrophy and cognitive scores, which showed a distinct pattern among the FTD syndromes. Comparing bvFTD and controls, attention and processing speed and working memory were correlated with cerebellar atrophy, while visuospatial function and language-motor correlated with cerebellar atrophy in SD and PNFA, respectively. When all participants were included in the analysis, working memory was found to correlate significantly with right cerebellar lobule VI, Crus I, and Crus II.
COMMENTARY
The main results from this paper suggest that cerebellar atrophy in FTDs can affect select cognitive domains, and that cerebellar atrophy is not simply due to global brain neurodegeneration but is specific to FTDs. These authors provided strong additional evidence for the role of the cerebellum in cognition; however, there were significant limitations that need to be addressed. First, replication is needed in additional cohorts to verify these results. Additionally, since FTD was diagnosed clinically, the use of FTD biomarkers (e.g., tau positron emission tomography), once validated, would be essential to avoid possible misclassification of patients in future studies. Furthermore, longitudinal studies in FTD patients examining the changes in the cerebellum as cognition declines over time would provide stronger evidence for cerebellar atrophy contributing to cognitive decline. Investigating cerebellar structural changes over time also would identify specific atrophy patterns in the cerebellum, which could provide further evidence that cerebellar atrophy is not simply a consequence of global brain atrophy. This study also cannot establish causality or identify the mechanisms underlying cerebellar atrophy. It is plausible that loss of cortical neurons feeds back to the cerebellum to cause atrophy. Alternatively, there may be specific neurodegenerative processes that occur in the cerebellum independent of those in the cortex. Longitudinal functional connectivity studies and mechanistic studies using cellular and animal models would be useful to address these questions. Finally, studies investigating the cerebellum and its role in other behaviors affected in FTD, such as empathy and eating behavior, are needed. Although significant work needs to be done, additional research building on this important study could further advance our understanding of the cerebellum in nonmotor function.
