Disruption of Reflexive Attention and Eye Movements in an Individual with a Collicular Lesion

Abstract

The superior colliculus (SC) plays a central role in the control of saccadic eye movements and has also been implicated in control of covert spatial attention. While there is a growing body of evidence from studies of awake behaving primates that supports these proposals, direct evidence from humans has been sparse. In the present study we tested a patient with thiamine deficiency and a lesion of the SC, who performed both eye movement tasks (prosaccades and antisaccades, with or without a gap) and a covert spatial attention task assessing inhibition of return (IOR). For eye movements, the gap effect was disrupted, and abnormal saccade metrics occurred, with reflexive eye movements being disrupted moreso than voluntary eye movements. Each of these effects resolved coincident with thiamine treatment. The covert attention task revealed a complete absence of IOR. The unequal disruption of voluntary and reflexive eye movements supports the idea that oculomotor responses can be generated in an independent fashion by frontal cortical and lower level neural systems. The role of the SC and other structures in these orienting processes is discussed.

Acknowledgments

This work was supported in part by grants from the National Alliance for Research in Schizophrenia and Depression to K.A. Briand, and by NIH R01 grants (MH63340, MH65492) to A.B. Sereno. Special thanks to Cate Cathey for technical support.

Notes

¹Other evidence for a specifically collicular role in IOR comes from developmental studies (Simion, Valenza, Umilta & Dalla Barba, 1995; Valenza, Simion & Umilta, 1994) and studies of hemianopia patients (Danziger, Fendrich & Rafal, 1997).

²We were unable to definitively conclude that SR's collicular abnormality and accompanying symptoms were specifically due to WE, because he unexpectedly passed away at home before any further diagnostic testing could be carried out and an autopsy was not performed. The pattern of results that follow, wherein performance of some behavioral tasks improved following thiamine treatment, suggests the diagnosis of a possible WE induced collicular lesion that was to some extent reversible. However, as we indicated previously, the presence of other diseases that could have affected the SC could not be ruled out.

³The eye tracking system was specifically set up to test the left eye (i.e., illumination levels and position of light source) and was not optimal for testing the right eye. Furthermore, SR's difficulty with vertical eye movements for his right eye made calibration almost impossible (i.e., to calibrate the eye tracker subjects have to gaze at highly eccentric points in horizontal and vertical directions before testing), thus preventing reliable measurement of true eye position.

⁴For both saccade and AS tasks, we chose to test SR in the gap condition only after he had been first tested in the overlap condition. Our rationale for using this order was that we predicted that SR's collicular damage would disrupt his gap effect. By having the gap condition presented second, we felt that practice effects would actually work against us, as saccadic RTs might improve slightly as SR became more familiar with the testing procedure. Thus, a demonstration of no gap effect even under these biased conditions is all the more impressive. We also did not believe that fatigue is a major issue, given the brief duration of each saccade block (40 or 28 trials) as well as the inclusion of frequent rest periods.

^ap < .005.

^bp < .03.

^cp < .06.

⁵We analyzed the IOR data by performing a three factor repeated measures ANOVA (Response × Cue Type × SOA). This was conducted as follows: (1) for each condition, missing trials (e.g., due to errors or anticipations) were replaced with that cell's mean RT. (2) Each row in the repeated measures ANOVA consisted of the n^th trial in each cell/condition, for example the first row (the first “subject”) consisted of the RTs from the first trial presented in each of the testing conditions, the second row consisted of the RTs from the second trial presented in each testing condition, down through to the last (16^th) row (the last RT from each testing condition). In effect, each successive trial within a testing condition was treated as coming from a different subject, but sets of trials corresponding to the n^th instance of each type were treated as coming from the same subject.

⁶While IOR is robust with this paradigm, not every subject in a given study will show it. For the study in which we originally reported the control data used here (Briand et al., 2001), 6 of 8 controls showed IOR, and 6 of 7 PD patients showed the effect. While it could be argued that SR's data is merely due to his being one of that minority of individuals who fail to show IOR, we feel this would be an unjustified assumption. The converging evidence in the literature of a SC/IOR connection suggests a more parsimonious explanation, that is, that the collicular damage observed in SR is somehow the cause.