Adaptation to right-deviating prisms is a promising intervention for the rehabilitation of patients with left spatial neglect. In order to test the lateral specificity of prism adaptation on left neglect, the present study evaluated the effect of left-deviating prism on straight-ahead pointing movements and on several classical neuropsychological tests in a group of five right brain-damaged patients with left spatial neglect. A group of healthy subjects was also included for comparison purposes. After a single session of exposing simple manual pointing to left-deviating prisms, contrary to healthy controls, none of the patients showed a reliable change of the straight-ahead pointing movement in the dark. No significant modification of attentional paper-and-pencil tasks was either observed immediately or 2 hours after prism adaptation. These results suggest that the therapeutic effect of prism adaptation on left spatial neglect relies on a specific lateralized mechanism. Evidence for a directional effect for prism adaptation both in terms of the side of the visuomanual adaptation and therefore possibly in terms of the side of brain affected by the stimulation is discussed.
Patients with right cerebral hemisphere lesions often show a reduced tendency to respond to stimuli and to search actively for them in the contralateral part of space [
Among these, prism adaptation is one of the most promising therapeutic interventions [
This impressive generalization and long-standing effects of prism adaptation have revived interest in the neuro-cognitive mechanisms by which it has been achieved. The most two basic questions about the mechanisms of action of prism adaptation are (i) whether adaptation per se is necessary to produce cognitive after-effects or whether simple visuomanual pointing could produce similar effects, and (ii) whether this adaptation is specific in terms of its direction. As a matter of facts, such specificity has been demonstrated in healthy individuals (see for review [
Patients were selected from the Neuro-rehabilitation Department of the Hospices Civils de Lyon, France. Inclusion criteria were right-handed patients with left spatial neglect after right hemispheric ischemic or hemorrhagic stroke. Patients with previous history of stroke, psychiatric diseases, global cognitive deterioration, or any impairment that could compromise comprehension and compliance with the tasks were excluded.
For all patients screened, hand preference was assessed by the Edinburgh inventory [
A group of healthy subjects was included for comparison purposes.
This study was conducted with the informed consent of the participants, in agreement with the French law (March 2002) and the Helsinki declaration relative to patient’s rights.
The sample comprised six healthy subjects and five patients aged between 67 and 80 years old (see Table
Clinical profiles of each patient.
Patients number | 1 | 2 | 3 | 4 | 5 |
---|---|---|---|---|---|
Sex | M | F | F | F | F |
Age | 80 | 75 | 73 | 67 | 74 |
Time after onset (mt) | 2 | 1.5 | 2.5 | 1 | 1 |
Motor deficit | L hemiparesis | L hemiplegia | L hemiplegia (transient) | L hemiplegia | L hemiparesis |
Somatosensory deficit | + | + | − | + | + |
Hemianopia | + | − | + | − | − |
Constructive apraxia | − | + | − | − | + |
Type of lesion | I (MCA) | H | I (MCA) | I (MCA) | I (MCA) |
Motor and somato-sensory deficits were assessed by a classical clinical examination. Presence of hemianopia was assessed by means of the Goldman perimetry. Constructive apraxia was assessed on copying geometrical drawings.
Abbreviations—+: present; −: absent; mt: month; F: female; M: male; L: left, H: hemorrhagic; I: ischemic; MCA: middle cerebral artery.
Lesion analysis showed the involvement of the inferior, middle, and superior temporal gyri in three patients (patient 1, patient 3, and patient 5); the temporoparietooccipital junction was damaged in two patients (patient 1 and patient 3). Lesions of other brain structures involved the somatosensory parietal cortex (patient 1 and patient 5), the primary motor cortex (patient 1 and patient 5), the occipital cortex (patient 1 and patient 3), the prefrontal and the orbito-frontal cortex (patient 5), the insula (patient 1, patient 2, patient 4, and patient 5), the thalamus (patient 2 and patient 4), the putamen and pallidum (patient 2, patient 4, and patient 5), the internal capsule (patient 2 and patient 4), the caudate nucleus, the hippocampus, and parahippocampus (patient 4). Figure
Lesion anatomy. For each patient, all lesions were mapped using the free MRIcro software and were drawn manually on slices of the high-resolution 3D T1-weighted template MRI scan. This template is oriented to match the Talairach space. Lesions were mapped onto the horizontal slices that correspond to Z-coordinates −16, −8, 0, 8, 16, 24, 32, 64 in the Talairach space by using the identical or the closest matching horizontal slices of each individual. Following radiological convention, the right cerebral hemisphere is displayed on the left side.
Patients’ performance was investigated in sessions that took place before prism adaptation (referred to as “pre”), immediately after (post), and 2 hours after (late). Healthy subjects performed the same tasks as patients before (pre) and immediately after (post) prism adaptation. During each session, left spatial neglect was assessed using line cancellation, balloon test, line bisection, copy of a scene, drawing from memory, and a simple text reading (cf. Section
The participant was seated blindfolded in front of a horizontal box that allowed for an electronic measurement of the finger movement endpoints with an accuracy of 1 deg. Participants were required to point straightahead while their head was kept aligned with the body’s sagittal axis. Seven pointing trials were performed during each of the two assessments.
This test consists of an A4 page containing 40 lines arranged in different direction. The page is placed at body midline. Participants were instructed to cross out all the lines on the page. The score was the total number of lines crossed.
This test consists of two subtests, carried out on two A3 landscape-orientated stimulus sheets, each containing 202 items (circles or balloons). In the first subtest “pop-out”, 22 target balloons are interspersed between 180 circles which play the role of distractor. Subjects were asked to cross out as many balloons as they could find. This test is based on the phenomenon of perceptual “po-pout,” that is, the time taken to detect target of this kind does not increase significantly as the number of distractors increase [
Participants were presented with an A4 page, in front of their body midline, containing twenty lines of different length ranging from 100 mm to 200 mm. Participants were instructed to cut each line in half by placing a small pencil mark through each line as close to its center as possible. The score was the mean percentage of deviation from the true center of the line (the score is positive when the deviation is in the right direction and negative when the deviation is in the left direction).
Participants were required to reproduce a picture made up of five items (4 trees and a house) in the space bellow it. Performances were assessed by two scores: (i) the number of items reproduced and (ii) the number of items symmetrically depicted.
Participants were simply asked to draw a daisy without any model. A score of 1 was given when the daisy was highly asymmetrical, 2 when the drawing was moderately asymmetrical, and 3 when the drawing was symmetrical.
Patients were required to read a simple text. The score on this test represented the number of words omitted or modified.
Left-deviating prism adaptation schematic procedure.
Baseline (open-loop pointing)
Exposition to left-deviating prisms (manuel pointing)
After-effects (open loop pointing)
The adaptation procedure involved the participants having to wear prismatic goggles that produced a 10° leftward shift of the visual wide-field that is in the opposite direction to Rossetti et al. (1998) [
The first analysis was to test whether healthy subjects and patients correctly adapted to left-deviating prisms. We carried out
In order to evaluate the presence of an amelioration of left neglect symptoms after prism adaptation, an analysis of variance with repeated measure (ANOVA) was performed on each neuropsychological test, using sessions (pre, post, late) as factor. Hence, for a specified test, the null hypothesis (
For the healthy controls, a significant displacement of the straightahead pointings to the right was observed after exposure to left deviating prisms without significant modification of the performances on the attentional paper and pencil tests.
For the neglect patients, no significant effect of prism exposure was observed neither on the straight-ahead pointing task nor on the neuropsychological tests.
Controls. Before left-deviating prism adaptation (pre), the group analysis showed that the mean end-position of 7 straightahead pointing trials was shifted 1.3 degrees to the right of the body midline (range: −2.3° to 5.1°). After prism adaptation (post), the mean deviation was significantly displaced to the right (mean position after prism adaptation: 5.8 degrees to the right of the body midline; range: 0.7° to 9.0°). Comparison between trials performed before and after prism adaptation was significant (
Straighthead pointing movements before and after prism adaptation for healthy controls (left) and neglect patients (right). (a) For each subject, the average end-position of straightahead pointing movements is represented before (pre) and after (post) left-deviating prism adaptation. Deviation from body midline is displayed in degrees of angle in positive value for right deviation and negative value for left deviation. Numbers refer to patient’s identification (cf. Table
Patients. Before left-deviating prism adaptation (pre), the group analysis showed that the mean end-position of 7 straightahead pointing trials was shifted 3.7 degrees to the right of the body midline (range: 2.0° to 4.9°). After prism adaptation (post), the mean end-position was unchanged (mean: 3.7 degrees to the right; range: 2.4° to 4°). Comparison between trials performed before and after prism adaptation was not significant (
None of the paper and pencil attentional tests have been significantly modified by left-deviating prisms in the control group. The 95% confident interval of healthy subjects’ performances is displayed on Figure
For the neglect group, numerical results are reported for each test in the following section and in Figure
An average of 36.4 lines were cancelled before prism adaptation, 33.8 immediately after prism adaptation, and 35.6 two hours later (standard error of mean = 3.3). Analysis of variance showed no significant difference between sessions,
In the “pop-out” subtest, a mean of 11.2 targets balloons were crossed before prism adaptation, 9.8 immediately after, and 15.2 two hours later (standard error of mean = 3.0). Analysis of variance showed no significant difference between sessions
In the “search” subtest, a mean of 8.4 circles were crossed before prism adaptation, 8.0 immediately after prism adaptation, and 8.2 two hours later (standard error of mean = 2.1). Analysis of variance showed no significant difference between sessions
Before prism adaptation, patients bisected lines with a mean deviation calculated at 50.3 percent on the right of the true centre; this deviation was 35.8 immediately after prism adaptation and 39.5 two hours later (standard error of mean = 9.0). Analysis of variance showed no significant difference between sessions
Before prism adaptation, an average of 3.6 of the five items was copied and an average of 2.2 items was symmetrically copied. Immediately after prism adaptation, an average of 4 items was copied and an average of 2.4 items was symmetrically copied. Two hours after prism adaptation, an average of 2.4 items was copied and an average of 1.6 items was symmetrically copied. Standard error of mean was 0.64 for the total number of items copied and 0.91 for items symmetrically copied. Analysis of variance showed no significant difference between sessions both for the total number of items copied
The daisy was moderately asymmetrical before prism adaptation (mean = 2), immediately after (mean = 2.2), and two hours later (mean = 2). Standard error of mean for this test was 0.34. Analysis of variance showed no significant difference between sessions
Before prism adaptation, an average of 13.5 words were omitted, 8.5 immediately after prism adaptation, and 8.0 two hours later (standard error of mean = 6.7). Analysis of variance showed no significant difference between sessions
The present study showed that patients with left spatial neglect are not affected by prism adaptation to a leftward optical shift. Indeed, neither the rightward deviation of straightahead pointing nor left spatial neglect, as assessed by a battery of classical paper and pencil tests, has been significantly improved or modified after a single session of visuomotor adaptation to left-deviating prisms. Not only do these results suggest that there is a directional specificity of the prisms, but they also show that no cognitive effects are found in the absence of adaptation. The present results play against the hypothesis that active exposure to a simple modification of sensori-motor coordinates is sufficient to reduce left spatial neglect. The short duration of the adaptation procedure cannot explain independently the absence of sensorimotor after-effects given that healthy controls adapt to prisms with the same procedure and neglect patients show sensori-motor after-effects, even larger than controls, when exposed to right-deviating prisms during the same amount of time [
In our experiment, none of the 5 neglect patients showed a consistent sensori-motor adaptation to left-deviating prisms. A similar result was already reported in experiment 1 of the original research performed by Rossetti et al. [
These results suggest that patients with left spatial neglect after right-brain damage are not able to adapt to left-deviating prisms whereas they are able to adapt to right-deviating prisms. This result contrasts with the finding of Weiner et al. [
Alternatively, it is possible that visual realignment after leftward-deviating prisms critically requires the integrity of the right hemisphere in contrast to visual realignment after rightward deviating prisms. As regard to this hypothesis, it is interesting to consider the directional asymmetry for visual after-effects observed in healthy subjects after a visual adaptation to leftward versus rightward prism displacement [
Our results showed for the first time that left-deviating prisms had no effect on various symptoms of left spatial neglect. Previous studies have reported a similar lateralized specificity of prism adaptation on several neglect-related symptoms. Tilikete et al. [
Hence, these results favour a specificity of prism adaptation in terms of the direction of optical shift: only adaptation to right-deviating prisms can improve left spatial neglect. The most obvious explanation to account for these results is related to the absence of adaptability to left-deviating prism for patients with left spatial neglect (cf. Section
Interestingly, the cognitive effects of prism, in non-brain-damaged subjects, are also supported by an asymmetrical pattern of performance. Colent et al. [
Striemer et al. [
The neural substrate underlying the therapeutic effect of this method remains to be fully elucidated. Our study was not specifically designed to deal with this issue but argues at least for an initial lateralized bottom-up activation implicated in the detection of the right visual error during the first pointing movements through prisms. In a recent functional imaging study performed on healthy subjects, we used event-related fMRI to analyze dynamic changes in brain activity during a prolonged exposure to visual prisms [
The mechanism by which such lateralized sensori-motor plasticity induced by prism adaptation can improve spatial neglect remains unclear. Moreover, the gap might be important between what we know about sensori-motor plasticity in normal subjects and what happens in brain-damaged neglect patients.
In a functional imaging PET study, we investigated the anatomical substrates underlying the beneficial effect of prism adaptation in five patients with left spatial neglect following right stroke [