Visual search as a cognitive marker to discriminate progressive supranuclear palsy from Parkinson’s disease

Abstract

Early and accurate diagnosis of Progressive Supranuclear Palsy is a high priority for both patients and clinicians. Nearly half of individuals with Progressive Supranuclear Palsy are initially given an alternative diagnosis—most commonly Parkinson’s disease—and the correct diagnosis of Progressive Supranuclear Palsy is delayed by an average of three years compared with Parkinson’s disease. Problems with cognition arise early in the progression of Progressive Supranuclear Palsy, including profoundly disturbed visual search. This search deficit is likely caused by a combination of oculomotor, visuospatial, attentional and executive dysfunction, which is not observed in Parkinson’s. Since assessment of visual search is not routinely included in clinical testing for Progressive Supranuclear Palsy, it may represent a promising method for differentiating these movement disorders. However, few studies have systematically compared visual search in Progressive Supranuclear Palsy and Parkinson’s disease, so the extent to which search deficits can reliably differentiate the two diseases is unknown.

To address this issue, patients with Progressive Supranuclear Palsy (N = 28), Parkinson’s disease (N=28) and age matched controls (N=27) completed four tests of visuospatial attention: an easy feature search task, a difficult feature search task, a conjunction search task and a paper-and-pencil cancellation task (the Bells test). We measured perceptual sensitivity (d prime), search time and detection accuracy. Search efficiency in the Bells test was also modelled using Quality of Search score. In addition, patients completed tests of oculomotor system integrity, the Montreal Cognitive Assessment and the National Eye Institute Visual Function Questionnaire 25.

As predicted, Progressive Supranuclear Palsy was associated with a profound search deficit: search was slow and highly disorganised on all tests. Furthermore, reflexive orienting along the vertical axis was impaired on the easy feature search. Critically, Receiver Operator Characteristic analyses revealed that the Quality of Search score derived from the Bells test classified Progressive Supranuclear Palsy with a sensitivity of 82%, a specificity of 96%, and Area under the Curve of .90. Given the Bells test is fast to administer (∼5 mins), free, relatively simple to score and already widely used in neuropsychology assessment, it has the potential to be a highly effective addition to the battery of tests used in the differential diagnosis of Progressive Supranuclear Palsy