Eye movements in neurocognitive disorders and frontotemporal dementia

Abstract

Introduction/Objective: Frontotemporal dementia (FTD) is a young onset heterogeneous neurodegenerative condition which presents with changes to behaviour and language and causes degeneration of a number of cortical regions, particularly the frontal and temporal lobes. Onset of the disease is often insidious, and a wide variety of presentations can be seen clinically. Signs and symptoms can be difficult to quantify early in the disease and there is a large overlap with a range of other diagnoses. Neuropsychological testing can often under-represent the deficit seen behaviourally by collateral observers and return a non-specific profile.

Vision is a dominant sense and the brain pathways and networks used for eye movements are widespread and well-documented. Therefore the aim of this thesis was to characterise eye movements in frontotemporal dementia across a hierarchical range of movements and to investigate if eye movements could be linked to neuropsychological deficits occurring in the disease.

Method: FTD and control participants were recruited to examine eye movement changes in the disease. A hierarchical battery of eye movements were examined beginning with the simplest movements, saccades, which require only a short pathway through the brainstem, and progressed to examining scan paths in more complex tasks including facial emotion recognition, visual search and driving, thus recruiting additional cortical regions and networks. These tasks were selected as FTD patients are reported to have difficulty correctly undertaking them. Saccadic eye movements included horizontal and vertical reflexive saccades, which were then built upon by examining higher-order saccades including antisaccades, self-paced, predictive and memory-guided saccades. Results were examined with regard to eye movement characteristics reported in the literature for other neurocognitive conditions including parkinson’s-plus disorders, schizophrenia and Alzheimer’s disease.

Results: A total of 28 FTD variant and 25 controls were recruited to participate. Not all participants undertook all tasks. Some participants were unable to undertake specific tasks either due to complexity or in the case of the driving task because they had never held a licence. In addition some participants had to stop partway through a task due to simulator sickness. Within the FTD group insufficient semantic dementia and progressive non-fluent aphasia patient participants were able to be recruited and so their data is presented in each task as pilot data. There was also a reduction in data due to some trials being excluded for poor tracking. Saccadic eye movements revealed an increased latency across all basic movements. Scan paths to emotional faces from the Ekman Pictures of Facial Affect series revealed a reduced gaze time to the right eye for both implicit and explicit face viewing but otherwise resembled those of controls. Eye movements whilst driving revealed gaze times not significantly different to any area of interest other than to pedestrians, which elicited less gaze time from bvFTD participants. Eye movements during visual search revealed an increased reaction time intercept but a similar processing time per item to control participants. Scan paths to all tasks were qualitatively normal.

Discussion: Eye movements were analysed across a hierarchical range of movements, starting with simple eye movement circuitry and progressively incorporating more brain regions and networks. Stimuli were selected to specifically target tasks known to produce errors in FTD. Our study revealed bvFTD patients display eye movement changes sharing two prominent features: a psychomotor slowing and a disruption of salience, possibly due to changes in their assignment of mind. Psychomotor slowing has been evidenced in both saccade and visual search tasks as patients were able to correctly undertake the task but at a significantly slower pace, seen in increased reaction times and increased latencies. The alterations in salience noted including reduced gaze time to the right eye and a lack of gaze time to pedestrians whilst maintaining qualitatively normal scan paths and gaze durations to other areas of interest, leads to the conclusion that social cues are detected but not processed and interpreted correctly. Qualitatively normal scan paths are in stark contrast to those seen in a number of overlapping neurocognitive diseases including schizophrenia and Alzheimer’s disease, which have been reported in the literature to lead to hypo-scanning and disorganised scan paths respectively. Our study did not differentiate different pathological groups; however there were a wide range of responses in each task which may reflect different underlying pathologies, this finding warrants future studies examining saccadic paradigms other than reflexive and antisaccades. Scan paths, such as to faces, may be able to be utilised diagnostically in conjunction with other examinations to differentiate between bvFTD and other degenerative groups. Performance on the driving simulator task suggests potential for a driving task to be developed to examine safety to drive.