Brain, Vol. 123, No. 4, 746-758,
April 2000
© 2000 Oxford University Press
Relationship of lesion location to cognitive outcome following microelectrode-guided pallidotomy for Parkinson's disease
Support for the existence of cognitive circuits in the human pallidum
1 Department of Neurosurgery, University of Utah, Salt Lake City, UT, USA, 2 Department of Psychology, The Toronto Hospital, 3 Department of Surgery, University of Toronto, Division of Neurosurgery, The Toronto Hospital, 4 Department of Psychology, York University, 5 Movement Disorders Centre, The Toronto Hospital, 6 Department of Medicine, University of Toronto, Division of Neurology, The Toronto Hospital, 7 Departments of Anatomy and Cell Biology and of Psychology, University of Toronto, Toronto, Ontario, Canada
Correspondence to:
Dr Wendy J. Lombardi, Department of Neurosurgery, University of Utah, 50 North Medical Drive, Salt Lake City, UT 84132, USA
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Current models of basal ganglia anatomy posit the existence of multiple parallel, anatomically segregated circuits. Anatomical data from non-human primates suggest that the circuits subserving motor functions are segregated from those subserving cognitive functions. Here we present data that demonstrate that, in humans, motor and cognitive frontosubcortical circuits are segregated. We studied a group of patients with Parkinson's disease undergoing surgical lesioning of the globus pallidus internus for relief of their symptoms. Lesion location along an anteromedial-to-posterolateral axis was found to be related to postsurgical outcome on both cognitive and motor measures. Performance on several neuropsychological measures, including the generation of category exemplars and continuous mental addition, was linearly related to distance along this axis, with anteromedial lesions leading to postsurgical impairment, intermediate lesions having little effect and posterolateral lesions leading to an improvement on several measures. The same relationship was found between memory performance under conditions of proactive interference and lesion location within the globus pallidus internus. In contrast, bradykinesia, assessed as the speed of finger-tapping, had a non-linear relationship to lesion location, intermediate lesions leading to greater postsurgical improvement than lesions in more extreme anteromedial or posterolateral locations. These data demonstrate that the cognitive effects of pallidotomy can be dissociated from the motor effects. These effects depend upon the placement of the lesions within the globus pallidus internus, supporting the segregation of functionally distinct circuits in the human pallidum.
Parkinson's disease; pallidotomy; globus pallidus; cognition; neuropsychology
CALT = Conditional Associative Learning Test; CVLT = California Verbal Learning Test; GDS = Geriatric Depression Scale; GPi = globus pallidus internus; PASAT = Paced Auditory Serial Addition Test; UPDRS = Unified Parkinson's Disease Rating Scale
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Posteroventral pallidotomy has now become well-established as a treatment for advanced Parkinson's disease. Numerous prospective studies have demonstrated that surgical lesioning of the globus pallidus internus (GPi) produces pronounced and lasting effects on the symptoms of Parkinson's disease (Baron et al., 1996
; Lang et al., 1997; Samuel et al., 1998). Electrical stimulation of the globus pallidus has been shown recently to produce similar effects (Siegfried and Lippitz, 1994; Gross et al., 1997; Pahwa et al., 1997; Galvez-Jimenez et al., 1998; Ghika et al., 1998). To various degrees, both treatments ameliorate the cardinal signs of Parkinson's disease, including bradykinesia, rigidity, tremor, and postural instability and gait disturbance, as well as levodopa-induced dyskinesias. As a result of these benefits, pallidotomy has become a widely used and accepted treatment for patients with advanced disease for whom medication is suboptimally effective and associated with disabling side effects.
The cognitive effects of posteroventral pallidotomy, although addressed in a number of recent papers, remain a controversial issue. Several early reports of neuropsychological sequelae in patients following pallidotomy were unable to demonstrate consistent changes (Baron et al., 1996; Soukup et al., 1997). Three recent studies, however, have demonstrated significant cognitive changes after pallidotomy or GPi stimulation on several tests, including letter-cued fluency, category-cued fluency and visuospatial construction (Troster et al., 1997; Scott et al., 1998). Trepanier and colleagues found both general and hemisphere-specific effects after unilateral, microelectrode-guided pallidotomy. Following either left or right pallidotomy, improvement occurred on a test of serial mental addition, and performance declined on a test of digit span. Performance on verbal learning and letter-cued fluency declined following left-sided lesions, and visuospatial construction declined following right-sided lesions (Trepanier et al., 1998).
Overall, studies of the cognitive sequelae of pallidotomy have hitherto produced variable results, ranging from no effects to an array of different consequences, with no consistent overlap in the measures or cognitive processes affected. The reason for these inconsistencies is not clear, inasmuch as there is remarkable consistency among studies (e.g. age of patients, postsurgical interval) and at least partial overlap in the tests administered. A more compelling explanation would be that variability in cognitive outcomes following pallidotomy is related to the precise side and location of lesions within the GPi. Current models of basal ganglia anatomy posit multiple parallel, segregated circuits, which can be localized to anatomically discrete regions within these basal ganglionic structures (Fig. 1) (Alexander et al., 1986; Hoover and Strick, 1993). Therefore, the effects of GPi lesions on neuropsychological tests may depend upon whether there is encroachment upon the circuits which play a role in cognition. This possibility receives support from recent evidence that the motor effects of pallidotomy depend upon the location of the lesion within the GPi. Gross and colleagues demonstrated that lesions placed anteromedially within the sensorimotor area of the GPi resulted in the greatest improvement of drug-related dyskinesias and rigidity, whereas lesions placed posterolaterally resulted in the greatest decrease in tremor (Gross et al., 1999a). Centrally placed lesions were more effective in ameliorating bradykinesia, as well as postural instability and gait disturbance. Two groups investigating the effects of pallidal stimulation on motor symptoms of Parkinson's disease have also reported location effects. In both instances, differences were found along the dorsal–ventral axis: stimulation at ventral locations tended to decrease rigidity and dyskinesias and to increase akinesia, whereas stimulation at dorsal locations tended to decrease akinesia and increase dyskinesias (Bejjani et al., 1997; Krack et al., 1998). If distinct frontosubcortical circuits subserving cognitive functions exist within the basal ganglia, and are distinct from motor circuits, the location of lesions within the GPi should also affect the results of neuropsychological performance following surgery. We now extend our earlier analyses of lesion location to an examination of the relationship between the location of lesions within the GPi and cognitive performance following pallidotomy.
|
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Subjects
The patient group consisted of 26 patients from the original 40 participants in a prospective series examining the effects of microelectrode-guided GPi pallidotomy on Parkinson's disease; this group of patients has been described in detail previously (Lozano et al., 1996
; Lang et al., 1997; Gross et al., 1999a, b). The patients in this study include the entire set of patients in the initial series for whom both postoperative volumetric MRI and neuropsychological data (presurgical and 3 or 6 months postsurgical) existed. All 26 patients underwent neurological assessments, including the Unified Parkinson's Disease Rating Scale (UPDRS; Langston et al., 1992; Lang et al., 1995) and modified Dyskinesia Rating Scale (Goetz et al., 1994), prior to surgery and 6 months after surgery in the unmedicated state (practically defined `off' period) and the medicated state (best `on' period). All subjects gave informed consent to participation in the study, for which approval was given by the Toronto Hospital Committee for Research on Human Subjects.
Volumetric MRI lesion characterization
Volumetric magnetic resonance imaging was performed on all available patients from the prospective series (33/40 patients) (Gross et al., 1999b). The postoperative volumetric MRI scans were obtained at various intervals (1 day to 49 months) following pallidotomy. The remaining patients could not participate due to illness, inconvenience or the implantation of a contralateral neurostimulation device. Image processing was performed using a reformatting program (Silhouette; ISG, Canada) on an UltraSPARC workstation (Sun Microsystems, Palo Alto, Calif., USA) to allow characterization of lesion location and volume, as previously described (Gross et al., 1999b). Briefly, volumetric MRIs were reformatted to correct for head rotation, twist and tilt prior to generating triplanar data sets parallel and perpendicular to the intercommissural line. The position of the centre of the lesion with respect to the conventional commissural landmarks, as well as to the wall of the third ventricle, was then measured. Measurements were normalized to an intercommissural distance of 23 mm.
Significant variability in the position of the lesion centre existed in both the anterior–posterior dimension (with respect to the anterior commissure) and the medial–lateral dimension (with respect to the third ventricle wall), whereas variability in the dorsal–ventral dimension (with respect to the intercommissural line) was relatively small (Gross et al., 1999b). Anterior–posterior and medial–lateral positions were correlated, such that lesions were distributed from anteromedial to posterolateral within the posteroventral region of the GPi, parallel to the internal capsule. As described previously (Gross et al., 1999b), the location of each lesion along the axis from anteromedial to posterolateral was determined by first projecting lesions at all dorsal–ventral positions onto a common axial plane. Next, a regression line through the lesions was generated and perpendicular lines were drawn from the lesions to the regression line. The distances along the regression line where these lines intersected it were measured and denoted distanceAM–PL, 0 being the most anteromedial.
Neuropsychological assessment
Subjects underwent extensive neuropsychological testing prior to surgery and following surgery at 3 months, 6 months, or both 3 and 6 months. In the event that neuropsychological testing was conducted at both postsurgical time points, the data collected at the 6-month time point was used in these analyses. The 6-month data were used whenever possible because they were available for a greater number of subjects than were the 3-month data, and it was felt that at the 6-month time point patients would be more stable and less likely to display temporary, general effects of surgery. Due to time constraints, not all tests were completed with all subjects at both the pre- and postpallidotomy time points. The numbers of subjects contributing to each analysis are given in Table 2.
|
Both pre- and postsurgical neuropsychological testing was performed in the medicated state (i.e. in the `on' period). The results of this outcome study have been reported previously (Trepanier et al., 1998
). Additional `off'-period testing was not practical because of the length of the test battery and the lack of validity of testing repeated at very short intervals. A subset of the tests included in the battery was selected for the current analyses on the basis of their sensitivity to damage to the frontal lobes, including both prefrontal cortex and motor areas. Frontal lobe function was analysed selectively on both theoretical and empirical grounds: (i) damage to basal ganglionic structures frequently produces deficits on frontal lobe tasks (for reviews, see Brown and Marsden, 1990; Dubois et al., 1991), and (ii) current models of basal ganglia function, in which information processing occurs in multiple cortical–basal ganglionic–thalamocortical loops, predict that frontal lobe processes will be affected by lesions within the globus pallidus (Alexander et al., 1986; Hoover and Strick, 1993). A test of depression was also included, to provide a check on the degree to which cognitive changes could be attributed to changes in mood. The tests included in these analyses are listed in Table 1.
|
Statistical analyses
Two sets of analyses were conducted. The first set of analyses compared the overall pre- and postpallidotomy test scores of the entire group of patients on each measure. The second set of analyses examined the relationship between lesion location and clinical outcome on each measure. Bonferroni correction was used to control the family-wise experimental error rate at 0.10 for the set of 13 cognitive measures, so that the level of statistical significance was set at P < 0.008 for these analyses.
Hierarchical multiple regression analyses were used to examine the relationship of lesion location to neuropsychological outcome, using the method reported previously (Gross et al., 1999a). The dependent variables in these analyses were the postoperative scores on each measure in the neuropsychological test battery. In each hierarchical analysis, independent variables were entered in a step-wise manner, in an order determined by logical priority. In this way, the effects of previously entered variables (such as preoperative score) could be controlled for when examining the effects of variables entered subsequently (e.g. lesion location) on neuropsychological measures. The independent variables and the order in which they were entered into the hierarchical regression analyses were as follows. (i) Preoperative score: the effect of preoperative score on the postoperative score was evaluated first. (ii) Volume: the amount of variance accounted for by the lesion volume, calculated previously (Gross et al., 1999b), was determined next. The likelihood of a lesion ablating a critical structure is a combination of its location and size. Since variance in outcome scores will be affected by each of these parameters, both must be accounted for. (iii) DistanceAM–PL: the variance in the outcome scores accounted for by lesion location from anteromedial to posterolateral along the posterior region of the GPi was determined next. (iv) DistanceAM–PL2 : to allow for the possibility of non-linear relationships between lesion position and outcome measures (e.g. that centrally located lesions may be more or less efficacious than anteromedial or posterolateral lesions), the final step of the multiple regression analysis was performed using quadratic regression. This was accomplished by including the second-order term, the square of the distance.
At each step in the hierarchical analysis, the increment in the proportion of the variance accounted for by the entered independent variable over that accounted for by all of the previously entered variables was determined; this represents the squared semipartial correlation coefficient (sr2). The significance of this relationship was then ascertained. The significance of the quadratic relationship was determined by assessing the combined variance accounted for by the first-order term (distanceAM–PL) and the second-order term (distanceAM–PL2) over that accounted for by preoperative score and lesion volume.
Because of the difficulty of representing the results of a multiple regression analysis graphically, the percentage change from the baseline score is shown as a function of linear distance along the anteromedial-to-posterolateral axis of lesion distribution. However, curve-fitting in these graphs was performed on the basis of the previous multiple regression analysis. If a linear relationship was observed (i.e. a significant percentage of variance was accounted for by the linear regression with distance), then a linear regression line was superimposed on the graph; if a non-linear relationship was observed (i.e. a significant percentage of the variance was accounted for by the quadratic regression with distanceAM–PL combined with distanceAM–PL2 but not with distanceAM–PL alone), then a quadratic linear regression line was superimposed. Percentage change scores were not used for any statistical analyses, however.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Comparison of pre- and postsurgical neuropsychological test scores
Cognitive measures
Mean pre- and postsurgical test scores on each of the tests included in the battery are shown in Table 2
. Letter-cued fluency was significantly changed from its preoperative baseline [t(25) = 3.18, P < 0.005]. The number of words generated to an initial letter cue was significantly lower following pallidotomy than it was prior to surgery. Following the Bonferroni correction, there was a trend for digit span to be lower after pallidotomy than it was before surgery [t(25) = 2.13, P < 0.05]. The ratio of the backward to the forward digit span score was significantly lower following pallidotomy [t(25) = 3.72, P < 0.001], indicating that the backward span was selectively affected by surgery. These latter two comparisons suggest that working memory is adversely affected by pallidotomy. One test of recent verbal memory from the California Verbal Learning Test (CVLT), memory for a new list of words after five learning trials on a previous list, was significantly lower following pallidotomy than it was prior to surgery [t(24) = 3.04, P < 0.007]. However, no other measure of verbal memory, including learning, clustering during recall, delayed recall or delayed recognition of the first list, approached significance. Isolated effects on the second study list, in the absence of effects on the first list, can be interpreted as proactive interference effects, the effects of earlier learning on new learning. Pallidotomy therefore appeared to increase the effects of proactive interference on verbal memory but had no impact on other aspects of memory.
The difference between pre- and postsurgical scores did not approach significance for either of the measures of set-shifting, Trail Making or Odd Man Out, or for the Conditional Associative Learning Test (CALT).
Depression
The difference between pre- and postsurgical scores was significant for the Geriatric Depression Scale (GDS) [t(25) = 2.95, P < 0.01]. Scores were lower, indicating fewer depressive symptoms following surgery.
Motor performance
One of the measures of motor functioning, the Purdue Pegboard test performed with the hand contralateral to the side of pallidotomy, was significantly improved following surgery [t(23) = -2.09, P < 0.05]. No overall change in ipsilateral Purdue Pegboard performance or either ipsilateral or contralateral finger tapping was found.
Relationship of lesion side to clinical outcome
We have reported previously on the relationship of lesion side to cognitive outcome following pallidotomy (Trepanier et al., 1998). To rule out the possibility that lesion side might account for any observed relationship of lesion location to cognitive outcome, we first determined whether any of the postsurgical test score variance in this subset of patients was attributable to the side of the lesion. A multiple regression analysis was therefore conducted with prescore entered on the first step and lesion side entered on the second step. Four of the subjects in this analysis were left-handed. These subjects were treated in the same manner as right-handers in this analysis, since the majority of left-handers are left hemisphere-dominant for language (e.g. Branch et al., 1964; Strauss and Wada, 1983). Postoperative performance on two of the measures, digit span and the interference trial of the CVLT, had a significant relationship to the side of the lesion (sr2 = 0.084, P < 0.02 for digit span; sr2 = 0.252, P < 0.01 for interference), and performance on both letter-cued fluency and the Purdue Pegboard test in the contralateral hand had a marginally significant relationship to the side of the lesion (sr2 = 0.057, P < 0.06 for letter-cued fluency; sr2 = 0.085, P < 0.06 for the pegboard test). In the case of digit span, left-sided lesions appeared to have little effect on performance (postsurgical performance = 101% of baseline), whereas right-sided lesions appeared to decrease performance (postsurgical performance = 89% of baseline). The opposite pattern of results was obtained for letter-cued fluency, on which right-sided lesions had little effect (postsurgical performance = 97% of baseline), whereas left-sided lesions appeared to decrease performance postoperatively (postsurgical performance = 77% of baseline). Right- and left-sided lesions appeared to have opposite effects on recall of a list of words under conditions of proactive interference, right-sided lesions improving performance (postsurgical performance = 112% of baseline) and left-sided lesions leading to a postoperative decrement in performance (postsurgical performance = 73% of baseline). Both right- and left-sided lesions improved performance on the Purdue Pegboard test in the contralateral hand, although right-sided lesions led to greater improvements (139% of baseline) than did left-sided lesions (125% of baseline).
Relationship of lesion volume to clinical outcome
The relationship of lesion volume to postpallidotomy test scores was assessed using hierarchical regression. The variance in the postsurgical test score that was related to the presurgical test score was first removed, and the degree to which the remaining variance in the postsurgical score was related to lesion volume was determined. The volume of the lesion was not significantly related to postsurgical outcome on any of the measures included in the cognitive assessment battery, accounting for 0–15% of the variance on these tests. The size of the relationship between lesion volume and each cognitive outcome measure is included in Table 3.
|
Relationship of lesion location to clinical outcome
The effect of lesion location on postpallidotomy test scores was assessed after accounting for the effects of prescore and lesion volume. For each measure, the incremental variance in the postsurgical score that was accounted for by lesion location is indicated by the semipartial correlation coefficients for distanceAM–PL (Table 3
). To allow for the possibility of a non-linear relationship between lesion location and clinical outcome, the incremental variance accounted for by adding both distanceAM–PL and its square was also assessed.
Cognitive measures
Category-cued fluency, the Paced Auditory Serial Addition Test (PASAT) and digit span were linearly related to distance of the lesion along the anteromedial-to-posterolateral axis. The relationship between distanceAM–PL and category-cued fluency was significant when the Bonferroni correction for multiple tests was applied. As illustrated in Fig. 2, lesions within the anteromedial region of the axis led to a decrement in postsurgical performance compared with the presurgical baseline, whereas lesions within the posterolateral region of the axis led to a postsurgical improvement in performance. Lesions falling within the mid-range of the axis did not appear to lead to a change in performance. Although not statistically significant when the Bonferroni correction was applied, the trend in the relationship of lesion location to PASAT performance was in the same direction as that found for category-cued fluency (Fig. 3). In both cases, better postpallidotomy performance occurred with more posterolaterally placed lesions.
|
|
The opposite pattern occurred in the relationship of digit span to lesion location, more anteromedially placed lesions resulting in better postsurgical performance. Because postsurgical performance on digit span was found to have a significant relationship to the side of the lesion (see above), a second multiple regression analysis was conducted, with preoperative score, lesion volume, lesion side and distanceAM–PL entered as independent variables. In this analysis, lesion side had a significant relationship to postsurgical performance (sr2 = 0.089, P < 0.01). DistanceAM–PL did not have a significant relationship to performance when the effects of lesion side were partialled out (sr2 = 0.019, P < 0.20).
No significant relationship between lesion location and letter-cued fluency or visual span was found. The addition of the distanceAM–PL2 term did not add significantly to the variance accounted for by any of the tests of verbal fluency or working memory (Table 3).
Learning of a categorized list of words on the CVLT was not related to distanceAM–PL, nor was a relationship detected when the distanceAM–PL2 term was added to the analysis. Delayed recall and recognition were also not related to lesion location. However, a significant linear relationship was detected between the learning of an unrelated second list of words and distanceAM–PL, more posteriorly located lesions leading to better recall of this list (Fig. 4). Lesion location within the GPi was therefore related to proactive interference in memory. DistanceAM–PL2 did not add a significant amount of variance to that accounted for by the linear relationship. Given the relationship that was found between lesion side and postoperative memory interference, a second regression analysis was conducted with preoperative score, lesion volume, lesion side and distanceAM–PL entered as independent variables. The relationship between postoperative memory interference and distanceAM–PL remained significant when the effects of lesion side were partialled out (sr2 = 0.128, P < 0.03).
|
Performance on tests of set-shifting was not related to lesion location in this group of patients. DistanceAM–PL did not explain a significant amount of variance in postsurgical performance of either of the two tests of mental set-shifting included in the battery, Trail Making or Odd Man Out. Performance on the CALT was also not linearly related to lesion location along the anteromedial-to-posterolateral axis. The second-order term, distanceAM–PL2, did not add significantly to the amount of variance accounted for by the regression equation.
Depression
Postpallidotomy scores on the GDS did not appear to be related to the location of the lesion. The addition of the second-order term distanceAM–PL2 to the regression equation did not significantly increase the amount of variance accounted for, ruling out the possibility of a non-linear relationship between depression and lesion location.
Motor functions
Two aspects of motor functioning, finger-tapping speed and manual dexterity, were included in the battery. Trail Making Part A provided a measure of psychomotor speed. Finger-tapping speed in the hand contralateral to the lesion was not significantly related to distanceAM–PL. However, it was related to the combination of the first- and second-order distance terms, distanceAM–PL + distanceAM–PL2, indicating that contralateral tapping speed had a non-linear relationship with location of the lesion along the anteromedial-to-posterolateral axis. Ipsilateral finger-tapping speed was not related to either lesion location or the combined variables distanceAM–PL + distanceAM–PL2.
Speed of performance on a test of manual dexterity of the contralateral hand did not have a significant linear or non-linear relationship to lesion location. Speed of performance on this task in the ipsilateral hand also did not have a significant relationship with lesion location, although there was a non-significant trend towards a curvilinear relationship. Speed of performance on Trail Making Part A did not have a relationship to lesion location along the anteromedial-to-posterolateral axis.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
The results of the present study demonstrate that surgical ablation of the GPi in Parkinson's disease patients has effects on several aspects of neuropsychological test performance. These effects were of several types, and are summarized in Table 4
. First, a uniform postsurgical effect across the entire cohort of patients was noted on two measures, the ratio of the backward to forward digit span and GDS. Pallidotomy appeared to affect backward digit span detrimentally relative to forward digit span, and this effect occurred regardless of the location of the lesion. Pallidotomy had a beneficial effect on symptoms of depression, and this effect also appeared to be uniform across lesion locations.
|
Effects of the second type were interhemispheric, postsurgical change occurring to a larger degree following left- or right-sided lesions; they were noted on four measures in the cognitive battery, including digit span, letter-cued fluency, recall of a word list under conditions of proactive interference, and the Purdue Pegboard test in the contralateral hand. In the case of digit span, postsurgical performance was decreased following right-sided lesions, but was essentially unchanged following left-sided lesions. In contrast, letter-cued fluency was essentially unchanged following right-sided lesions, and was worse following left-sided lesions. The effect of lesion side on letter-cued fluency was only marginally significant in these analyses, but was the same as that obtained by Trepanier and colleagues in the larger population of patients of which this group is a subset (Trepanier et al., 1998
). Proactive interference in recent memory was worse postsurgically in the case of left-sided lesions, and appeared slightly improved following right-sided lesions. Performance on the Purdue pegboard test with the contralateral hand improved postsurgically, and there was a trend towards greater improvement following right-sided lesions than left-sided lesions.
The third type of effect was an intrahemispheric lesion location effect, the degree and/or direction of postsurgical change depending upon the distance of the lesion along the anteromedial-to-posterolateral axis of the GPi (Table 4). Intrahemispheric lesion location, measured as distanceAM–PL within the GPi, was found to have an effect on several cognitive measures in the battery, including category fluency, the PASAT and recall under conditions of interference from previous learning. On all three measures, lesion location was found to have a linear relationship with performance. More anteromedially placed lesions were associated with worsening of performance postsurgically, whereas more posterolaterally placed lesions were associated with improvement in performance. In the case of the PASAT, however, this conclusion must be made with caution. Prepallidotomy PASAT data were not obtained for eight subjects, and although this loss of data was random, a number of these subjects had anteromedial lesions. The range of lesion locations in the analysis of the PASAT data is therefore restricted, making it difficult to draw conclusions about the effects of anteromedial lesion locations. No relationship was found between lesion location and recall of a word list without prior interfering learning. Isolated effects on the second study list, in the absence of effects on the first list, can be interpreted as proactive interference effects, the effects of earlier learning on new learning. An anteromedial lesion location, therefore, appeared to result in greater proactive interference effects in verbal memory than did more posterolateral lesion locations. In the absence of interference, however, learning and memory did not appear to be affected by lesion location.
The measures of set-shifting included in these analyses were not affected by pallidotomy. No evidence of a group outcome effect or of a relationship between lesion location and postsurgical performance was detected. This result is surprising, given that Parkinson's disease and other degenerative diseases affecting the basal ganglia typically produce impairments on these tests. One possible reason for this negative result is that the measures of set-shifting included in these analyses are among the easiest of these tests, and may therefore have been insensitive to changes following pallidotomy. It is possible that examination of the relationship between lesion location and performance on tests of set-shifting and mental flexibility other than those included in the present battery would have demonstrated that pallidotomy does alter these processes. However, other reports of cognitive functioning following pallidotomy have also failed to find group outcome effects on measures of executive functioning (Baron et al., 1996; Soukup et al., 1997; Troster et al., 1997; Scott et al., 1998), including two of the more sensitive measures in this category, the Wisconsin Card Sorting Test (Berg, 1948) and the Stroop test (Golden, 1978). It is also possible that presurgical degeneration of the basal ganglia has resulted in the use of compensatory pathways for executive functions, in a manner similar to that suggested by Marsden and Obeso (1994) for motor control, so that lesioning of the pallidum does not result in further decrements in executive functioning (for a discussion, see Trepanier et al., 1998).
The differential effects of lesion location on the various neuropsychological measures included in these analyses may provide interesting hints as to which underlying cognitive processes are affected by pallidal lesions. For example, the finding that lesion location has an effect on proactive interference in memory, but not on clustering in recall, may tell us that frontosubcortical circuits play a role in memory interference but not in conceptual clustering during learning. Similarly, the finding that left-sided lesions, regardless of location, impair letter-cued fluency, whereas anteromedial lesions located in either hemisphere impair category-cued fluency may be providing information about important differences between these two tests. We have avoided speculating here at any length about differences among the tests for several reasons. First, the tests are standard neuropsychological measures, rather than experimental tasks designed to isolate and manipulate individual cognitive processes, a fact that limits the insights that can be gained from such comparisons. Secondly, the lack of an effect of lesion location on some tests may be the result of differences among the tests in sensitivity to disruption, perhaps due to different degrees of difficulty rather than to fundamental differences in cognitive processes. Different patterns of lesion location effects on different cognitive tasks would have provided stronger evidence for dissociations among the tests in the effects of lesion location, but such differences were not observed in these data.
The pre- and postpallidotomy neuropsychological assessments were conducted only in the medicated `on' state. The length of the test battery, as well as the possibility that practice effects would contaminate retesting at short intervals, made it impractical to collect unmedicated `off'-state data. `On'-state testing was selected because of the likelihood that many patients would find the tasks in the cognitive battery too difficult to perform in the `off' state. However, it is possible that `off'-state testing would have revealed neuropsychological effects of pallidotomy which were missed in these analyses. Negative effects of the present study, including the failure to find effects of pallidotomy on mental set-shifting and flexibility, must therefore be interpreted with caution.
One measure of motor functioning, the speed of finger tapping on the side contralateral to the pallidotomy, was associated with the location of the lesion along the anteromedial-to-posterolateral axis in the GPi. The form of the relationship was curvilinear, lesions in an intermediate location having a more beneficial effect on the speed of finger-tapping than lesions located in the more extreme anteromedial or posterolateral locations. This finding is consistent with our earlier demonstration that intermediate lesion locations are associated with the most beneficial effects on ratings of bradykinesia on the UPDRS (Gross et al., 1999a).
The relationship of lesion location to the speed of finger-tapping after pallidotomy is in stark contrast to that found between lesion location and cognitive performance in this same group of patients. Lesions in the intermediate region of the anteromedial-to-posterolateral axis led to maximal effects on motor measures but produced little effect on any of the cognitive measures. The more extreme anteromedial and posterolateral locations both produced larger, although different, effects on cognitive measures than did the intermediate locations, but led to small effects on the motor measure. These data demonstrate that the effects of pallidotomy on cognition and motor symptoms cannot both be accounted for by its effects on a single variable such as the speed of information processing, because such effects would be expected to produce parallel changes on motor and cognitive variables. These data also provide evidence that the frontosubcortical circuits contributing to cognitive performance can be at least partly dissociated from motor circuits within the human pallidum (Fig. 1). This dissociation between cognitive and motor circuits is consistent with previous anatomical evidence obtained in non-human primates using both tract-tracing and physiological recording methods. Middleton and Strick demonstrated that, in the primate, regions of the GPi which project to cognitive areas of the frontal lobe are segregated from those which project to primarily motor areas (Middleton and Strick, 1994). Following injection of a retrograde tracer into the dorsolateral prefrontal cortex, labelled neurons in the GPi were located in dorsomedial regions, in contrast to the more ventrolateral regions labelled following injection of the primary motor, premotor or supplementary motor cortex. Yoshida and colleagues recorded from single pallidal neurons during electrical stimulation of the frontal lobe. Their data demonstrated that the pallidal neurons receiving inputs from the motor cortex were segregated from the neurons receiving inputs from the prefrontal, premotor and supplementary motor regions (Fig. 1) (Yoshida et al., 1993). Stimulation of the motor cortex influenced the firing of neurons located in the ventrolateral and caudal region of the globus pallidus, whereas stimulation of the prefrontal cortex, premotor cortex and supplementary motor area influenced neurons located in the dorsomedial and rostral region. The data reported here are the first demonstration that cognitive and motor circuits can be dissociated in the human pallidum.
A striking aspect of the data on cognitive sequelae of pallidotomy is that performance can be either improved or impaired following surgery, depending upon the placement of the lesion. In comparison with presurgical performance, category-cued fluency and PASAT performance appear to be improved and proactive interference in memory to be decreased following posterolateral lesions. In contrast, category-cued fluency and PASAT performance are impaired and proactive interference increased following anteromedial lesions. As noted above, previous research has demonstrated that frontosubcortical circuits which include cognitive areas of the frontal lobe are located in dorsomedial areas within the pallidum. It is therefore not surprising that lesions in the anteromedial region of the anteromedial-to-posterolateral axis tend to produce the largest postsurgical decrements in cognitive performance. The reason for the observed improvements in several aspects of cognitive performance following posterolateral lesions is not clear. One possibility is that the improvements are the result of the beneficial effects of pallidotomy on motor performance. Although the pattern of improvements in cognitive performance differs from that of all of the motor symptoms except for tremor (Gross et al., 1999a), it is possible that these data reflect the combined effects of lesions in both the cognitive and motor circuits: as lesions move away from the anteromedial extreme of the axis within the GPi, the direct detrimental effects of lesions within the cognitive circuits may gradually lessen. At the same time, the indirect benefits of improvements in bradykinesia and rigidity are greatest in the intermediate portions of the axis and should gradually lessen as lesions move towards the extreme ends of the axis. In combination, these effects may result in a net decrement in performance in the anteromedial region, no net effect in the intermediate region and a net improvement in performance in the posterolateral region. Alternatively, it is possible that lesions within the circuits of the posterolateral region of the pallidum, although thought to be primarily motor, have direct beneficial effects on cognition which are not presently understood. For example, it is possible that the supplementary motor area plays a role in working memory that has been unappreciated previously.
A possible alternative explanation of our results is that the lesions which produce maximal cognitive effects are those which encroach upon another structure, such as the globus pallidus externus, rather than upon a cognitive circuit within the GPi itself. This possibility seems unlikely, however, for two reasons. First, medial lesions are not more likely to impinge upon the globus pallidus externus than laterally placed lesions. Secondly, if encroachment upon another structure is the critical factor in producing cognitive effects, lesion volume would be expected to be associated with cognitive outcome, inasmuch as large lesions would be more likely to do so. However, no relationship of lesion volume with performance was found for any of the cognitive measures.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
In the present study, the cognitive effects of pallidotomy were found to depend in part upon the location of the lesion within the GPi. Performance on several neuropsychological measures, including the generation of category exemplars and continuous mental addition, was linearly related to distance along the anteromedial-to-posterolateral axis, anteromedial lesions leading to postsurgical impairment, intermediate lesions having little effect and posterolateral lesions leading to an improvement on several measures. The same relationship was found between memory performance under conditions of proactive interference and lesion location within the GPi. In contrast, the effects of pallidotomy on the motor symptom of bradykinesia, assessed as the speed of finger-tapping, had a non-linear relationship with lesion location, intermediate lesions leading to greater postsurgical improvement than lesions in more extreme anteromedial or posterolateral locations. The relationship between lesion location and cognitive effects may provide a partial explanation for the large degree of individual variability in cognitive outcome following pallidotomy that has typically been noted in various studies.
These data provide the first evidence of distinct cognitive frontosubcortical circuits in humans. Although research in non-human primates has provided an anatomical basis for separate cognitive and motor circuits, there has been no demonstration that these physically distinct circuits play unique roles in behaviour. The data reported here demonstrate that regions within the GPi which can be predicted on anatomical grounds to play differing roles in behaviour do, in fact, make unique contributions. The demonstration of a dissociation between the relationship of lesion location to the cognitive and motor effects of pallidotomy supports the segregation of functionally distinct circuits in the human pallidum.
|
Acknowledgments |
|---|
The work presented in this paper was supported by the Parkinson's Foundation of Canada grants to A.M.L., A.E.L. and J.A.S. and a Center of Excellence grant from the Parkinson Foundation to A.E.L.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsReferences |
|---|
Alexander GE, DeLong MR, Strick PL. Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annu Rev Neurosci 1986; 9: 357–81.[Web of Science][Medline]
Baron MS, Vitek JL, Bakay RAE, Green J, Kaneoke Y, Hashimoto T, et al. Treatment of advanced Parkinson's disease by posterior GPi pallidotomy: 1-year results of a pilot study. Ann Neurol 1996; 40: 355–66.[Web of Science][Medline]
Bejjani B, Damier P, Arnulf I, Vonnet AM, Bidailhet M, Dormont D, et al. Pallidal stimulation for Parkinson's disease: two targets? Neurology 1997; 49: 1564–9.
Berg EA. A simple objective technique for measuring flexibility in thinking. J Gen Psychol 1948; 39: 15–22.[Medline]
Branch C, Milner B, Rasmussen T. Intracarotid sodium amytal for the lateralization of cerebral speech dominance. J Neurosurg 1964; 21: 399–405.[Web of Science][Medline]
Brink TL, Yesavage JA, Lum O, Heersema PH, Adey M, Rose TS. Screening tests for geriatric depression. Clin Gerontol 1982; 1: 37–43.
Brown RG, Marsden CD. Cognitive function in Parkinson's disease: from description to theory. [Review]. Trends Neurosci 1990; 13: 21–29.[Web of Science][Medline]
Delis DC, Kramer JH, Kaplan E, Ober BA. California Verbal Learning Test: Adult Version. San Antonio (TX): Psychological Corporation; 1987.
Dubois B, Boller F, Pillon B, Agid Y. Cognitive deficits in Parkinson's disease. In Boller F, Grafman J, editors. Handbook of neuropsychology, Vol. 5. Amsterdam: Elsevier; 1991. p. 195–240.
Flowers KA, Robertson C. The effect of Parkinson's disease on the ability to maintain a mental set. J Neurol Neurosurg Psychiatry 1985; 48: 517–29.
Galvez-Jimenez N, Lozano A, Tasker R, Duff J, Hutchison W, Lang AE. Pallidal stimulation in Parkinson's disease patients with a prior unilateral pallidotomy. Can J Neurol Sci 1998; 25: 300–5.[Web of Science][Medline]
Ghika J, Villemure JG, Fankhauser H, Favre J, Assal G, Ghika-Schmid F. Efficiency and safety of bilateral contemporaneous pallidal stimulation (deep brain stimulation) in levodopa-responsive patients with Parkinson's disease with severe motor fluctuations: a 2-year follow-up review. J Neurosurg 1998; 89: 713–8.[Web of Science][Medline]
Goetz CG, Stebbins GT, Shale HM, Lang AE, Chernik DA, Chmura TA, et al. Utility of an objective dyskinesia rating scale for Parkinson's disease: inter- and intrarater reliability assessment. Mov Disord 1994; 9: 390–4.[Web of Science][Medline]
Golden CJ. Stroop Color and Word Test. Chicago (IL): Stoelting; 1978.
Gronwall DM. Paced auditory serial-addition task: a measure of recovery from concussion. Percept Mot Skills 1977; 44: 367–73.[Web of Science][Medline]
Gross C, Rougier A, Guehl D, Boraud T, Julien J, Bioulac B. High-frequency stimulation of the globus pallidus internalis in Parkinson's disease: a study of seven cases. J Neurosurg 1997; 87: 491–8.[Web of Science][Medline]
Gross RE, Lombardi WJ, Lang AE, Duff J, Hutchison WD, Saint-Cyr, JA, et al. Relationship of lesion location to clinical outcome following microelectrode-guided pallidotomy for Parkinson's disease. Brain 1999a; 122: 405–16.
Gross RE, Lombardi WJ, Hutchison WD, Narula S, Saint-Cyr JA, Dostrovsky JO, et al. Variability in lesion location following microelectrode-guided pallidotomy: anatomical, physiological and technical factors that determine lesion distribution. J Neurosurg 1999b; 90: 468–77.[Web of Science][Medline]
Hoover JE, Strick PL. Multiple output channels in the basal ganglia. Science 1993; 259: 819–21.
Krack P, Pollak P, Limousin P, Hoffmann D, Benazzouz A, Le Bas JF, et al. Opposite motor effects of pallidal stimulation in Parkinson's disease. Ann Neurol 1998; 43: 180–92.[Web of Science][Medline]
Lang AE, Benabid A-L, Koller WC, Lozano AM, Obeso JA, Olanow CW, et al. The core assessment program for intracerebral transplantation [letter]. Mov Disord 1995; 10: 527–8.[Web of Science][Medline]
Lang AE, Lozano AM, Montgomery E, Duff J, Tasker R, Hutchison W. Posteroventral medial pallidotomy in advanced Parkinson's disease. N Engl J Med 1997; 337: 1036–42.
Langston JW, Widner H, Goetz CG, Brooks D, Fahn S, Freeman T, et al. Core assessment program for intracerebral transplantations (CAPIT). Mov Disord 1992; 7: 2–13.[Web of Science][Medline]
Lozano A, Hutchison W, Kiss Z, Tasker R, Davis K, Dostrovsky J. Methods for microelectrode-guided posteroventral pallidotomy. J Neurosurg 1996; 84: 194–202.[Web of Science][Medline]
Marsden CD, Obeso JA. The functions of the basal ganglia and the paradox of stereotaxic surgery in Parkinson's disease. [Review]. Brain 1994: 117: 877–97.
Middleton FA, Strick PL. Anatomical evidence for cerebellar and basal ganglia involvement in higher cognitive function. Science 1994; 266: 458–61.
Pahwa R, Wilkinson S, Smith D, Lyons K, Miyawaki E, Koller WC. High-frequency stimulation of the globus pallidus for the treatment of Parkinson's disease. Neurology 1997; 49: 249–53.
Reitan RM, Wolfson D. The Halstead–Reitan Neuropsychological Test Battery: theory and clinical interpretation. Tucson (AZ): Neuropsychology Press; 1993.
Samuel M, Caputo E, Brooks DJ, Schrag A, Scaravilli T, Branston NM, et al. A study of medial pallidotomy for Parkinson's disease: clinical outcome, MRI location and complications. Brain 1998; 121: 59–75.
Scott R, Gregory R, Hines N, Carroll C, Hyman N, Papanasstasiou V, et al. Neuropsychological, neurological and functional outcome following pallidotomy for Parkinson's disease: a consecutive series of eight simultaneous bilateral and twelve unilateral procedures. Brain 1998; 121: 659–75.
Siegfried J, Lippitz B. Bilateral chronic electrostimulation of ventroposterolateral pallidum: a new therapeutic approach for alleviating all parkinsonian symptoms. Neurosurgery 1994; 35: 1126–30.[Web of Science][Medline]
Soukup VM, Ingram F, Schiess MC, Bonnen JG, Nauta HJ, Calverley JR. Cognitive sequelae of unilateral posteroventral pallidotomy. Arch Neurol 1997; 54: 947–50.
Spreen O, Strauss E. A compendium of neuropsychological tests, 2nd ed. New York: Oxford University Press; 1998.
Strauss E, Wada J. Lateral preferences and cerebral speech dominance. Cortex 1983; 19: 165–77.[Web of Science][Medline]
Taylor AE, Saint-Cyr JA, Lang AE. Memory and learning in early Parkinson's disease: evidence for a `frontal lobe syndrome'. Brain Cogn 1990; 13: 211–32.[Web of Science][Medline]
Tiffan J. Purdue Pegboard: Examiner Manual. Rosemont (IL): London House; 1968.
Trepanier LL, Saint-Cyr JA, Lozano AM, Lang AE. Neuropsychological consequences of posteroventral pallidotomy for the treatment of Parkinson's disease. Neurology 1998; 51: 207–15.
Troster AI, Fields JA, Wilkinson SB, Pahwa R, Miyawaki E, Lyons KE, et al. Unilateral pallidal stimulation for Parkinson's disease: neurobehavioral functioning before and 3 months after electrode implantation. Neurology 1997; 49: 1078–83.
Wechsler D. Wechsler Adult Intelligence Scale—revised. WAIS-R manual. New York: Psychological Corporation; 1981.
Wechsler D. Wechsler Memory Scale—revised manual. San Antonio (TX): Psychological Corporation; 1987.
Yoshida S, Nambu A, Jinnai K. The distribution of the globus pallidus neurons with input from various cortical areas in the monkeys. Brain Res 1993; 611: 170–4.[Web of Science][Medline]
Received March 29, 1999. Revised September 7, 1999. Second revision on October 25, 1999. Accepted November 1, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. F. Dulay, A. M. Strutt, H. S. Levin, J. Jankovic, E. C. Lai, R. G. Grossman, and M. K. York Depressed Mood and Memory Impairment Before and After Unilateral Posteroventral Pallidotomy in Parkinson's Disease J Neuropsychiatry Clin Neurosci, August 1, 2008; 20(3): 357 - 363. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Pillon, C. Ardouin, K. Dujardin, P. Vittini, A. Pelissolo, O. Cottencin, L. Vercueil, J. -L. Houeto, P. Krystkowiak, Y. Agid, et al. Preservation of cognitive function in dystonia treated by pallidal stimulation Neurology, May 23, 2006; 66(10): 1556 - 1558. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Sage, S. G. Anagnostaras, S. Mitchell, J. M. Bronstein, A. De Salles, D. Masterman, and B. J. Knowlton Analysis of Probabilistic Classification Learning in Patients With Parkinson's Disease Before and After Pallidotomy Surgery Learn. Mem., May 1, 2003; 10(3): 226 - 236. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K R Turner, W G Reid, J Homewood, and R J Cook Neuropsychological sequelae of bilateral posteroventral pallidotomy J. Neurol. Neurosurg. Psychiatry, October 1, 2002; 73(4): 444 - 446. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. A. Middleton and P. L. Strick Basal-ganglia 'Projections' to the Prefrontal Cortex of the Primate Cereb Cortex, September 1, 2002; 12(9): 926 - 935. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Green, W. M. McDonald, J. L. Vitek, M. Haber, H. Barnhart, R. A.E. Bakay, M. Evatt, A. Freeman, N. Wahlay, S. Triche, et al. Neuropsychological and psychiatric sequelae of pallidotomy for PD: Clinical trial findings Neurology, March 26, 2002; 58(6): 858 - 865. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. B. Scott, J. Harrison, C. Boulton, J. Wilson, R. Gregory, S. Parkin, P. G. Bain, C. Joint, J. Stein, and T. Z. Aziz Global attentional-executive sequelae following surgical lesions to globus pallidus interna Brain, March 1, 2002; 125(3): 562 - 574. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Baron, T. Wichmann, D. Ma, and M. R. DeLong Effects of Transient Focal Inactivation of the Basal Ganglia in Parkinsonian Primates J. Neurosci., January 15, 2002; 22(2): 592 - 599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Litvan, M. Phipps, V. L. Pharr, M. Hallett, J. Grafman, and A. Salazar Randomized placebo-controlled trial of donepezil in patients with progressive supranuclear palsy Neurology, August 14, 2001; 57(3): 467 - 473. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Fukuda, M. J. Mentis, Y. Ma, V. Dhawan, A. Antonini, A. E. Lang, A. M. Lozano, J. Hammerstad, K. Lyons, W. C. Koller, et al. Networks mediating the clinical effects of pallidal brain stimulation for Parkinson's disease: A PET study of resting-state glucose metabolism Brain, August 1, 2001; 124(8): 1601 - 1609. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Dagher, A. M. Owen, H. Boecker, and D. J. Brooks The role of the striatum and hippocampus in planning: A PET activation study in Parkinson's disease Brain, May 1, 2001; 124(5): 1020 - 1032. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Chen, R. R. Garg, A. M. Lozano, and A. E. Lang Effects of internal globus pallidus stimulation on motor cortex excitability Neurology, March 27, 2001; 56(6): 716 - 723. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Saint-Cyr, L. L. Trepanier, R. Kumar, A. M. Lozano, and A. E. Lang Neuropsychological consequences of chronic bilateral stimulation of the subthalamic nucleus in Parkinson's disease Brain, October 1, 2000; 123(10): 2091 - 2108. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Lang Surgery for Parkinson Disease: A Critical Evaluation of the State of the Art Arch Neurol, August 1, 2000; 57(8): 1118 - 1125. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||