- Email: [email protected]
The Prefrontal Substrate of Reflexive Saccade Inhibition in Humans
The Prefrontal Substrate of Reflexive Saccade Inhibition in Humans Christoph J. Ploner, Bertrand M. Gaymard, Sophie Rivaud-Péchoux, and Charles Pierrot-Deseilligny Background: Prefrontal dysfunction in neuropsychiatric disorders such as schizophrenia has been shown to impair inhibition of reflexive saccadic eye movements; however, it is unclear whether reflexive saccade inhibition can be attributed to a distinct subregion of the human prefrontal cortex. Methods: We tested 15 patients with acute unilateral ischemic lesions of the prefrontal cortex and 20 control subjects with an antisaccade task. Lesions were reconstructed using Talairach coordinates, and possible candidate regions for reflexive saccade inhibition were identified. Results: Significantly increased antisaccade error rates were observed in patients with lesions affecting a region in mid-dorsolateral prefrontal cortex or the white matter between this region and the anterior portions of the internal capsule. Antisaccade error rates of patients with lesions outside this region were normal. These findings were largely independent of lesion volume, postlesion delay, and subject age. Conclusions: Our findings suggest that inhibition of reflexive saccades depends on a circumscribed subregion of the human dorsolateral prefrontal cortex. This region closely corresponds to Brodmann area 46 as defined by recent cytoarchitectonic studies. Increased antisaccade error rates in patients with prefrontal pathology may be explained by dysfunction of this region.
Key Words: antisaccades, Brodmann area 46, dorsolateral prefrontal cortex, eye movements, frontal lobe, human
P
refrontal dysfunction contributes importantly to disability in many neuropsychiatric disorders (Stuss and Knight 2002). In addition to conventional neuropsychologic testing, prefrontal function can be assessed with simple oculomotor tasks (Pierrot-Deseilligny et al 1991, 2003). Among these, the antisaccade paradigm is the most frequently used (Broerse et al 2001; Everling and Fischer 1998; McDowell and Clementz 2001). In this task, the subject is presented a lateral target while fixating straight ahead and is requested to look to the opposite side as soon as the target appears (Hallett 1978). Defective inhibition of reflexive eye movements to the target results in increased frequency of erroneous prosaccades, which can easily be quantified as antisaccade error rate (AER), even outside an oculomotor laboratory (Currie et al 1991; Pierrot-Deseilligny et al 1991, 2003). Until now, increased AERs have been found with schizophrenia, Alzheimer disease, Huntington disease, and several other neuropsychiatric diseases affecting the prefrontal cortex (Broerse et al 2001; Everling and Fischer 1998; McDowell and Clementz 2001). For patients suffering from schizophrenia and their relatives, analysis of AERs has become a widely used research tool for prefrontal function (Broerse et al 2001; McDowell and Clementz 2001). Despite a clear-cut clinical association between prefrontal dysfunction and increased AERs, previous lesion studies have failed to identify unequivocally a distinct neural substrate of reflexive saccade inhibition in the human prefrontal cortex. An initial lesion study by Guitton and colleagues (1985) suggested
From the Klinik für Neurologie (CJP), Charité, Berlin, Germany, and INSERM U 289 and Service de Neurologie I (BMG, SRP, CPD), Hôpital de la Salpêtrière, Paris, France. Address reprint requests to PD Dr. Christoph J. Ploner, Klinik für Neurologie, Charité, Augustenburger Platz 1, D-13353 Berlin, Germany; E-mail: [email protected]. Received August 18, 2004; revised November 18, 2004; revised February 1, 2005; accepted February 15, 2005.
0006-3223/05/$30.00 doi:10.1016/j.biopsych.2005.02.017
that lesions of the frontal lobe involving the frontal or supplementary eye fields may lead to defective cancellation of reflexive glances. In subsequent studies, Pierrot-Deseilligny and colleagues (1991, 2003) found that lesions affecting Brodmann area 46 (BA 46) in the dorsolateral prefrontal cortex (DLPFC) invariably led to increased AERs, whereas this was rarely the case with lesions restricted to the frontal and supplementary eye fields. Fukushima and colleagues (1994), however, reported patients with lesions affecting lateral prefrontal cortex outside BA 46 and increased AERs. Functional imaging has not resolved this issue: activation of the DLPFC during antisaccade tasks was either absent (O’Driscoll et al 1995; Paus et al 1993; Raemakers et al 2002) or differed considerably in anteroposterior and dorsoventral directions (Doricchi et al 1997; Matsuda et al 2004; McDowell et al 2002; Müri et al 1998; Sweeney et al 1996). Considering these ambiguities, it has been suggested that the specificity of the antisaccade task for the location of dysfunctional brain structures may be low (Everling and Fischer 1998). This contrasts with recent results from experiments in nonhuman primates, suggesting that inhibition of reflexive saccades is exerted by a distinct subregion of the DLPFC in the principal sulcus (Gaymard et al 2003a). This region appears to be the source of a projection to the superior colliculus (Leichnetz et al 1981), lesions of which have recently been shown to increase AERs in humans (Condy et al 2004; Gaymard et al 2003b). The origin of this projection in the human prefrontal cortex has not yet been identified. We investigated whether inhibition of reflexive saccades can be attributed to a distinct subregion of the human lateral prefrontal cortex. In previous lesion studies on this matter, patients were grouped according to affected regions of interest in the prefrontal cortex (Pierrot-Deseilligny et al 1991, 2003). This approach did not exclude the possibility of regions outside BA 46 involved in reflexive saccade inhibition, possible damage to fibers from cortex outside this region in the subjacent white matter, or cumulative effects with increasing lesion size (Bates et al 2003). In our study, patients with focal ischemic lesions affecting the lateral frontal cortex were examined with a standard antisaccade task and grouped according to their AERs. The composite lesions of patients with normal and abnormal AERs were then superimposed in Talairach coordinates. We thus BIOL PSYCHIATRY 2005;57:1159 –1165 © 2005 Society of Biological Psychiatry
1160 BIOL PSYCHIATRY 2005;57:1159 –1165 aimed to define a prefrontal candidate region involved in reflexive saccade inhibition which allows for a more direct brain– behavior correlation in patients with neuropsychiatric diseases and abnormal AERs.
Methods and Materials Subjects Patients included 5 women and 10 men (mean age 48.3 ⫾ 3.1 years) recruited from the Departments of Neurology at the Salpêtrière Hospital in Paris, France. All patients sustained a unilateral atherothrombotic or embolic infarction affecting the frontal convexity (5 right-sided lesions, 10 left-sided lesions). Unusual causes of stroke, such as vasculitis, coagulation disorders, dissections of cerebral arteries, or venous thrombosis were not included. Mean time since lesion was 19.3 ⫾ 4.1 days and hence clearly beyond the upper temporal limit of infarct maturation in humans (Back et al 2004; Pantano et al 1999). The basal ganglia were spared in all patients. There was no clinical or neuroradiologic evidence of additional lesions. All patients were fully cooperative during eye movement recordings. All patients were free of additional neurologic or psychiatric disorders and free of psychoactive drugs. Visual fields were full, ocular motility was full range and conjugate, and there was no clinical evidence of neglect in all patients. A group of 11 women and 9 men (mean age 48.9 ⫾ 2.2 years) without any history of neurologic and psychiatric disorders served as control subjects. This number of control subjects was chosen because it permits the definition of reference ranges where the probability of a patient to be outside this range is p ⫽ 1/20 ⫽ .05. All subjects gave understanding and written consent before participation in the study, which was approved by the local ethics committee (Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale, Hôpital de la Salpêtrière, Paris) and conducted in conformity with the Declaration of Helsinki. Lesion Reconstruction For reconstruction of the patient’s lesions, we used seven transverse brain sections from ⫹4 mm to ⫹60 mm parallel above the anterior commissure–posterior commissure (AC-PC) line taken from the atlas by Talairach and Tournoux (1988). The sections were chosen to show the entire extent of the frontal convexity. Lesions were first identified on transverse computed tomography (CT) scans in 12 patients, and on transverse magnetic resonance imaging (MRI) scans in 3 patients. Similarly to Thompson-Schill et al (1998) and Karnath et al (2004), the segmented lesions were then transposed manually on the reference sections by using anatomic landmarks and measurement of distances between lesion borders (Figure 1). Lesion volumes were determined by outlining lesions on individual sections with the help of an image analysis program (Scion Image, Scion, Frederick, Maryland). Lesion areas were then multiplied with the corresponding reference section thickness and the resulting volumes were added to obtain a patients’ total lesion volume in cm3. This approach has the advantage to control for slight interindividual variations of brain volumes and therefore permits comparison of lesion volumes across subjects (Fiez et al 2000). Eye Movement Recordings Eye movements were recorded using horizontal direct current electrooculography with bitemporal electrodes. Data were sampled at a frequency of 200 Hz. The system had a www.elsevier.com/locate/biopsych
C.J. Ploner et al band with of 0 –100 Hz and a spatial resolution of .5 degrees. Subjects were seated in complete darkness with the head fixed at the temples. Visual cues were presented at a distance of 85 cm with red light-emitting diodes (LEDs) forming a curved horizontal array. The LEDs were .18° in size and 5 cd/m2 in luminance. Each recording session was preceded by 10 min of dark adaptation. Antisaccade Task At the beginning of a trial, the subject fixated on a central fixation point for an unpredictable duration of 3000 to 5500 msec, after which the central fixation point was switched off. A horizontal target appeared with unpredictable laterality at 25° eccentricity 200 msec after central fixation point offset (“gap”). The subject was instructed to move his or her eyes in the opposite direction as soon as the target appeared. After 1000 msec, the target was switched off, and the next trial began. An antigap paradigm with large eccentricities was chosen because it has been shown that antisaccade errors are more likely to occur with a temporal gap and with large eccentricities (Fischer and Weber 1997). A total of 18 to 20 targets was presented on each side. Subjects were allowed several practice trials to ensure comprehension of instructions. Data Analysis For each subject, right- and left-sided performance in the antisaccade task was quantified as the AER, that is, the percentage of erroneous prosaccades to the targets. Hence, a right-sided AER describes the percentage of erroneous prosaccades directed to the right-hand target. Latencies of correct antisaccades were also determined in each subject. The number of these saccades was low in patients with increased AERs, and these values therefore were not further analyzed. Trials with artifacts or premature saccades were excluded from analysis (5% in patients and control subjects). Statistics were done with reference to Altman (1991). Throughout the text, means are given with standard errors. Two-tailed Mann–Whitney tests and Wilcoxon tests were used for statistical analysis. In control subjects, mean right and left AER did not differ significantly (right: 7.7% ⫾ 1.8%; left: 8.2% ⫾ 1.6%; p ⫽ .48). For calculation of the reference range, right and left AERs were therefore pooled in control subjects. Thus, mean AER in control subjects was 8.0% ⫾ 1.5%. Mean latency of correct antisaccades was 283 ⫾ 20 msec right and 292 ⫾ 19 msec left. Patient AERs were considered abnormal if outside the range of control AERs, that is, greater than 26%. This upper range cutoff agrees with our previous studies of AERs in normal subjects performed with the same type of paradigm and a comparable number of trials (Condy et al 2004; Gaymard et al 1998, 1999, 2003b; Pierrot-Deseilligny et al 2003).
Results Antisaccade Error Rates The AERs of nine patients (three with right-sided lesions and six with left-sided lesions) fell within the normal range, despite considerable damage to prefrontal cortex and its subjacent white matter (Figure 1, second column; Figure 2, second and third columns). Mean AER in these patients was 7.9% ⫾ 1.7% ipsilaterally and 9.8% ⫾ 2.4% contralaterally to the lesion side (p ⫽ .98 and p ⫽ .59 compared with control subjects). Mean latency of correct antisaccades was 351 ⫾ 57 msec ipsilaterally and 311 ⫾ 50 msec contralaterally to the lesion side. Mean age of these patients was 44.1 ⫾ 3.5 years (p ⫽ .23 compared with control
C.J. Ploner et al
BIOL PSYCHIATRY 2005;57:1159 –1165 1161
Figure 1. Anatomy of the frontal cortex and lesion drawings. Seven transverse sections parallel above the anterior commissure–posterior commissure line (AC-PC line) from the atlas of Talairach and Tournoux (1988) arranged from dorsal to ventral. Each row corresponds to a distinct horizontal level from ⫹60 mm to ⫹4 mm above the AC-PC line (z-level of the Talairach coordinate system). VAC, vertical anterior commissure line; VPC, vertical posterior commissure line. The black bars in the lower right corner of the figure correspond to a length of 50 mm in right–left and anteroposterior directions (x- and ylevels of the Talairach coordinate system). First column: normal anatomy. R, right; L, left; F1, superior frontal gyrus; F2, middle frontal gyrus; F3, inferior frontal gyrus; PrC, precentral gyrus. Second column: lesions of patients with antisaccade error rates (AERs) within the control range. Third column: lesions of patients with increased AERs. Light gray, lesion; dark gray, overlap of two lesions; nearly black, overlap of three or more lesions. Fourth column: lesion subtraction. Gray, regions affected in patients with increased AERs and spared in patients with AERs within the control range. Arrows point to the anterior and posterior limits of candidate regions for reflexive saccade inhibition.
subjects), mean time since lesion was 15.4 ⫾ 3.2 days, mean lesion volume was 43 ⫾ 9.7 cm3. The AERs of six patients (two with right-sided lesions and four with left-sided lesions), at least contralaterally to the lesion side, clearly fell outside the normal range (Figure 1, third column; Figure 2, fourth and fifth columns). Mean AER in these patients
was 57.7% ⫾ 14.7% ipsilaterally and 75.2% ⫾ 8.2% contralaterally to the lesion side (p ⫽ .001 and p ⬍ .001 compared with control subjects). No significant difference was found between AERs ipsilaterally and contralaterally to the lesion side (p ⫽ .12). Mean latency of correct antisaccades was 524 ⫾ 68 msec ipsilaterally and 562 ⫾ 99 msec contralaterally to the lesion side. Mean age of www.elsevier.com/locate/biopsych
1162 BIOL PSYCHIATRY 2005;57:1159 –1165
Figure 2. Antisaccade error rates (AERs). The gray-shaded area corresponds to the control range. ipsi, AERs ipsilateral to the lesion side; cont, AERs contralateral to the lesion side. The second and third columns show patients with AERs within the control range, the fourth and fifth column show patients with increased AERs. White dots show AERs of patients with rightsided lesions.
these patients was 54.7 ⫾ 4.9 years (p ⫽ .22 difference with control subjects), mean time since lesion was 25 ⫾ 9 days, mean lesion volume was 48.2 ⫾ 6.9 cm3. There was no significant difference in age, time since lesion, and lesion volume between both patient groups (p ⫽ .11, p ⫽ .46, and p ⫽ .33, respectively). Candidate Regions for Reflexive Saccade Inhibition Lesions from patients with AERs within the control range (Figure 1, second column) and lesions from patients with increased AERs (Figure 1, third column) were then subtracted from each other because regions affected in patients with increased AERs but spared in patients with AERs within the control range should contain the region(s) involved in reflexive saccade inhibition. This subtraction yielded several regions in various parts of the prefrontal cortex, which were then transposed on separate sections to define candidate regions for reflexive saccade inhibition (Figure 1, fourth column, gray-shaded regions). It must be borne in mind, however, that a region affected in patients with increased AERs and spared in patients with AERs within the control range is not necessarily involved in reflexive saccade inhibition. For example, if we would have had a large frontoparietal lesion in the patient group with increased AERs, it would not indicate that the parietal cortex is involved in reflexive saccade inhibition. For definition of candidate regions, it is thus important to take the course of the fibers projecting from frontal cortex to the anterior part of the internal capsule into account. It has been shown previously that lesions to this region of the internal capsule affect the corticotectal pathway that controls reflexive saccade inhibition (Condy et al 2004; Gaymard et al 2003b; Leichnetz 1981). Therefore, cortical regions on the sections in the fourth column of Figure 1 were considered candidate regions when lesion to the subjacent caudal and ventral white www.elsevier.com/locate/biopsych
C.J. Ploner et al matter containing the efferent fibers of this region to the anterior limb of the internal capsule yielded increased AERs and vice versa. The course of these fibers was determined on coronal and horizontal brain sections of Déjerine (1895, 1901) and on coronal and sagittal brain sections of Talairach and Tournoux (1988). Fiber tracts were located on the horizontal sections of this study by using anatomic landmarks and the Talairach coordinate system. For example, the efferent fibers of the gray-shaded region in the right precentral gyrus of the ⫹60 and ⫹50 mm sections in the fourth column of Figure 1 must descend laterally from the fundus of the callosomarginal sulcus and the corpus callosum and at least 20 mm laterally from the midline through the ⫹40 mm horizontal section located ventrally of this region. Lesions to this part of the ⫹40 mm section yielded no increase in AERs, however, as can be seen in the second column of Figure 1. Hence, the gray-shaded region in the right precentral gyrus of the ⫹60 and ⫹50 mm sections in the fourth column of Figure 1 is not a candidate region. By contrast, in the right hemisphere, a region on the ⫹20 mm section fulfilled our criteria for a candidate region (Figure 1, fourth column, arrows). This region is largely congruent with the middle portions of the middle frontal gyrus (F2) and extends from about 28 to 60 mm anterior to the vertical anterior commissure line (VAC) line. Lesions to the superior frontal gyrus (F1), the dorsal parts of F2, the inferior frontal gyrus (F3), and the precentral gyrus (PrC) yielded no increase in AER. The ventral limits of this region could not be determined, because there were no patients with increased AERs and lesions on more ventral sections. A corresponding region was found in the left hemisphere on section ⫹20 mm (Figure 1, fourth column, arrows). Here, the candidate region extended from about 10 to 50 mm anterior to the VAC line and included the middle portion of F2 but also parts of F3. Compared with the right hemisphere, however, this region also extended more dorsally to section ⫹28 mm, where it extended from about 19 to 48 mm anterior to the VAC line and included the middle portion of F2 but not F3. Caudally, this region extended to section ⫹12 mm and included ventral parts of F2 and F3. Lesions to F1, to the dorsal parts of F2, and to the PrC yielded no increased AERs.
Discussion This study confirms that unilateral lesions affecting the prefrontal cortex may lead to increased AERs. Nonetheless, even patients with large lesions of lateral prefrontal cortex showed AERs within the control range, as long as their lesions did not affect a region in the mid-DLPFC or the white matter between this region and the anterior part of the internal capsule. Significant increases in AER appeared to be largely independent of lesion volume, postlesion delay, and age of patients. We suggest that the candidate region in DLPFC identified in this study delineates the possible extent of a prefrontal subregion involved in reflexive saccade inhibition. In the following paragraphs, we discuss how these findings relate to previous physiologic and anatomic studies in humans and nonhuman primates. Recently, the cytoarchitectonic parcellation of human prefrontal cortex has been reanalyzed for the left hemisphere (Petrides and Pandya 1999; Rajkowska and Goldman-Rakic 1995). Compared with these maps, the left candidate region of our study comprises the entire constant portion of BA 46 in the middle portions of F2, as well as transitional cortex to BA 9 posterior to this region and areas BA 44 and BA 45 in F3; BA 8, the invariable portion of BA9, and BA10 were not parts of this region. There are no comparable studies for the right hemi-
C.J. Ploner et al sphere, and it is therefore difficult to relate unequivocally the right candidate region of our study to a distinct cytoarchitectonic subdivision of the right prefrontal cortex. It appears, however, that with the exception of the speech-dependent asymmetries of BA 44 and BA 45, there are no major right–left asymmetries for the remaining prefrontal subregions, in particular for BA 46 (Rajkowska and Goldman-Rakic 1995). When the cytoarchitectonic parcellation of the left hemisphere is applied to the right hemisphere, as is regularly done in functional imaging studies, the right candidate region of our study mainly comprises the lower parts of the constant portion of BA 46; BA 8, the invariable portion of BA9, as well as BA10, BA 44, and BA 45 are not part of this region. Therefore, BA 46 is the only region that appears to be affected in right and left candidate regions. In the monkey homologue of this region, injection of anterograde tracers results in labeling of the deep layers of the superior colliculus via pathways through the anterior part of the internal capsule (Leichnetz et al 1981; Selemon and Goldman-Rakic 1988). Furthermore, recent studies have shown that pharmacologic deactivation of a small circumscribed region in Walker area 46 in monkeys results in increased AERs (Gaymard et al 2003a). We thus believe that the findings from our study are entirely consistent with our initial proposal that inhibition of reflexive saccades may depend on a circumscribed DLPFC subregion, which corresponds to BA 46 and which may be the source of a projection onto the superior colliculus where local inhibitory interneurons may finally mediate suppression of unwanted saccades (Condy et al 2004; Munoz and Istvan 1998; Pierrot-Deseilligny et al 1991). When the different angulations of the canthomeatal line and AC-PC line are taken into account (Weiss et al 2003), it is nevertheless obvious that this subregion is located more ventrally in F2 than assumed in our first study on this matter (PierrotDeseilligny et al 1991) and, at least in the right hemisphere, slightly more ventrally than assumed in a subsequent study with lesions transposed on sections parallel to the AC-PC line (PierrotDeseilligny et al 2003). The candidate regions defined in our study were also affected in DLPFC patients of these preceding studies, however. The findings from our study illustrate the limitations of lesion studies in which patient groups are solely based on affected regions of interest (Bates et al 2003). If we would have simply compared performance of all patients with prefrontal lesions to performance of control subjects, there would still have been increased AERs in patients (mean ipsilateral AER: 27.8% ⫾ 8.6%; mean contralateral AER: 36.0% ⫾ 9.2%) and a significant difference between patients and control subjects, at least for contralateral AERs (ipsilateral: p ⫽ .11; contralateral: p ⫽ .02). Hence, almost any prefrontal region of interest would have yielded significant results. Likewise, zones of lesion overlap in patients with increased AERs may be misleading when the subcortical projections of these regions are not taken into account. These points may explain, at least in part, the differences between our results and preceding studies of AERs in patients with lateral prefrontal lesions (Fukushima et al 1994; Guitton et al 1985) and in patients with lesions of the anterior cingulate cortex (Milea et al 2003). The putative oculomotor part of this latter region can be found in the medial frontal cortex on the ⫹28 and ⫹40 mm sections of Figure 1, where it extends several millimeters anterior of the VAC line (Gaymard et al 1998). From inspection of the ⫹12, ⫹20, and ⫹28 mm sections in the fourth column of Figure 1, however, it is clear that any lesion of this region that extends medially and ventrally into the white matter runs the risk of damaging fibers from F2 to the internal capsule. This hypothesis
BIOL PSYCHIATRY 2005;57:1159 –1165 1163 is supported by the fact that small vascular lesions of the anterior cingulate cortex yield no increase in AER, at least with the stimulation parameters of our study (Gaymard et al 1998), whereas more extensive postsurgical lesions caused clear impairments in an antisaccade task (Milea et al 2003). How do our results relate to the conflicting activation patterns observed in functional imaging studies with antisaccade tasks (Doricchi et al 1997; Matsuda et al 2004; McDowell et al 2002; Müri et al 1998; O’Driscoll et al 1995; Paus et al 1993; Raemakers et al 2002; Sweeney et al 1996)? Even when a slight and inevitable intra- and intersubject variance in manual lesion reconstruction is taken into account (Fiez et al. 2000), it is obvious that DLPFC activation in some studies is clearly within (Matsuda et al 2004; Müri et al 1998; Sweeney et al 1996), and in some studies clearly outside the candidate regions of our study (Doricchi et al 1997; McDowell et al 2002). Comparison of task parameters such as target eccentricities, gap–step antisaccade tasks, or control conditions provides no unequivocal explanation for these discrepancies and for the absence of DLPFC activation in other studies (O’Driscoll et al 1995; Paus et al 1993; Raemakers et al 2002). It should be conceded that it is difficult to design tasks that control for fixation demands and number and timing of saccades in an antisaccade task. For example, a corrected antisaccade error consists always of at least two saccades, a fact that renders difficult the comparison of antisaccade to prosaccade control conditions without eye movement recordings. Similarly, fixation control conditions require pronounced and presumably inhibitory fixation-related activity that may be difficult to disentangle from inhibition of reflexive saccades in an antisaccade task (Raemaekers et al 2002). Because the primate DLPFC appears to contain more regions responding to fixation and saccades than assumed previously (Moschovakis et al 2004), different DLPFC activation patterns in functional imaging studies may at least partially result from neuronal activity unrelated to inhibition of reflexive saccades. Event-related task designs with simultaneous eye movement recordings and separate analysis of correct antisaccades and antisaccade errors may finally provide a resolution of these discrepancies. Several lines of evidence have implicated DLPFC subregions in the pathophysiology of schizophrenia (Bogerts 1999; GoldmanRakic and Selemon 1997; Harrison 1999). For example, neuropathologic studies have found evidence for reduced interneuronal neuropil, decreased levels of synaptophysin, and elevated apolipoprotein D levels in BA 46 of schizophrenia patients (Glantz and Lewis 1997; Selemon et al 1998; Thomas et al 2003). The consistently elevated AERs of patients with schizophrenia and their relatives therefore have frequently been attributed to DLPFC dysfunction (e.g., Broerse et al 2001; McDowell and Clementz 2001). The candidate regions defined in our study are in good agreement with this hypothesis and argue against recent hypotheses that relate increased AERs in patients with schizophrenia to a dysfunctional interaction between frontal cortex and basal ganglia rather than to a selective dysfunction of the DLPFC itself (Raemaekers et al 2002). Longitudinal oculomotor studies in patients with DLPFC lesions will have to clarify whether chronic dysfunction of the DLPFC, as observed in schizophrenia, yields increased AERs similarly to our patients, which suffered from relatively acute vascular lesions. If this is the case, our candidate regions may help to define the borders of regions of interest for correlation of neuropathologic changes with increased AERs in patients with schizophrenia. Taken together, the findings from this study strongly suggest that, as in the nonhuman primate, the human DLPFC contains a www.elsevier.com/locate/biopsych
1164 BIOL PSYCHIATRY 2005;57:1159 –1165 distinct subregion that is involved in reflexive saccade inhibition as revealed by standard antisaccade tasks. Compared with recent cytoarchitectonic studies, this region appears to be located bilaterally in BA 46. Although successful performance in an antisaccade task requires additional operations in other frontal regions, in particular triggering of a voluntary saccade by the frontal eye field (Gaymard et al 1999), it is inhibition of reflexive saccades that can be quantified as AER. Increased AERs may therefore provide a simple behavioral marker of BA 46 dysfunction. We therefore hope that the candidate regions of this study provide more detailed anatomic information, allowing for testable anatomic predictions in neuropsychiatric patients with increased AERs.
This study was supported by the Deutsche Forschungsgemeinschaft (Pl 248/2-1). Altman DG (1991): Practical Statistics for Medical Research. London: Chapman & Hall. Back T, Hemmen T, Schüler OG (2004): Lesion evolution in cerebral ischemia. J Neurol 251:388 –397. Bates E, Wilson SM, Pinar Saygin A, Dick F, Sereno MI, Knight RT, Dronkers NF (2003): Voxel-based lesion-symptom mapping. Nature Neurosci 6:448 – 450. Bogerts B (1999): The neuropathology of schizophrenic disease: Historical aspects and present knowledge. Eur Arch Psychiatry Clin Neurosci 249(S4):IV/2–IV/3. Broerse A, Crawford TJ, den Boer JA (2001): Parsing cognition in schizophrenia using saccadic eye movements: A selective overview. Neuropsychologia 39:742–756. Condy C, Rivaud-Péchoux S, Ostendorf F, Ploner CJ, Gaymard B (2004): Neural control of antisaccades: Role of subcortical structures. Neurology 63:1571–1578. Currie J, Ramsden B, McArthur C, Maruff P (1991): Validation of a clinical antisaccadic eye movement test in the assessment of dementia. Arch Neurol 48:644 – 648. Déjerine J (1895): Anatomie des centres nerveux, tome premier (vol. 1). Paris: Rueff. Déjerine J (1901): Anatomie des centres nerveux, tome deuxième (vol. 2). Paris: Rueff. Doricchi F, Perani D, Incoccia C, Grassi F, Cappa SF, Bettinardi V, et al (1997) : Neural control of fast-regular saccades and antisaccades: An investigation using positron emission tomography. Exp Brain Res 116:50 – 62. Everling S, Fischer B (1998): The antisaccade: A review of basic research and clinical studies. Neuropsychologia 36:885– 899. Fiez JA, Damasio H, Grabowski TJ (2000): Lesion segmentation and manual warping to a reference brain: Intra- and interobserver reliability. Hum Brain Mapp 9:192–211. Fischer B, Weber H (1997): Effects of stimulus conditions on the performance of antisaccades in man. Exp Brain Res 116:191–200. Fukushima J, Fukushima K, Miyasaka K, Yamashita I (1994): Voluntary control of saccadic eye movement in patients with frontal cortical lesions and parkinsonian patients in comparison with that in schizophrenics. Biol Psychiatry 36:21–30. Gaymard B, Condy C, Wattiez N, Rivaud-Péchoux S (2003a): Anatomy and pharmacology of antisaccades in monkeys. Program No. 72.18. 2003 Abstract Viewer and Itinerary Planner. Washington DC: Society for Neuroscience. Gaymard B, François C, Ploner CJ, Condy C, Rivaud-Péchoux S (2003b): A direct prefrontotectal tract against distractibility in the human brain. Ann Neurol 53:542–545. Gaymard B, Ploner CJ, Rivaud-Pechoux S, Pierrot-Deseilligny C (1999): The frontal eye field is involved in spatial short-term memory but not in reflexive saccade inhibition. Exp Brain Res 129:288 –301. Gaymard B, Rivaud S, Cassarini JF, Dubard T, Rancurel G, Agid Y, PierrotDeseilligny C (1998): Effects of anterior cingulate cortex lesions on ocular saccades in humans. Exp Brain Res 120:173–183.
www.elsevier.com/locate/biopsych
C.J. Ploner et al Glantz LA, Lewis DA (1997): Reduction of synaptophysin immunoreactivity in the prefrontal cortex of subjects with schizophrenia. Regional and diagnostic specificity. Arch Gen Psychiatry 54:660 – 669. Goldman-Rakic PS, Selemon LD (1997): Functional and anatomical aspects of prefrontal pathology in schizophrenia. Schizophr Bull 23:437– 458. Guitton D, Buchtel DA, Douglas RM (1985): Frontal lobe lesions in man cause difficulties in suppressing reflexive glances and in generating goal-directed saccades. Exp Brain Res 58:455– 472. Hallett PE (1978): Primary and secondary saccades to goals defined by instructions. Vision Res 18:1279 –1296. Harrison PJ (1999): The neuropathology of schizophrenia. Brain 122:593– 624. Karnath HO, Fruhmann Berger M, Kuker W, Rorden C (2004): The anatomy of spatial neglect based on voxelwise statistical analysis: A study of 140 patients. Cereb Cortex 14:1164 –1172. Leichnetz GR, Spencer RF, Hardy SGP, Astruc J (1981): The prefrontal corticotectal projection in the monkey: An anterograde and retrograde horseradish peroxidase study. Neuroscience 6:1023–1041. Matsuda T, Matsuura M, Ohkubo T, Ohkubo H, Matsushima E, Inoue K, et al (2004) Functional MRI of brain activation during visually guided saccades and antisaccades: Cortical and subcortical networks. Psychiat Res 131:147–155. McDowell JE, Brown GG, Paulus M, Martinez A, Stewart SE, Dubowitz DJ, Braff DL (2002): Neural control of refixation saccades and antisaccades in normal and schizophrenia subjects. Biol Psychiatry 51:216 –223. McDowell JE, Clementz BA (2001): Behavioral and brain imaging studies of saccadic performance in schizophrenia. Biol Psychol 57:5–22. Milea D, Lehericy S, Rivaud-Pechoux S, Duffau H, Lobel E, Capelle L, et al (2003): Antisaccade deficit after anterior cingulate cortex resection. Neuroreport 14:283–287. Moschovakis AK, Gregoriou GG, Ugolini G, Doldan M, Graf W, Guldin W, et al (2004): Oculomotor areas of the primate frontal lobes: a transneuronal transfer of rabies virus and [14C]-2-deoxyglucose functional imaging study. J Neurosci 24:5726 –5740. Munoz DP, Istvan PJ (1998): Lateral inhibitory interactions in the intermediate layers of monkey superior colliculus. J Neurophysiol 79:1193–1209. Müri RM, Heid O, Nirkko AC, Ozdoba C, Felblinger J, Schroth G, Hess CW (1998): Functional organization of saccades and antisaccades in the frontal lobe in humans: A study with echo planar functional magnetic resonance imaging. J Neurol Neurosurg Psychiatry 65:374 –377. O’Driscoll GA, Alpert NA, Matthysse SW, Levy DL, Rauch SL, Holzman PS (1995): Functional neuroanatomy of antisaccde eye movements investigated with positron emission tomography. Proc Natl Acad Sci U S A 92: 925–929. Pantono P, Caramina F, Bozzao L, Dieler C, von Kummer R (1999): Delayed increase in infarct volume after cerebral ischemia. Stroke 30:502–507. Paus T, Petrides M, Evans AC, Meyer E (1993): Role of the human anterior cingulate cortex in the control of oculomotor, manual, and speech responses: A positron emission tomography study. Proc Natl Acad Sci U S A 70:453– 469. Petrides M, Pandya DN (1999): Dorsolateral prefrontal cortex: Comparative cytoarchitectonic analysis in the human and the macaque brain and corticocortical connection patterns. Eur J Neurosci 11:1101–1136. Pierrot-Deseilligny C, Müri RM, Ploner CJ, Gaymard B, Demeret S, RivaudPechoux S (2003): Decisional role of the dorsolateral prefrontal cortex in ocular motor behaviour. Brain 126:1460 –1473. Pierrot-Deseilligny C, Rivaud S, Gaymard B, Agid Y (1991): Cortical control of reflexive visually guided saccades. Brain 114:1473–1485. Raemaekers M, Jansma JM, Cahn W, Van der Geest JN, van der Linden JA, Kahn RS, Ramsey NF (2002): Neuronal substrate of the saccadic inhibition deficit in schizophrenia investigated with 3-dimensional eventrelated functional magnetic resonance imaging. Arch Gen Psychiatry 59: 313–320. Rajkowska G, Goldman-Rakic PS (1995): Cytoarchitectonic definition of prefrontal areas in the normal human cortex: II. Variability in locations of areas 9 and 46 and relationship to the Talairach coordinate system. Cereb Cortex 5:323–337. Selemon LD, Goldman-Rakic PS (1988): Common cortical and subcortical targets of the dorsolateral prefrontal and posterior parietal cortices in the rhesus monkey: Evidence for a distributed neural network subserving spatially guided behavior. J Neurosci 8:4049 – 4068. Selemon LD, Rajkowska G, Goldman-Rakic PS (1998): Elevated neuronal density in prefrontal area 46 in brains from schizophrenic patients: ap-
C.J. Ploner et al plication of a three-dimensional, stereologic counting method. J Comp Neurol 392:402– 412. Stuss DT, Knight RT (2002): Principles of Frontal Lobe Function. Oxford: Oxford University Press. Sweeney JA, Mintun MA, Kwee S, Wiseman MB, Brown DL, Rosenberg DR, Carl JR (1996): Positron emission tomography study of voluntary saccadic eye movements and spatial working memory. Proc Natl Acad Sci U S A 75:454 – 468. Talairach J, Tournoux P (1988): Co-Planar Stereotaxic Atlas of the Human Brain. Stuttgart: Thieme.
BIOL PSYCHIATRY 2005;57:1159 –1165 1165 Thomas EA, Dean B, Scarr E, Copolov D, Sutcliffe JG (2003): Differences in neuroanatomical sites of apoD elevation discriminate between schizophrenia and bipolar disorder. Mol Psychiatry 8:167–175. Thompson-Schill SL, Swick D, Farah MJ, D’Esposito M, Kan IP, Knight RT (1998): Verb generation in patients with focal frontal lesions: a neuropsychological test of neuroimaging findings. Proc Natl Acad Sci USA 95: 15855–15860. Weiss KL, Pan H, Storrs J, Strub W, Weiss JL, Jia L, Eldevik OP (2003): Clinical brain MR imaging prescriptions in Talairach space: Technologist- and computer-driven methods. Am J Neuroradiol 24:922–929.
www.elsevier.com/locate/biopsych
Key Words: antisaccades, Brodmann area 46, dorsolateral prefrontal cortex, eye movements, frontal lobe, human
P
refrontal dysfunction contributes importantly to disability in many neuropsychiatric disorders (Stuss and Knight 2002). In addition to conventional neuropsychologic testing, prefrontal function can be assessed with simple oculomotor tasks (Pierrot-Deseilligny et al 1991, 2003). Among these, the antisaccade paradigm is the most frequently used (Broerse et al 2001; Everling and Fischer 1998; McDowell and Clementz 2001). In this task, the subject is presented a lateral target while fixating straight ahead and is requested to look to the opposite side as soon as the target appears (Hallett 1978). Defective inhibition of reflexive eye movements to the target results in increased frequency of erroneous prosaccades, which can easily be quantified as antisaccade error rate (AER), even outside an oculomotor laboratory (Currie et al 1991; Pierrot-Deseilligny et al 1991, 2003). Until now, increased AERs have been found with schizophrenia, Alzheimer disease, Huntington disease, and several other neuropsychiatric diseases affecting the prefrontal cortex (Broerse et al 2001; Everling and Fischer 1998; McDowell and Clementz 2001). For patients suffering from schizophrenia and their relatives, analysis of AERs has become a widely used research tool for prefrontal function (Broerse et al 2001; McDowell and Clementz 2001). Despite a clear-cut clinical association between prefrontal dysfunction and increased AERs, previous lesion studies have failed to identify unequivocally a distinct neural substrate of reflexive saccade inhibition in the human prefrontal cortex. An initial lesion study by Guitton and colleagues (1985) suggested
From the Klinik für Neurologie (CJP), Charité, Berlin, Germany, and INSERM U 289 and Service de Neurologie I (BMG, SRP, CPD), Hôpital de la Salpêtrière, Paris, France. Address reprint requests to PD Dr. Christoph J. Ploner, Klinik für Neurologie, Charité, Augustenburger Platz 1, D-13353 Berlin, Germany; E-mail: [email protected]. Received August 18, 2004; revised November 18, 2004; revised February 1, 2005; accepted February 15, 2005.
0006-3223/05/$30.00 doi:10.1016/j.biopsych.2005.02.017
that lesions of the frontal lobe involving the frontal or supplementary eye fields may lead to defective cancellation of reflexive glances. In subsequent studies, Pierrot-Deseilligny and colleagues (1991, 2003) found that lesions affecting Brodmann area 46 (BA 46) in the dorsolateral prefrontal cortex (DLPFC) invariably led to increased AERs, whereas this was rarely the case with lesions restricted to the frontal and supplementary eye fields. Fukushima and colleagues (1994), however, reported patients with lesions affecting lateral prefrontal cortex outside BA 46 and increased AERs. Functional imaging has not resolved this issue: activation of the DLPFC during antisaccade tasks was either absent (O’Driscoll et al 1995; Paus et al 1993; Raemakers et al 2002) or differed considerably in anteroposterior and dorsoventral directions (Doricchi et al 1997; Matsuda et al 2004; McDowell et al 2002; Müri et al 1998; Sweeney et al 1996). Considering these ambiguities, it has been suggested that the specificity of the antisaccade task for the location of dysfunctional brain structures may be low (Everling and Fischer 1998). This contrasts with recent results from experiments in nonhuman primates, suggesting that inhibition of reflexive saccades is exerted by a distinct subregion of the DLPFC in the principal sulcus (Gaymard et al 2003a). This region appears to be the source of a projection to the superior colliculus (Leichnetz et al 1981), lesions of which have recently been shown to increase AERs in humans (Condy et al 2004; Gaymard et al 2003b). The origin of this projection in the human prefrontal cortex has not yet been identified. We investigated whether inhibition of reflexive saccades can be attributed to a distinct subregion of the human lateral prefrontal cortex. In previous lesion studies on this matter, patients were grouped according to affected regions of interest in the prefrontal cortex (Pierrot-Deseilligny et al 1991, 2003). This approach did not exclude the possibility of regions outside BA 46 involved in reflexive saccade inhibition, possible damage to fibers from cortex outside this region in the subjacent white matter, or cumulative effects with increasing lesion size (Bates et al 2003). In our study, patients with focal ischemic lesions affecting the lateral frontal cortex were examined with a standard antisaccade task and grouped according to their AERs. The composite lesions of patients with normal and abnormal AERs were then superimposed in Talairach coordinates. We thus BIOL PSYCHIATRY 2005;57:1159 –1165 © 2005 Society of Biological Psychiatry
1160 BIOL PSYCHIATRY 2005;57:1159 –1165 aimed to define a prefrontal candidate region involved in reflexive saccade inhibition which allows for a more direct brain– behavior correlation in patients with neuropsychiatric diseases and abnormal AERs.
Methods and Materials Subjects Patients included 5 women and 10 men (mean age 48.3 ⫾ 3.1 years) recruited from the Departments of Neurology at the Salpêtrière Hospital in Paris, France. All patients sustained a unilateral atherothrombotic or embolic infarction affecting the frontal convexity (5 right-sided lesions, 10 left-sided lesions). Unusual causes of stroke, such as vasculitis, coagulation disorders, dissections of cerebral arteries, or venous thrombosis were not included. Mean time since lesion was 19.3 ⫾ 4.1 days and hence clearly beyond the upper temporal limit of infarct maturation in humans (Back et al 2004; Pantano et al 1999). The basal ganglia were spared in all patients. There was no clinical or neuroradiologic evidence of additional lesions. All patients were fully cooperative during eye movement recordings. All patients were free of additional neurologic or psychiatric disorders and free of psychoactive drugs. Visual fields were full, ocular motility was full range and conjugate, and there was no clinical evidence of neglect in all patients. A group of 11 women and 9 men (mean age 48.9 ⫾ 2.2 years) without any history of neurologic and psychiatric disorders served as control subjects. This number of control subjects was chosen because it permits the definition of reference ranges where the probability of a patient to be outside this range is p ⫽ 1/20 ⫽ .05. All subjects gave understanding and written consent before participation in the study, which was approved by the local ethics committee (Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale, Hôpital de la Salpêtrière, Paris) and conducted in conformity with the Declaration of Helsinki. Lesion Reconstruction For reconstruction of the patient’s lesions, we used seven transverse brain sections from ⫹4 mm to ⫹60 mm parallel above the anterior commissure–posterior commissure (AC-PC) line taken from the atlas by Talairach and Tournoux (1988). The sections were chosen to show the entire extent of the frontal convexity. Lesions were first identified on transverse computed tomography (CT) scans in 12 patients, and on transverse magnetic resonance imaging (MRI) scans in 3 patients. Similarly to Thompson-Schill et al (1998) and Karnath et al (2004), the segmented lesions were then transposed manually on the reference sections by using anatomic landmarks and measurement of distances between lesion borders (Figure 1). Lesion volumes were determined by outlining lesions on individual sections with the help of an image analysis program (Scion Image, Scion, Frederick, Maryland). Lesion areas were then multiplied with the corresponding reference section thickness and the resulting volumes were added to obtain a patients’ total lesion volume in cm3. This approach has the advantage to control for slight interindividual variations of brain volumes and therefore permits comparison of lesion volumes across subjects (Fiez et al 2000). Eye Movement Recordings Eye movements were recorded using horizontal direct current electrooculography with bitemporal electrodes. Data were sampled at a frequency of 200 Hz. The system had a www.elsevier.com/locate/biopsych
C.J. Ploner et al band with of 0 –100 Hz and a spatial resolution of .5 degrees. Subjects were seated in complete darkness with the head fixed at the temples. Visual cues were presented at a distance of 85 cm with red light-emitting diodes (LEDs) forming a curved horizontal array. The LEDs were .18° in size and 5 cd/m2 in luminance. Each recording session was preceded by 10 min of dark adaptation. Antisaccade Task At the beginning of a trial, the subject fixated on a central fixation point for an unpredictable duration of 3000 to 5500 msec, after which the central fixation point was switched off. A horizontal target appeared with unpredictable laterality at 25° eccentricity 200 msec after central fixation point offset (“gap”). The subject was instructed to move his or her eyes in the opposite direction as soon as the target appeared. After 1000 msec, the target was switched off, and the next trial began. An antigap paradigm with large eccentricities was chosen because it has been shown that antisaccade errors are more likely to occur with a temporal gap and with large eccentricities (Fischer and Weber 1997). A total of 18 to 20 targets was presented on each side. Subjects were allowed several practice trials to ensure comprehension of instructions. Data Analysis For each subject, right- and left-sided performance in the antisaccade task was quantified as the AER, that is, the percentage of erroneous prosaccades to the targets. Hence, a right-sided AER describes the percentage of erroneous prosaccades directed to the right-hand target. Latencies of correct antisaccades were also determined in each subject. The number of these saccades was low in patients with increased AERs, and these values therefore were not further analyzed. Trials with artifacts or premature saccades were excluded from analysis (5% in patients and control subjects). Statistics were done with reference to Altman (1991). Throughout the text, means are given with standard errors. Two-tailed Mann–Whitney tests and Wilcoxon tests were used for statistical analysis. In control subjects, mean right and left AER did not differ significantly (right: 7.7% ⫾ 1.8%; left: 8.2% ⫾ 1.6%; p ⫽ .48). For calculation of the reference range, right and left AERs were therefore pooled in control subjects. Thus, mean AER in control subjects was 8.0% ⫾ 1.5%. Mean latency of correct antisaccades was 283 ⫾ 20 msec right and 292 ⫾ 19 msec left. Patient AERs were considered abnormal if outside the range of control AERs, that is, greater than 26%. This upper range cutoff agrees with our previous studies of AERs in normal subjects performed with the same type of paradigm and a comparable number of trials (Condy et al 2004; Gaymard et al 1998, 1999, 2003b; Pierrot-Deseilligny et al 2003).
Results Antisaccade Error Rates The AERs of nine patients (three with right-sided lesions and six with left-sided lesions) fell within the normal range, despite considerable damage to prefrontal cortex and its subjacent white matter (Figure 1, second column; Figure 2, second and third columns). Mean AER in these patients was 7.9% ⫾ 1.7% ipsilaterally and 9.8% ⫾ 2.4% contralaterally to the lesion side (p ⫽ .98 and p ⫽ .59 compared with control subjects). Mean latency of correct antisaccades was 351 ⫾ 57 msec ipsilaterally and 311 ⫾ 50 msec contralaterally to the lesion side. Mean age of these patients was 44.1 ⫾ 3.5 years (p ⫽ .23 compared with control
C.J. Ploner et al
BIOL PSYCHIATRY 2005;57:1159 –1165 1161
Figure 1. Anatomy of the frontal cortex and lesion drawings. Seven transverse sections parallel above the anterior commissure–posterior commissure line (AC-PC line) from the atlas of Talairach and Tournoux (1988) arranged from dorsal to ventral. Each row corresponds to a distinct horizontal level from ⫹60 mm to ⫹4 mm above the AC-PC line (z-level of the Talairach coordinate system). VAC, vertical anterior commissure line; VPC, vertical posterior commissure line. The black bars in the lower right corner of the figure correspond to a length of 50 mm in right–left and anteroposterior directions (x- and ylevels of the Talairach coordinate system). First column: normal anatomy. R, right; L, left; F1, superior frontal gyrus; F2, middle frontal gyrus; F3, inferior frontal gyrus; PrC, precentral gyrus. Second column: lesions of patients with antisaccade error rates (AERs) within the control range. Third column: lesions of patients with increased AERs. Light gray, lesion; dark gray, overlap of two lesions; nearly black, overlap of three or more lesions. Fourth column: lesion subtraction. Gray, regions affected in patients with increased AERs and spared in patients with AERs within the control range. Arrows point to the anterior and posterior limits of candidate regions for reflexive saccade inhibition.
subjects), mean time since lesion was 15.4 ⫾ 3.2 days, mean lesion volume was 43 ⫾ 9.7 cm3. The AERs of six patients (two with right-sided lesions and four with left-sided lesions), at least contralaterally to the lesion side, clearly fell outside the normal range (Figure 1, third column; Figure 2, fourth and fifth columns). Mean AER in these patients
was 57.7% ⫾ 14.7% ipsilaterally and 75.2% ⫾ 8.2% contralaterally to the lesion side (p ⫽ .001 and p ⬍ .001 compared with control subjects). No significant difference was found between AERs ipsilaterally and contralaterally to the lesion side (p ⫽ .12). Mean latency of correct antisaccades was 524 ⫾ 68 msec ipsilaterally and 562 ⫾ 99 msec contralaterally to the lesion side. Mean age of www.elsevier.com/locate/biopsych
1162 BIOL PSYCHIATRY 2005;57:1159 –1165
Figure 2. Antisaccade error rates (AERs). The gray-shaded area corresponds to the control range. ipsi, AERs ipsilateral to the lesion side; cont, AERs contralateral to the lesion side. The second and third columns show patients with AERs within the control range, the fourth and fifth column show patients with increased AERs. White dots show AERs of patients with rightsided lesions.
these patients was 54.7 ⫾ 4.9 years (p ⫽ .22 difference with control subjects), mean time since lesion was 25 ⫾ 9 days, mean lesion volume was 48.2 ⫾ 6.9 cm3. There was no significant difference in age, time since lesion, and lesion volume between both patient groups (p ⫽ .11, p ⫽ .46, and p ⫽ .33, respectively). Candidate Regions for Reflexive Saccade Inhibition Lesions from patients with AERs within the control range (Figure 1, second column) and lesions from patients with increased AERs (Figure 1, third column) were then subtracted from each other because regions affected in patients with increased AERs but spared in patients with AERs within the control range should contain the region(s) involved in reflexive saccade inhibition. This subtraction yielded several regions in various parts of the prefrontal cortex, which were then transposed on separate sections to define candidate regions for reflexive saccade inhibition (Figure 1, fourth column, gray-shaded regions). It must be borne in mind, however, that a region affected in patients with increased AERs and spared in patients with AERs within the control range is not necessarily involved in reflexive saccade inhibition. For example, if we would have had a large frontoparietal lesion in the patient group with increased AERs, it would not indicate that the parietal cortex is involved in reflexive saccade inhibition. For definition of candidate regions, it is thus important to take the course of the fibers projecting from frontal cortex to the anterior part of the internal capsule into account. It has been shown previously that lesions to this region of the internal capsule affect the corticotectal pathway that controls reflexive saccade inhibition (Condy et al 2004; Gaymard et al 2003b; Leichnetz 1981). Therefore, cortical regions on the sections in the fourth column of Figure 1 were considered candidate regions when lesion to the subjacent caudal and ventral white www.elsevier.com/locate/biopsych
C.J. Ploner et al matter containing the efferent fibers of this region to the anterior limb of the internal capsule yielded increased AERs and vice versa. The course of these fibers was determined on coronal and horizontal brain sections of Déjerine (1895, 1901) and on coronal and sagittal brain sections of Talairach and Tournoux (1988). Fiber tracts were located on the horizontal sections of this study by using anatomic landmarks and the Talairach coordinate system. For example, the efferent fibers of the gray-shaded region in the right precentral gyrus of the ⫹60 and ⫹50 mm sections in the fourth column of Figure 1 must descend laterally from the fundus of the callosomarginal sulcus and the corpus callosum and at least 20 mm laterally from the midline through the ⫹40 mm horizontal section located ventrally of this region. Lesions to this part of the ⫹40 mm section yielded no increase in AERs, however, as can be seen in the second column of Figure 1. Hence, the gray-shaded region in the right precentral gyrus of the ⫹60 and ⫹50 mm sections in the fourth column of Figure 1 is not a candidate region. By contrast, in the right hemisphere, a region on the ⫹20 mm section fulfilled our criteria for a candidate region (Figure 1, fourth column, arrows). This region is largely congruent with the middle portions of the middle frontal gyrus (F2) and extends from about 28 to 60 mm anterior to the vertical anterior commissure line (VAC) line. Lesions to the superior frontal gyrus (F1), the dorsal parts of F2, the inferior frontal gyrus (F3), and the precentral gyrus (PrC) yielded no increase in AER. The ventral limits of this region could not be determined, because there were no patients with increased AERs and lesions on more ventral sections. A corresponding region was found in the left hemisphere on section ⫹20 mm (Figure 1, fourth column, arrows). Here, the candidate region extended from about 10 to 50 mm anterior to the VAC line and included the middle portion of F2 but also parts of F3. Compared with the right hemisphere, however, this region also extended more dorsally to section ⫹28 mm, where it extended from about 19 to 48 mm anterior to the VAC line and included the middle portion of F2 but not F3. Caudally, this region extended to section ⫹12 mm and included ventral parts of F2 and F3. Lesions to F1, to the dorsal parts of F2, and to the PrC yielded no increased AERs.
Discussion This study confirms that unilateral lesions affecting the prefrontal cortex may lead to increased AERs. Nonetheless, even patients with large lesions of lateral prefrontal cortex showed AERs within the control range, as long as their lesions did not affect a region in the mid-DLPFC or the white matter between this region and the anterior part of the internal capsule. Significant increases in AER appeared to be largely independent of lesion volume, postlesion delay, and age of patients. We suggest that the candidate region in DLPFC identified in this study delineates the possible extent of a prefrontal subregion involved in reflexive saccade inhibition. In the following paragraphs, we discuss how these findings relate to previous physiologic and anatomic studies in humans and nonhuman primates. Recently, the cytoarchitectonic parcellation of human prefrontal cortex has been reanalyzed for the left hemisphere (Petrides and Pandya 1999; Rajkowska and Goldman-Rakic 1995). Compared with these maps, the left candidate region of our study comprises the entire constant portion of BA 46 in the middle portions of F2, as well as transitional cortex to BA 9 posterior to this region and areas BA 44 and BA 45 in F3; BA 8, the invariable portion of BA9, and BA10 were not parts of this region. There are no comparable studies for the right hemi-
C.J. Ploner et al sphere, and it is therefore difficult to relate unequivocally the right candidate region of our study to a distinct cytoarchitectonic subdivision of the right prefrontal cortex. It appears, however, that with the exception of the speech-dependent asymmetries of BA 44 and BA 45, there are no major right–left asymmetries for the remaining prefrontal subregions, in particular for BA 46 (Rajkowska and Goldman-Rakic 1995). When the cytoarchitectonic parcellation of the left hemisphere is applied to the right hemisphere, as is regularly done in functional imaging studies, the right candidate region of our study mainly comprises the lower parts of the constant portion of BA 46; BA 8, the invariable portion of BA9, as well as BA10, BA 44, and BA 45 are not part of this region. Therefore, BA 46 is the only region that appears to be affected in right and left candidate regions. In the monkey homologue of this region, injection of anterograde tracers results in labeling of the deep layers of the superior colliculus via pathways through the anterior part of the internal capsule (Leichnetz et al 1981; Selemon and Goldman-Rakic 1988). Furthermore, recent studies have shown that pharmacologic deactivation of a small circumscribed region in Walker area 46 in monkeys results in increased AERs (Gaymard et al 2003a). We thus believe that the findings from our study are entirely consistent with our initial proposal that inhibition of reflexive saccades may depend on a circumscribed DLPFC subregion, which corresponds to BA 46 and which may be the source of a projection onto the superior colliculus where local inhibitory interneurons may finally mediate suppression of unwanted saccades (Condy et al 2004; Munoz and Istvan 1998; Pierrot-Deseilligny et al 1991). When the different angulations of the canthomeatal line and AC-PC line are taken into account (Weiss et al 2003), it is nevertheless obvious that this subregion is located more ventrally in F2 than assumed in our first study on this matter (PierrotDeseilligny et al 1991) and, at least in the right hemisphere, slightly more ventrally than assumed in a subsequent study with lesions transposed on sections parallel to the AC-PC line (PierrotDeseilligny et al 2003). The candidate regions defined in our study were also affected in DLPFC patients of these preceding studies, however. The findings from our study illustrate the limitations of lesion studies in which patient groups are solely based on affected regions of interest (Bates et al 2003). If we would have simply compared performance of all patients with prefrontal lesions to performance of control subjects, there would still have been increased AERs in patients (mean ipsilateral AER: 27.8% ⫾ 8.6%; mean contralateral AER: 36.0% ⫾ 9.2%) and a significant difference between patients and control subjects, at least for contralateral AERs (ipsilateral: p ⫽ .11; contralateral: p ⫽ .02). Hence, almost any prefrontal region of interest would have yielded significant results. Likewise, zones of lesion overlap in patients with increased AERs may be misleading when the subcortical projections of these regions are not taken into account. These points may explain, at least in part, the differences between our results and preceding studies of AERs in patients with lateral prefrontal lesions (Fukushima et al 1994; Guitton et al 1985) and in patients with lesions of the anterior cingulate cortex (Milea et al 2003). The putative oculomotor part of this latter region can be found in the medial frontal cortex on the ⫹28 and ⫹40 mm sections of Figure 1, where it extends several millimeters anterior of the VAC line (Gaymard et al 1998). From inspection of the ⫹12, ⫹20, and ⫹28 mm sections in the fourth column of Figure 1, however, it is clear that any lesion of this region that extends medially and ventrally into the white matter runs the risk of damaging fibers from F2 to the internal capsule. This hypothesis
BIOL PSYCHIATRY 2005;57:1159 –1165 1163 is supported by the fact that small vascular lesions of the anterior cingulate cortex yield no increase in AER, at least with the stimulation parameters of our study (Gaymard et al 1998), whereas more extensive postsurgical lesions caused clear impairments in an antisaccade task (Milea et al 2003). How do our results relate to the conflicting activation patterns observed in functional imaging studies with antisaccade tasks (Doricchi et al 1997; Matsuda et al 2004; McDowell et al 2002; Müri et al 1998; O’Driscoll et al 1995; Paus et al 1993; Raemakers et al 2002; Sweeney et al 1996)? Even when a slight and inevitable intra- and intersubject variance in manual lesion reconstruction is taken into account (Fiez et al. 2000), it is obvious that DLPFC activation in some studies is clearly within (Matsuda et al 2004; Müri et al 1998; Sweeney et al 1996), and in some studies clearly outside the candidate regions of our study (Doricchi et al 1997; McDowell et al 2002). Comparison of task parameters such as target eccentricities, gap–step antisaccade tasks, or control conditions provides no unequivocal explanation for these discrepancies and for the absence of DLPFC activation in other studies (O’Driscoll et al 1995; Paus et al 1993; Raemakers et al 2002). It should be conceded that it is difficult to design tasks that control for fixation demands and number and timing of saccades in an antisaccade task. For example, a corrected antisaccade error consists always of at least two saccades, a fact that renders difficult the comparison of antisaccade to prosaccade control conditions without eye movement recordings. Similarly, fixation control conditions require pronounced and presumably inhibitory fixation-related activity that may be difficult to disentangle from inhibition of reflexive saccades in an antisaccade task (Raemaekers et al 2002). Because the primate DLPFC appears to contain more regions responding to fixation and saccades than assumed previously (Moschovakis et al 2004), different DLPFC activation patterns in functional imaging studies may at least partially result from neuronal activity unrelated to inhibition of reflexive saccades. Event-related task designs with simultaneous eye movement recordings and separate analysis of correct antisaccades and antisaccade errors may finally provide a resolution of these discrepancies. Several lines of evidence have implicated DLPFC subregions in the pathophysiology of schizophrenia (Bogerts 1999; GoldmanRakic and Selemon 1997; Harrison 1999). For example, neuropathologic studies have found evidence for reduced interneuronal neuropil, decreased levels of synaptophysin, and elevated apolipoprotein D levels in BA 46 of schizophrenia patients (Glantz and Lewis 1997; Selemon et al 1998; Thomas et al 2003). The consistently elevated AERs of patients with schizophrenia and their relatives therefore have frequently been attributed to DLPFC dysfunction (e.g., Broerse et al 2001; McDowell and Clementz 2001). The candidate regions defined in our study are in good agreement with this hypothesis and argue against recent hypotheses that relate increased AERs in patients with schizophrenia to a dysfunctional interaction between frontal cortex and basal ganglia rather than to a selective dysfunction of the DLPFC itself (Raemaekers et al 2002). Longitudinal oculomotor studies in patients with DLPFC lesions will have to clarify whether chronic dysfunction of the DLPFC, as observed in schizophrenia, yields increased AERs similarly to our patients, which suffered from relatively acute vascular lesions. If this is the case, our candidate regions may help to define the borders of regions of interest for correlation of neuropathologic changes with increased AERs in patients with schizophrenia. Taken together, the findings from this study strongly suggest that, as in the nonhuman primate, the human DLPFC contains a www.elsevier.com/locate/biopsych
1164 BIOL PSYCHIATRY 2005;57:1159 –1165 distinct subregion that is involved in reflexive saccade inhibition as revealed by standard antisaccade tasks. Compared with recent cytoarchitectonic studies, this region appears to be located bilaterally in BA 46. Although successful performance in an antisaccade task requires additional operations in other frontal regions, in particular triggering of a voluntary saccade by the frontal eye field (Gaymard et al 1999), it is inhibition of reflexive saccades that can be quantified as AER. Increased AERs may therefore provide a simple behavioral marker of BA 46 dysfunction. We therefore hope that the candidate regions of this study provide more detailed anatomic information, allowing for testable anatomic predictions in neuropsychiatric patients with increased AERs.
This study was supported by the Deutsche Forschungsgemeinschaft (Pl 248/2-1). Altman DG (1991): Practical Statistics for Medical Research. London: Chapman & Hall. Back T, Hemmen T, Schüler OG (2004): Lesion evolution in cerebral ischemia. J Neurol 251:388 –397. Bates E, Wilson SM, Pinar Saygin A, Dick F, Sereno MI, Knight RT, Dronkers NF (2003): Voxel-based lesion-symptom mapping. Nature Neurosci 6:448 – 450. Bogerts B (1999): The neuropathology of schizophrenic disease: Historical aspects and present knowledge. Eur Arch Psychiatry Clin Neurosci 249(S4):IV/2–IV/3. Broerse A, Crawford TJ, den Boer JA (2001): Parsing cognition in schizophrenia using saccadic eye movements: A selective overview. Neuropsychologia 39:742–756. Condy C, Rivaud-Péchoux S, Ostendorf F, Ploner CJ, Gaymard B (2004): Neural control of antisaccades: Role of subcortical structures. Neurology 63:1571–1578. Currie J, Ramsden B, McArthur C, Maruff P (1991): Validation of a clinical antisaccadic eye movement test in the assessment of dementia. Arch Neurol 48:644 – 648. Déjerine J (1895): Anatomie des centres nerveux, tome premier (vol. 1). Paris: Rueff. Déjerine J (1901): Anatomie des centres nerveux, tome deuxième (vol. 2). Paris: Rueff. Doricchi F, Perani D, Incoccia C, Grassi F, Cappa SF, Bettinardi V, et al (1997) : Neural control of fast-regular saccades and antisaccades: An investigation using positron emission tomography. Exp Brain Res 116:50 – 62. Everling S, Fischer B (1998): The antisaccade: A review of basic research and clinical studies. Neuropsychologia 36:885– 899. Fiez JA, Damasio H, Grabowski TJ (2000): Lesion segmentation and manual warping to a reference brain: Intra- and interobserver reliability. Hum Brain Mapp 9:192–211. Fischer B, Weber H (1997): Effects of stimulus conditions on the performance of antisaccades in man. Exp Brain Res 116:191–200. Fukushima J, Fukushima K, Miyasaka K, Yamashita I (1994): Voluntary control of saccadic eye movement in patients with frontal cortical lesions and parkinsonian patients in comparison with that in schizophrenics. Biol Psychiatry 36:21–30. Gaymard B, Condy C, Wattiez N, Rivaud-Péchoux S (2003a): Anatomy and pharmacology of antisaccades in monkeys. Program No. 72.18. 2003 Abstract Viewer and Itinerary Planner. Washington DC: Society for Neuroscience. Gaymard B, François C, Ploner CJ, Condy C, Rivaud-Péchoux S (2003b): A direct prefrontotectal tract against distractibility in the human brain. Ann Neurol 53:542–545. Gaymard B, Ploner CJ, Rivaud-Pechoux S, Pierrot-Deseilligny C (1999): The frontal eye field is involved in spatial short-term memory but not in reflexive saccade inhibition. Exp Brain Res 129:288 –301. Gaymard B, Rivaud S, Cassarini JF, Dubard T, Rancurel G, Agid Y, PierrotDeseilligny C (1998): Effects of anterior cingulate cortex lesions on ocular saccades in humans. Exp Brain Res 120:173–183.
www.elsevier.com/locate/biopsych
C.J. Ploner et al Glantz LA, Lewis DA (1997): Reduction of synaptophysin immunoreactivity in the prefrontal cortex of subjects with schizophrenia. Regional and diagnostic specificity. Arch Gen Psychiatry 54:660 – 669. Goldman-Rakic PS, Selemon LD (1997): Functional and anatomical aspects of prefrontal pathology in schizophrenia. Schizophr Bull 23:437– 458. Guitton D, Buchtel DA, Douglas RM (1985): Frontal lobe lesions in man cause difficulties in suppressing reflexive glances and in generating goal-directed saccades. Exp Brain Res 58:455– 472. Hallett PE (1978): Primary and secondary saccades to goals defined by instructions. Vision Res 18:1279 –1296. Harrison PJ (1999): The neuropathology of schizophrenia. Brain 122:593– 624. Karnath HO, Fruhmann Berger M, Kuker W, Rorden C (2004): The anatomy of spatial neglect based on voxelwise statistical analysis: A study of 140 patients. Cereb Cortex 14:1164 –1172. Leichnetz GR, Spencer RF, Hardy SGP, Astruc J (1981): The prefrontal corticotectal projection in the monkey: An anterograde and retrograde horseradish peroxidase study. Neuroscience 6:1023–1041. Matsuda T, Matsuura M, Ohkubo T, Ohkubo H, Matsushima E, Inoue K, et al (2004) Functional MRI of brain activation during visually guided saccades and antisaccades: Cortical and subcortical networks. Psychiat Res 131:147–155. McDowell JE, Brown GG, Paulus M, Martinez A, Stewart SE, Dubowitz DJ, Braff DL (2002): Neural control of refixation saccades and antisaccades in normal and schizophrenia subjects. Biol Psychiatry 51:216 –223. McDowell JE, Clementz BA (2001): Behavioral and brain imaging studies of saccadic performance in schizophrenia. Biol Psychol 57:5–22. Milea D, Lehericy S, Rivaud-Pechoux S, Duffau H, Lobel E, Capelle L, et al (2003): Antisaccade deficit after anterior cingulate cortex resection. Neuroreport 14:283–287. Moschovakis AK, Gregoriou GG, Ugolini G, Doldan M, Graf W, Guldin W, et al (2004): Oculomotor areas of the primate frontal lobes: a transneuronal transfer of rabies virus and [14C]-2-deoxyglucose functional imaging study. J Neurosci 24:5726 –5740. Munoz DP, Istvan PJ (1998): Lateral inhibitory interactions in the intermediate layers of monkey superior colliculus. J Neurophysiol 79:1193–1209. Müri RM, Heid O, Nirkko AC, Ozdoba C, Felblinger J, Schroth G, Hess CW (1998): Functional organization of saccades and antisaccades in the frontal lobe in humans: A study with echo planar functional magnetic resonance imaging. J Neurol Neurosurg Psychiatry 65:374 –377. O’Driscoll GA, Alpert NA, Matthysse SW, Levy DL, Rauch SL, Holzman PS (1995): Functional neuroanatomy of antisaccde eye movements investigated with positron emission tomography. Proc Natl Acad Sci U S A 92: 925–929. Pantono P, Caramina F, Bozzao L, Dieler C, von Kummer R (1999): Delayed increase in infarct volume after cerebral ischemia. Stroke 30:502–507. Paus T, Petrides M, Evans AC, Meyer E (1993): Role of the human anterior cingulate cortex in the control of oculomotor, manual, and speech responses: A positron emission tomography study. Proc Natl Acad Sci U S A 70:453– 469. Petrides M, Pandya DN (1999): Dorsolateral prefrontal cortex: Comparative cytoarchitectonic analysis in the human and the macaque brain and corticocortical connection patterns. Eur J Neurosci 11:1101–1136. Pierrot-Deseilligny C, Müri RM, Ploner CJ, Gaymard B, Demeret S, RivaudPechoux S (2003): Decisional role of the dorsolateral prefrontal cortex in ocular motor behaviour. Brain 126:1460 –1473. Pierrot-Deseilligny C, Rivaud S, Gaymard B, Agid Y (1991): Cortical control of reflexive visually guided saccades. Brain 114:1473–1485. Raemaekers M, Jansma JM, Cahn W, Van der Geest JN, van der Linden JA, Kahn RS, Ramsey NF (2002): Neuronal substrate of the saccadic inhibition deficit in schizophrenia investigated with 3-dimensional eventrelated functional magnetic resonance imaging. Arch Gen Psychiatry 59: 313–320. Rajkowska G, Goldman-Rakic PS (1995): Cytoarchitectonic definition of prefrontal areas in the normal human cortex: II. Variability in locations of areas 9 and 46 and relationship to the Talairach coordinate system. Cereb Cortex 5:323–337. Selemon LD, Goldman-Rakic PS (1988): Common cortical and subcortical targets of the dorsolateral prefrontal and posterior parietal cortices in the rhesus monkey: Evidence for a distributed neural network subserving spatially guided behavior. J Neurosci 8:4049 – 4068. Selemon LD, Rajkowska G, Goldman-Rakic PS (1998): Elevated neuronal density in prefrontal area 46 in brains from schizophrenic patients: ap-
C.J. Ploner et al plication of a three-dimensional, stereologic counting method. J Comp Neurol 392:402– 412. Stuss DT, Knight RT (2002): Principles of Frontal Lobe Function. Oxford: Oxford University Press. Sweeney JA, Mintun MA, Kwee S, Wiseman MB, Brown DL, Rosenberg DR, Carl JR (1996): Positron emission tomography study of voluntary saccadic eye movements and spatial working memory. Proc Natl Acad Sci U S A 75:454 – 468. Talairach J, Tournoux P (1988): Co-Planar Stereotaxic Atlas of the Human Brain. Stuttgart: Thieme.
BIOL PSYCHIATRY 2005;57:1159 –1165 1165 Thomas EA, Dean B, Scarr E, Copolov D, Sutcliffe JG (2003): Differences in neuroanatomical sites of apoD elevation discriminate between schizophrenia and bipolar disorder. Mol Psychiatry 8:167–175. Thompson-Schill SL, Swick D, Farah MJ, D’Esposito M, Kan IP, Knight RT (1998): Verb generation in patients with focal frontal lesions: a neuropsychological test of neuroimaging findings. Proc Natl Acad Sci USA 95: 15855–15860. Weiss KL, Pan H, Storrs J, Strub W, Weiss JL, Jia L, Eldevik OP (2003): Clinical brain MR imaging prescriptions in Talairach space: Technologist- and computer-driven methods. Am J Neuroradiol 24:922–929.
www.elsevier.com/locate/biopsych