Cortical control of saccades

NEUROLOGICAL PROGRESS

Cortical Control of Saccades Charles Pierrot-Deseilligny, MD, Sophie Rivaud, Bertrand Gaymard, MD, RenC Miiri, MD, and Anne-Isabelle Vermersch, MD

A scheme for the cortical control of saccadic eye movements is proposed based partly on defects revealed by specific test paradigms in humans with discrete lesions. Three different cortical areas are capable of triggering saccades. The frontal eye field disengages fixation, and triggers intentional saccades to visible targets, to remembered target locations, or to the location where it is predicted that the target will reappear (i.e., saccades concerned with intentional exploration of the visual environment). The parietal eye field triggers saccades made reflexively on the sudden appearance of visual targets (i.e., saccades concerned with reflexive exploration of the visual environment). The supplementary eye field is important for triggering sequences of saccades and in controlling saccades made during head or body movement (i.e., saccades concerned with complex motor programming). Three other areas contribute to the preparation of certain types of saccades. The prefrontal cortex (area 46 of Brodmann) plays a crucial role for planning saccades to remembered target locations. The inferior parietal lobule is involved in the visuospatial integration used for calculating saccade amplitude. The hippocampus appears t o control the temporal working memory required for memorization of the chronological order of sequences of saccades.

Pierrot-DeseiilignyC, Rivaud S, Gaymard B, Miiri R, Vermersch A-I. Cortical control of saccades. Ann Neurol 1995;37:557-567 Recently substantial progress has been made in the knowledge of the cortical control of saccades, owing to not only experimental results in the monkey, but also human studies using cerebral lesions, transcranial magnetic stimulation (TMS), positron emission tomography (PET), and functional magnetic resonance imaging (MRI). Furthermore, the use of relatively sophisticated saccade paradigms has helped to determine the involvement of new cortical areas in the preparation of these eye movements. Saccades are used to catch rapidly an image of interest on the fovea (the fine central vision). There are different types of saccades. Refexive saccades are externally triggered by the sudden appearance of a visual target on the peripheral part of the retina (reflexive visually guided saccades) (Fig l),or a sudden noise in the immediate environment (reflexive auditory saccades). Intentional (or voluntary) saccades are internally triggered, with a goal. The goal may be to catch on the fovea a target that has been visible on the peripheral part of the retina for a period of time (intentional visually guided saccades) (Fig 1).The goal also may be to find the image of a target not yet or no longer visible: predictive saccades, when the image of the target is expected at a specific location (Fig 2); and memory-guided saccades toward the remembered position of a target perceived a moment before on the peripheral part of the retina (i.e., with visual input), or toward the remembered position to which gaze was

directed before a body rotation (i.e., with vestibular input) (Fig 2). Antisaccades, made in the direction opposite to a suddenly appearing lateral visual target, are also intentional saccades (see Fig 1).Spontaneous saccades are internally triggered, but without a goal, and occur during another motor activity (such as speech), or at rest in darkness. All these saccades are controlled by different cortical areas, which could be partially specialized both in the triggering of a specific type of saccade (e.g., intentional or reflexive saccades) and for those triggering either type, in a specific calculation of saccade amplitude. Such a calculation is made using three different types of coordinate systems depending on the saccade to be performed: In the retinotopic coordinate system, target location is determined in respect to the position of the eye, and the motor vector (i.e., saccade amplitude) is equal to the retinal vector given by the position of the target image on the retina; in the craniotopic coordinate system, target location is determined in respect to the position of the head, and the motor vector is calculated to reach a specific position in the orbit, requiring the use of extraretinal signals; in the spariotopic coordinate system, target location is determined in respect to the position of the body, and the calculation of the motor vector also requires the use of extraretinal signals. Therefore, saccades may be “retinotopic,” “craniotopic,” or “spatiotopic.”

From Service de Neurologie and INSERM 289, HBpital de la SalpCtrii.re, Paris, France.

Address correspondence to Dr Pierrot-Deseilligny, Clinique Paul Castaigne, H B p d de la SalpCtrii.re, 47 Bd de I’Hbpital, 75651 Paris 13’ France’

Received Sep 27, 1994, and in revised form Dec 9, 1994, and Jan 31, 1995. Accepted for publication Feb 1, 1995.

Copyright 0 1995 by the American Neurological Association

557

The cortical areas involved in the triggering or preparation of saccades are reviewed successively.

Cortical Areas Triggering Saccades Three areas in the cerebral cortex are able to trigger saccades, the frontal eye field (FEF), supplementary eye field (SEF), and parietal eye field (PEF). Frontal Eye Field In the monkey, the FEF projects to the deep layers of the superior colliculus (SC), which is an important relay for saccades, and also directly to the premotor reticular formations of the brainstem 11, 21, that is, the saccade generator (Fig 3). The FEF receives multiple cortical afferent tracts, in particular from the PEF, SEF, and prefrontal cortex (PFC), that is, area 46 of Brodmann

c3, 41.

T

c : :

. ....e...

Eye

-200ms

->

!at

1

M

L

Fig I . Paradigms of visually guided saccades, antisaccades, and double-step saccades. (A)Refexive visually guided saccades: (a) gap task, i n which the central fixation Point is switched o f f before the onset of the lateral target; (bj overlap task, in which the central fixation point remains switched on while the lateral target occurs. Latency (lat) and amplitude (am)of visually guided saccades are measured. (Bj Intentional visually guided saccades: The lateral target is already present when the central fixation point is switched off; which is the signal t o pee.foon the saccade. (Cj Antisaccades: Stimulation is the same as in the gap task (a), but the subject is instructed to look in the opposite direction to the lateral target; the percentage of misdirected saccades (ef and the latency of correct antisaccades are determined. (D) Double-step paradigm: The subject is instructed to make two successive saccades in response to two successive,Jashed kzteral targets; the amplitude (a') of the first saccade ( 1) is calculated using retinotopic coordinates (a' = r', the retinal vector of the first target), but the amplitude (a") of the second saccade (2)is calculated using craniotopi6 coordinates (a" = r", i.e., the retinal vector of the second target f a', given by extraretinalsignals). F = central fixation point; G = gap (in stimulation); L = left; M = midline; R = right; T = lateral target.

558 Annals of Neurology Vol 37 No 5 May 1995

The human FEF is located around-the lateral part of the precentral sulcus, involving both the posterior extremity of the middle frontal gyrus and the adjacent precentral sulcus and gyrus, just anteriorly to the motor cortex [5-91. In patients with an FEF lesion, the triggering of different types of intentional saccades is impaired. After a unilateral FEF lesion, memoryguided saccades with visual input (see Fig 1) have bilateral increased latency [lo]. In a predictive paradigm (see Fig 2), the percentage of predictive saccades, occurring before the predictable onset of visual targets, is decreased [1I}. Furthermore, antisaccades, performed away from a suddenly appearing visual target (see Fig l), are delayed after frontal lesions [127, even those limited to the FEF 1111, or after TMS of the FEF region 1131. In contrast, in the gap task of reflexive visually guided saccades (see Fig l),triggering is much less disturbed in patients with an FEF lesion { 11, 147. In this task, the central fixation point is switched off just before (i.e., in general, 200 msec) the onset of the lateral visual target t 151. Therefore, disengagement from current fixation is performed externally just before the occurrence of the visually guided saccade. On the other hand, in the overlap task of reflexive visually guided saccades (see Fig l),in which the central fixation point remains switched on when the lateral target occurs, latency is increased in patients with an FEF lesion [I 1, 161. In this task, active disengagement from the central fixation point is required before the saccade can be triggered. The FEF could be involved in this function, perhaps through its fixation cells 1171 projecting both to the brainstem reticular formations [2) and to the SC 1181, in which other fixation cells exist 1191. It should be noted that an SC lesion also results in increased saccade latency in the overlap task and not in the gap task 1201. Such a result supports the existence of an active role of the SC, occurring after that of the FEF, in the disengagement from fixation. The FEF also appears to control accuracy of retino-

2 Virunlly guided

2

snccnde to F

3 Delny (2s)

Flashed target 0

Switched on

O T fl,

4

3 Virunlly guided

rnccnde to T

Predictive rnccnde to T

4

Memory-guided snccnde

4

Memory-guided saccade

0 Switched off

OT

:I:

. I .

. I .

2

Rotation wifh VOR suppression

3 Delay (5s)

~~

Fig 2. Predictive and memoy-guided saccade paradigm. (A)Predictive saccades. (B) Memoy-guided saccade with uisual input. (C) Memoy-guided saccade with vestibular input. F = central fixation point; T = lateral target.

topic saccades. This applies to situations in which the target is visual and the head remains immobile, that is, to visually guided saccades (reflexive or intentional) (see Fig l), memory-guided saccades (with visual input), and predictive saccades (see Fig 2). In patients with an FEF lesion, accuracy of all these saccades is impaired contralateral to the lesion 1111 (Fig 4). This confirms experimental results suggesting that the FEF controls retinotopic saccades E21, 221. In contrast, amplitude of memory-guided saccades with vestibular in-

Neurological Progress: Pierrot-Deseilligny et al: Cortical Control of Saccades 559

Motor

.Drograms -

T

vc

oc

Reflexive vbually0

guided snccndes

I

YPC L

v

PEF

PFC

L

FEF

SEF

F

lnhibilion of reflexive

b

0

visually-guided spccndes

SC-RF

b

L

r

H

~

Intentional visually-

e

guided saccndes Saccades programmed with extra-retinal signals

L r

L

r

Memory-guided saccades

0

with visual input

L r

Memory-guided saccades with vestibular input

r

L

0

0

+

L

r

-

L

L

.

Predictive saccndes 0

b

L

L

r

r

L

r

b

b b

Sequences of memoryguided saccades (with visual input)

Fig 3 . Hypothetical cortical pathways involved in saccade control. The cortical relays of saccade pathways are represented by arrowheads. FEF = fmntal eye field; H = hippocampal f a m a tion; OC = occipital cortex; PEF = parietal eye field; PFC =

560 Annals of Neurology Vol 37 No 5 May 1995

prefrontal cortex (i.e., area 46 of Brodmann); PPC = posterior parietal cortex; SC and RF = superior colliculus and reticular &nations; SEF = supplementary eye field; SPL = superior parietal lobule; T = thalamus; VC = vestibular cortex.

1

2

3

Fig 4. Example of a patient with a lesion in the ldt frontal q e field. (A) Reflexive visually guided saccades (gap task): Note the marked hypometria of rightward saccades (in which amplitude was calculated using retinotopic coordinates). (B) Double-step paradigm: The first target was flashed for 100 msec and the second target (immediateb afterward and in the opposite direction) for SO mec; targets were randomly presented with an ercentricity of 10 or 20 degrees, left or right; note that the first rightward saccade (single arrows) was markedly hypometric in trials 1 and 4 and normometric in trial S (with significant hypometria on average), and that the second rightward saccade (double arrows) was normometric in trials 2 and G and slightly hypemetric in trial 3 (with normal amplitude on average). Therefore, the j r s t (retinotopic) rightward saccades were hypometric (as were all other types of retinotopic saccades in this patient), whereas the second (craniotopic) rightward saccades were n o m l in amplitude. E = eye movement; L = lejt; M = midline; R = right; T = target. Parts of thisjgure have been published elsewhere {I I ) .

4

5

put (see Fig 2) is not disturbed in patients with an FEF lesion [23]. In this paradigm, there is a body rotation, together with vestibuloocular reflex (VOR) suppression (i.e., with vestibular input), and the subject has to make a memory-guided saccade to the remembered position to which gaze was directed before rotation. In this case, target location has to be determined in respect to body position, and the saccade is spatiotopic. Furthermore, in the double-step paradigm, accuracy of the second saccade was not impaired in a patient with a left FEF lesion 111) (see Fig 4).In this paradigm, the subject has to make two successive saccades, in immediate response to two successive and different visual targets, briefly flashed (see Fig 1). The first saccade is retinotopic. However, the first saccade amplitude has to be taken into account in calculating the second saccade amplitude. Information on the eye position reached after the first saccade is probably transmitted to the cerebral cortex by extraretinal signals originating in the motor or premotor structures of the brainstem (i.e., the efference copy) and passing through the internal medullary lamina of the thalamus E24). Therefore, the second saccade of the double-step paradigm is craniotopic (reaching a specific position in the orbit) and is preserved after a left FEF lesion. The ampli-

Neurological Progress: Pierrot-Deseilligny et al: Cortical Control of Saccades 561

tude of other types of craniotopic saccades, which are performed just after another eye movement (such as the VOR), was preserved in a patient with a left FEF lesion (B. Gaymard and C. Pierrot-Deseilligny, unpublished data, 1995).Thus, the human FEF could control retinotopic saccades but is probably less involved in the control of spatiotopic or even craniotopic saccades. The results of these human lesion studies confirm the findings of monkey stimulation studies 125-281, suggesting that the FEF controls mainly retinotopic saccades. Such dependence of the FEF on visual stimulation is supported by other experimental electrophysiological results showing that the FEF cells are active during natural scanning eye movements, and not during spontaneous saccades in the dark 1291. In summary, the FEF controls (1) disengagement from fixation, ( 2 ) triggering of intentional retinotopic saccades (intentional visually guided saccades, memoryguided saccades with visual input, predictive saccades, and correct antisaccades), and ( 3 ) amplitude of all (i.e., reflexive and intentional) contralateral retinotopic saccades. Finally, it may be hypothesized that the main role of the FEF is to explore the visual environment with intentional saccades, for which the simple retinotopic framework is sufficient to calculate saccade amplitude.

Supplementsy Eye Field Stimulation studies in the monkey have shown that the SEF, located at the anterior part of the supplementary motor area (SMA), is able to trigger saccades 126, 30341. The SEF receives multiple cortical afferent tracts, in particular from the PFC and the posterior part of the cerebral hemisphere 13,353. The SEFprojects to the FEF 141 and like the FEF, to the SC and to the premotor reticular formations 1361, on which it may act directly (without passing through the FEF or the SC) c341. In humans, the SEF lies in the posteromedial part of the superior frontal gyrus, in the SMA region [ 5 , 6 , 8 , 9 ] . Patients with an SEF lesion have no abnormalities of reflexive visually guided saccades or memory-guided saccades with visual input 110, 14, 231, that is, retinotopic saccades. In contrast, in the same patients, there is marked inaccuracy of memory-guided saccades made with vestibular input, that is, spatiotopic saccades [231. In the double-step paradigm (see above), accuracy of the second (craniotopic) saccade could also be impaired in patients with an SEF lesion 1373. It remains to be determined whether all saccades in which amplitude is calculated using extraretinal signals are preferentially controlled by the SEF rather than the FEF (see Fig 3). It should be noted that saccades elicited by stimulation of the SEF in the monkey are goal directed, that is, directed to a specific position in the orbit (whatever the starting position of the saccade), as in the craniotopic 562 Annals of Neurology Vol 37 No 5 May 1995

coordinate system {26, 30-343. Therefore, these experimental findings also suggest that the SEF controls craniotopic saccades. The SEF also appears to control spontaneous saccades 1301. Accordingly, several lines of evidence based on human lesion and experimental stimulation studies suggest that the FEF is more involved in the control of retinotopic saccades, whereas the SEF is more involved in that of craniotopic or spatiotopic saccades. Yet, some electrophysiological results suggested that the FEF and the SEF have similar activity 1381. These different types of experimental and clinical findings are not necessarily conflicting. The specialization of each of these two areas could be only partial. Furthermore, any given cortical ocular motor area involved in the triggering of a specific type of saccades could also be informed of the triggering in other cortical regions of the other types of saccades. Such information could result in an activity within this area, even before the saccade is triggered. However, such an activity does not imply an active role of this area in the triggering. Therefore, simple electrophysiological or even PET studies cannot easily differentiate between the specific roles of the areas concerned, whereas stimulation or lesion studies can determine, through the type of saccade elicited or the deficit observed, which role is predominant or crucial in each area. The SEF also appears to control the chronological aspect of saccade sequences, since the order of these saccades is markedly disturbed when there are left (but not right) lesions affecting this area [39,403. However, it cannot be determined from such lesion studies which precise function is impaired during this relatively complex, multiphase paradigm (Fig 5). First, there is the presentation of different successive lateral visual targets, while the subject fixates a central point. During this initial phase, the subject has to learn only the order of appearance of the visual targets (presentation phase), and not their location, since they remain switched on during the rest of the paradigm. Then, there is a delay of a few seconds during which the subject has to memorize the presentation order (memorization phase). Lastly, the central point is switched off (“go signal”), and the subject has to reproduce the sequence with successive saccades (execution phase), to be performed in the same order as that of presentation. TMS, which lasts a few milliseconds and is supposed to inhibit briefly the stimulated cortical area [41), allows the different phases of this paradigm of saccade sequences to be studied. If such stimulation is performed on the SEF region in normal subjects during this paradigm, the order of saccade sequences is disturbed when stimulation takes place during the presentation phase 142) or at the time of the “go signal” 1431 (see Fig 5). However, when SEF stimulation takes place during the memorization or execution phases, no impairment of

Go signal

- - - - - ------- _ - - - _- -----4 _ - _ - _ -_ _ _ _ _ _ - - - - - - _ _ - - - ---- - - - - -------

- - - - - - _ - -- - - - - - - - - - - - - - -

___

----- _ _ _ _ - _ _ - _ ---------------___-

_______

.iiL,T3 E .-

1

_ _ _ __ __ __ __ __ _______________ __ __ _- -___-------------- --- -- - - - - - - - - ----------

T4

*

A + U %A g

Presentation

Memorization

Execution

*

No Stimulation _ _

_ _ _ a -

_ _ _ _ - - - - - -- - - - - - - - - - - - - - - - - - - - - - t

B F i g 5 . Transcranial magnetic stimulation of the supplementary eye field (SEF) during sequences of memory-guided saccades in normal subjects. (A) Paradigm. During the presentationphase, lateral targets (TI -T4) are switched on successively (and remain illuminated during the rest of the paradigm) while the subject fixates a central fixation point (F). This phase is followed by a dekzy of a few seconds, forming the memorization pbaJe. Lastly, when the central paint is fwitched 08(“go signap), the subject has t o reproduce the sequence with saccades, by looking successively at the different targets, zn the same chronological order as presentation (execution phase). (B) Schematic representation of the errors in the chronology of saccad sequences with or without SEF stimulation. Note that with stimulation during the presentation phase or at the time of the go signal, errorj are near 60%, but with stimulation during the memorization and execution phases, ewors remain near 20% (i.e., at the same level as without stimulation). E = eye movement.

saccade sequences is observed [421. These phases could be controlled by other cortical areas (see below). Therefore, in this paradigm, the SEF appears to be involved during the presentation (i.e., learning) phase and at the instant of initiation of the saccade sequence (i.e., at two different stages in the preparation of such a sequence). Thus, like the SMA for sequences involving limbs [44], the SEF could control motor programs made up of several saccades. In summary, the SEF controls (1) triggering and amplitude of memory-guided saccades with vestibular input (using spatiotopic information); (2) in all probability, amplitude of the second saccade in the double-step paradigm (using craniotopic information); and (3) saccade sequences (during the learning and initiation

phases). Finally, the main role of the SEF could be to prepare motor programs, either combining several intentional saccades or coordinating intentional saccades with other body movements, which require the use of craniotopic or spatiotopic coordinates for calculating saccade amplitude.

parietal E~~ Field The PEF is the lateral intraparietal (LIP) area of the monkey, located in the intraparietal sulcus C45,46]. In stimulation studies that this area the is to saccades[46-491, and lesions at this level result in increased latency of reflexive visually guided saccades [SO]. The PEF projects to the FEF and the SC, but not directly to the brainstem reticular formations involved in saccades [ 3 , 511. The equivalent area in humans could also be located in the region of the intraparietal sulcus, that is, in the superior part of the angular gyrus and supramarginal gyms (areas 39 and 40 of Brodmann) [14, 52). Functional MRI shows that this sulcus is involved during visually guided saccades (C. Pierrot-Deseilligny and colleagues, unpublished data, 1995). Unilateral lesions affecting the posterior parietal cortex (PPC) (including the PEF) result in a bilateral increase in reflexive visually guided saccade latency in the gap task [ 14, 531, as well as in the overlap task, where this effect is even more marked [54, 551 (C. Pierrot-Deseilligny and colleagues, unpublished data, 1995). Increased latency of reflexive visually guided saccades in both overlap and

Neurological Progress: Pierrot-Deseilligny et al: Cortical Control of Saccades 563

gap tasks suggests that their triggering is controlled by the PEF. However, a difference between the gap and the overlap tasks also suggests that the PEF, like the FEF and the SC (see above), is involved in the disengagement from fixation. Therefore, it may be that, upstream of the FEF, the pathway involved in such a function includes the PEF. TMS of the PEF region 1561 and degenerative lesions affecting the parietal lobe 1571, such as those observed in corticobasal degeneration, also result in increased latency of reflexive visually guided saccades. The PEF could facilitate the triggering of reflexive visually guided saccades through its direct projection to the SC. It should be noted that right lesions usually result in more marked impairment of visually guided or memory-guided saccades than left lesions 110, 147. This suggests that there is a degree of cerebral lateralization for these saccades, like that existing for a number of other visuospatial functions 158, 591. The PPC is also involved in lateral shifts of visual attention 1607. However, this influence appears to be different from that controlling visually guided saccade latency, since after human posterior parietal lesions, there is no correlation between visual neglect and increased latency 114, 531. In fact, visual attention could be controlled mainly by the superior parietal lobule in humans C611, which is adjacent to the PEF. Bilateral lesions affecting the PPC (including the PEF) 162) or both this cortex and the FEF 1631 result in Balint’s syndrome and “acquired ocular motor apraxia,” respectively, with severe disturbances of visually guided saccades in both cases (for review, see 1641). In summary, the PEF controls (1) perhaps, disengagement from fixation (upstream of the FEF), and ( 2 ) triggering of reflexive visually guided saccades (in the gap and overlap tasks). Finally, the roles of the PEF and FEF overlap somewhat in the control of visually guided saccades, but it appears that ( 1 ) the PEF plays a predominant role in the triggering of reflexive visually guided saccades (in the gap task), and ( 2 )the FEF plays a major (or even predominant) role in that of intentional visually guided saccades. Thus, the PEF could be more involved in the reflexive exploration of the visual environment, and the FEF more involved in the intentional exploration of this environment (see Fig 3).

Cortical Areas Preparing Saccades Other cortical areas, while not directly involved in saccade triggering, play an important role in the preparation of some types of saccades. Thus, in the posterior hemispheric cortex, the PPC and the vestibular cortex (VC) control saccade accuracy. Furthermore, the dorsolateral PFC and the hippocampal formation could be involved in working memory, probably controlling the memorization period in memory-guided saccades and saccade sequences.

564 Annals of Neurology Vol 37 No 5 May 1995

Posterior Hemispheric Cortex The PPC is equivalent to area 7a in the monkey. This area is involved in visuospatial integration 146, 651 and projects both to the PEF C661 and to the PFC 13, 671. The exact location of this area in humans is not known, but it could also be near the intraparietal sulcus, that is, adjacent to the PEF, in the inferior parietal lobule [lo]. In patients with lesions affecting this region, accuracy of memory-guided saccades with visual input was impaired 1101. In normal subjects, TMS of the PPC region results in a similar impairment of saccade accuracy, but only if stimulation is performed at the very beginning of the memorization period, that is, during the visuospatial integration stage (R. Muri and C. Pierrot-Deseilligny, unpublished data, 1995). In contrast, accuracy of memory-guided saccades with vestibular input was preserved in patients with PPC lesions 1681. This suggests that the PPC is less crucial for vestibular inputs than for visual inputs. Instead of the PPC, the VC, presumably located in the posterior part of the superior temporal gyrus 169, 701, could be involved at the integration stage in memory-guided saccades with vestibular input, since such saccades are impaired when there are lesions affecting this region 1681. Prefrontal Cortex The dorsolateral PFC (i.e., area 46 of Brodmann) is located anteriorly to the FEF C35, 71, 72). In the monkey, this area receives an afferent tract from the PPC and projects to the FEF, SEF, and SC [3, 351. In patients with a PFC lesion, there is a marked increase in the percentage of errors (i.e., unwanted reflexive visually guided saccades) in the antisaccade paradigm {141 (see Fig 1). Such an increase in the percentage of unwanted reflexive visually guided saccades has also been documented in patients with larger focal frontal lesions 1121 or progressive supranuclear palsy C731, that is, with functionally impaired frontal lobe. Therefore, the PFC could control inhibition of these saccades, probably via the SC 120) rather than through the FEF 1141. The dorsolateral PFC E23, 74, 751, but not the PEF 176, 771, is also involved in the control of the spatial aspect of working memory. In memory-guided saccade paradigms, this area holds external integrated information in memory, for a maximum of approximately 15 seconds 1341. In patients with a unilateral left or right PFC lesion, memory-guided saccades with visual or vestibular input have increased latency and impaired accuracy [ l o , 231. TMS of the PFC also impairs memory-guided saccade accuracy 178). Since the PFC appears to memorize information both in retinotopic and in spatiotopic frameworks, a truly spatial memory could be organized in this area. Finally, the PFC appears to be the only common cortical area involved during memory-guided saccades with visual input and those with vestibular input (see Fig 3). The PFC also

may be involved in the control of predictive saccades (see Fig 3). When there are PFC lesions, these saccades are less frequent than in normal subjects (C. PierrotDeseilligny and colleagues, unpublished data, 1995). As similar results are observed in the presence of FEF lesions [l 11, such predictive saccades are probably triggered by the FEF (see Fig 3). Therefore, besides its role at the memorization stage in all types of memoryguided saccades, the PFC may prevent unwanted reflexive saccades, via the SC, and elicit predictive saccades, through the FEF. Even if such an area does not appear capable of triggering saccades by itself, its different actions in eye movement control suggest that it plays a major role in what could be called ocdar motor behavior.

6.

7.

8.

9. 10.

11.

Hippocampal Fomtion TMS during saccade sequences in humans provides results suggesting that the SEF is not crucial during the memorization phase [42] (see Fig 5). Therefore, the chronological aspect of working memory could be stored in memory in a structure other than the SEF. In patients with medial temporal lesions, affecting the hippocampal formation, the chronological order of saccade sequences is markedly disturbed, whereas spatial memory is preserved [791. Therefore, the hippocampal formation, which is indirectly connected to the SEF via the anterior cingulate cortex [SO], could control the chronological aspect but not the spatial aspect (at least during the first few seconds) of working memory. Thus, for saccade sequences, it may be suggested that after visual integration in the PPC, information is sent successively (1) to the SEF for learning the sequence during its presentation 142); (2) to the hippocampal formation for short-term memorization of the chronological aspect 1791; (3)back to the SEF at the beginning of the reproduction stage of the sequence, for initiating the motor program 1431;and finally (4)to the FEF for triggering each of these intentional visually guided saccades (see above) (see Fig 3).

12.

13.

14.

15. 16.

17.

18.

19. 20.

2 1.

22.

References

23.

1. Stanton GB, Goldberg ME, Bruce CJ. Frontal eye field efferent in the macaque monkey. J Comp Neurol 1988;271:473-506 2. Segraves MA. Activity of monkey frontal eye field neurons projecting to oculomotor regions of the pons. J Neurophysiol1992; 68:1967-1985 3. Cavada C, Goldman-Rakic PS. Posterior parietal cortex in rhesus monkey. 11. Evidence for segregated cortico-cortical networks linking sensory and limbic areas with the frontal lobe. J Comp Neurol I989;287: 393-44 5 4. Schall JD, Morel A, Kaas JH. Topography of supplementary eye field afferents to frontal eye field in macaque: implications for mapping between saccade coordinate systems. Visual Neurosci 1993;10:386-393 5. Melamed E, Larsen B. Cortical activation pattern during saccadic

24.

25.

26.

27.

eye movements in humans: localization by focal cerebral blood flow increases. Ann Neurol 1979;5:79-88 Fox FT,Fox JM, Raichle ME, Burde RM. The role of cerebral cortex in the generation of voluntary saccade: a positron emission tomographic study. J Neurophysiol 1985;54:348-369 Godoy J, Liiders H, Dinner DS, et al. Versive eye movements elicited by cortical stimulation of the human brain. Neurology 1990;40:296-299 Petit L, Orssaud C, Tzourio N, et al. PET study of voluntary saccadic eye movements in humans: basal gangliathalamocortical system and cingulate cortex involvement. J Neurophysiol 1993;69:1009-101 7 Kleinschmidt A, Merboldt KD, Requardt M, et al. Functional MRI of cooperative cortical activation patterns during eye movements. SOCNeurosci Abstr 1994;20:1402 Piertot-Deseilligny C, Rivaud S, Gaymard B, Agid Y. Cortical control of memory-guided saccades in man. Exp Brain Res 1991;83:607-6 17 Rivaud S, Miiri RM, Gaymard B, et a]. Eye movement disorders after frontal eye field lesions in humans. Exp Brain Res 1994; 102:llO-120 Guitton D, Buchtel HA, Douglas RM. Frontal lobe lesions in man cause difficultiesin suppressing reflexive glances and in generating goal-directed saccades. Exp Brain Res 1985;58:455-472 Miiri RM, Hess CW, Meienberg 0.Transcranial stimulation of the human frontal eye field by magnetic pulses. Exp Brain Res 1991;86:2 19-223 Pierrot-Deseilliny C, Rivaud S , Gaymard B, Agid Y. Cortical control of reflexive visually guided saccades in man. Brain 1991; 114~1473-1485 Fischer B. The preparation of visually guided saccades. Rev Physiol Biochem Pharmacol 1987;106:1-35 Sharpe JA. Adaptation to frontal lobe lesions. In: Keller EL, Zee DS, eds. Adaptive processes in visual and oculomotor systems. Oxford: Pergamon, 1986:239-246 Petit L, Tzourio N, Orssaud C, et al. Functional neuroanatomy of the human visual fixation system. Eur J Neurosci 1995 (in press) Segraves MA, Goldberg ME. Functional properties of corticotectal neurons in the monkey's frontal eye field. J Neurophysiol 1987;58:1387-1419 Munoz DP, Wurtz RH. Fixation cells in monkey superior colliculus. J Neurophysiol 1993;70:559-589 Pierrot-Deseilligny C, Rosa A, Masmoudi K, Rwaud S. Saccade deficits after a unilateral lesion affecting the superior colliculus. J Neurol Neurosurg Psychiatry 1991;54:1106-1109 Goldberg ME, Bruce CJ. Primate frontal eye fields. 111. Maintenance of a spatially accurate saccade signal. J Neurophysiol 199Q64~489-5 08 Segraves MA, Park K. The relationship of monkey frontal eye field activity to saccade dynamics. J Neurophysiol 1993;69: 1880-1889 Pierrot-Deseilligny C, Israel I, Berthoz A, et al. Role of the different frontal lobe areas in the control of the horizontal component of memory-guided saccades in man. Exp Brain Res 1993; 95:166-1 7 1 Gaymard B, Rivaud S, Pierrot-Deseilligny C. Impairment of extraretinal eye position signals after central thalamic lesions in humans. Exp Brain Res 1994;102:1-9 Bruce CJ, Goldberg ME. Primate eye fields. I. Single neurons discharging before saccades. J Neurophysiol 1985;53:603-635 Mitz AR, Godschalk M. Eye movement represenration in the frontal lobe of rhesus monkeys. Neurosci Lett 1989;106:157162 Schall JD. Neuronal activity related to visually guided saccades in the frontal eye fields of rhesus monkeys: comparison with supplementary eye fields. J Neurophysiol 1991;66:559-579

Neurological Progress: Pierrot-Deseilligny e t al: Cortical Control of Saccades

565

28. Dassonville P, Schlag J, Schlag-Rey M. The frontal eye field provides the goal of saccadic eye movement. Exp Brain Res 1992;89:300-310 29. Burman DD, Segraves MA. Primate frontal eye field activity during natural scanning eye movements. J Neurophysiol 1994; 7 1:1266- 1271 30. SchlagJ, Schlag-Rey M. Evidence for a supplementary eye field. J Neurophysiol 1987;57:179-200 31. Schall JD. Neuronal activity related to visually guided saccadic eye movements in the supplementary motor area of rhesus monkeys. J Neurophysiol 1991;66:530-558 32. Bon L, Luccheti C. The dorsomedial frontal Cortex of the macaque monkey: fixation and saccade-related activity. Exp Brain Res 1992;89:571-580 33. Tehovnik EJ, Lee K. The dorsomedial frontal cortex of the rhesus monkey: topographic representation of saccades evoked by electrical stimulation. Exp Brain Res 1993;96:430-442 34. Tehovnik EJ, Lee KM, Schiller PH. Stimulation-evoked saccades from the dorsomedial frontal cortex of the rhesus monkey following lesions of the frontal eye fields and superior colliculus. Exp Brain Res 1994;98:179-190 35. Goldman-Rakic PS. Circuitry of primate prefrontal cortex and regulation of behavior by representational memory. In: Mountcastle VB, ed. Handbook of physiology, section 1, vol 5 , part I. Bethesda: American Physiological Society, 1987:373-417 36. Shook BL, Schlag-Rey M, SchlagJ. Primate supplementary eye field. I. Comparative aspects of mesencephalic and pontine connections. J Comp Neurol 1990;301:618-642 37. Heide W, Zimmermann E, Kompf D. Double-step saccades in patients with frontal or parietal lesions. SOCNeurosci Abstr 1993;19:427 38. Russo GS, Bruce CJ. Effect of eye position within the orbit on electrically elicited saccadic eye movements: a comparison of the macaque monkey’s frontal and supplementary eye fields. J Neurophysiol 1993;69:800-8 18 39. Gaymard B, Pierrot-Deseilligny C, Rivaud S. Impairment of sequences of memory guided saccades after supplementary motor area lesions. Ann Neurol 1990;28:622-626 40. Gaymard B, Rivaud S, Pierrot-Deseilligny C. Role of the left and right supplementary motor areas in memory-saccade sequences. Ann Neurol 1993;34:404-406 41. Day BL, Rochwell JC, Thompson PD, et al. Delay in the execution of voluntary movement by electrical or magnetic brain stimulation in intact man? Brain 1989;112:649-663 42. Miiri RM, Rivaud S, Vermersch AI, et al. Effects of transcranial magnetic stimulation on the supplementary motor area region during sequences of memory-guided saccades. Exp Brain Res 1995 (in press) 43. Miiri RM, Roessler KM, Hess CW. Influence of transcranial magnetic stimulation on the execution of memorized sequences of saccades in man. Exp Brain Res 1994;101:521-524 44. Tanji J. The supplementary motor area in the cerebral cortex. Neurosci Res 1994;19:251-268 45. Andersen RA. Visual and eye movement functions of the posterior parietal cortex. Annu Rev Neurosci 1989;12:377-403 46. Thier P, Andersen RA. Electrical microstimulation delineates 3 distinct eye-movement related areas in the posterior parietal cortex of the rhesus monkey. SOCNeurosci Abstr 1991;17: 1281 47. Keating EG, Gooley SG, Pratt SE, Kelsey JE. Removing the superior colliculus silences eye movements normally evoked from stimulation of-the parietal and occipital eye fields. Brain Res 1983;269:145-148 48. Shibutani H , Sakata H , Hyvarinen J. Saccade and blinking evoked by microstimulation of the posterior parietal association cortex of the monkey. Exp Brain Res 1984;55:1-8 49. Kurylo DD, Skavenski AA. Eye movements elicited by electri-

566 Annals of Neurology Vol 37 No 5 May 1995

cal stimulation of area PG in the monkey. J Neurophysiol 1991; 65:1243-1253 50. Lynch JC, McLaren JW. Deficits of visual attention and saccadic eye movements after lesions of parieto-occipital cortex in monkeys. J Neurophysiol 1989;61:74-90 51. Lynch JC, Graybiel AM, Lobeck LJ. The differential projection of two cytoarchitectonic subregions of the inferior parietal lobule of macaque upon the deep layers of the superior colliculus. J Comp Neurol 1985;235:241-254 52. Anderson TJ, Jenkins IH, Brooks DJ, et al. Cortical control of saccades and fixation in man; a PET study. Brain 1994;117: 1073-1084 53. Pierrot-Deseilligny C, Rivaud S, Penet C, Rigolet MH. Latencies of visually-guided saccades in unilateral hemispheric cerebral lesions. Ann Neurol 1987;21:138-148 54. Walker R, Findlay JM, Young AW, Welch J. Disentangling neglect and hemianopia. Neuropsychologia 1991;29:1019-1027 55. Kompf D, Heide W. Saccades in parietal lesions. Neuroophthalmology 1994;14(suppl):535 56. Elkington PTG, Kerr GK, Stein JS. The effect of electromagnetic stimulation of the posterior parietal cortex on eyemovements. Eye 1992;6:510-5 14 57. Vidailhet M, Rivaud S, Gouider-Khouja N, et al. Eye movements in parkinsonian syndromes. Ann Neurol 1994;35:420426 58. Mesulam MM. A cortical network for directed attention and unilateral neglect. Ann Neurol 1981;10:309-325 59. Weintraub S, Mesulam MM. Right cerebral dominance in spatial attention. Arch Neurol 1987;44:621-625 60. Posner MI, Walker JA, Fredrich FJ, Rafal RD. Effects of parietal injury on covert orienting of attention. J Neurosci 1984;4: 1863-1874 61. Corbetta M, Miezin FM, Shulman GL, Petersen SE. A PET study of visuospatial attention. J Neurosci 1993;13:1202-1226 62. Pierrot-Deseilligny C, Gray F, Brunet P. Infarcts of both inferior parietal lobules with impairment of visually guided eye movements, peripheral visual inattention and optic ataxia. Brain 1986; 109:81-97 63. Pierrot-Deseilligny C, Gautier JC, Loron P. Acquired ocular motor apraxia due to bilateral frontoparietal infarcts. Ann Neurol 1988;23:199-202 64. Pierrot-Deseilligny C. Saccade and smooth pursuit impairment after cerebral hemispheric lesions. Eur Neurol 1994;34:121134 65. Andersen RA, Bracewell M, Barash S. Eye position effects on visual, memory, and saccade-related activity in areas LIP and 7a of macaque. J Neurosci 1990;10:1176-1196 66. Tian JR, Lynch JC, Hardy SGP. Area LIP input to area 7a in the inferior parietal lobule of macaque. SOCNeurosci Abstr 1991;17:1020 67. Andersen RA, Asanuma C, Essick G, Siege1 RM. Corticocortical connections of anatomically and physiologically defined subdivisions within the inferior parietal lobule. J Comp Neurol 1990;296:65-113 68. Israel I, Rivaud S, Gaymard B, et al. Cortical control of vestibular-derived goal-directed saccades after body rotation in humans. Brain 1995 (in press) 69. Friberg L, Olsen TS, Roland PE, et al. Focal increase of blood flow in the cerebral cortex of man during vestibular stimulation. Brain 1985;108:609-62 3 70. Bottini G, Sterzi R, Paulesu E, et al. Identification of the central vestibular projections in man: a positron emission tomography activation study. Exp Brain Res 1994;99:164-169 71. Miezin FM, Corbetta M, Webb GL, et al. PET activations for visual and memory guided saccades. SOCNeurosci Abstr 1992; 18:699 72. McCarthy G, Blamire AM, Puce A, et al. Functional magnetic

73.

74.

75.

76.

resonance imaging of human prefrontal cortex activation during a spatial working memory task. Proc Natl Acad Sci USA 1994; 91:8690-8694 Pierrot-Deseilligny C , Rivaud S, Pillon B, et al. Lateral visually guided saccades in progressive supranuclear palsy. Brain 1989; 112:471-487 Funahashi S , Bruce CJ, Goldman-Rakic PS. Visuospatial coding in primate prefrontal neurons revealed by oculomotor paradigms. J Neurophysiol 1990;63:814-83 I Funahashi S, Bruce CJ, Goldman-Rakic PS. Dorsolateral prefrontal lesions and oculomotor delayed response performance: evidence for mnemonic “scotornas.”J Neurosci 1993;13:14791497 Chafee M, Goldman-Rakic PS. Prefrontal cooling dissociates

77.

78.

79.

80.

memory- and sensory-guided oculomotor delayed response functions. SOCNeurosci Abstr 19944;20808 Goldman-Rakic P, Chafee M. Feedback processing in prefrontoparietal circuits during memory-guided saccades. SOCNeurosci Abstr 1994;20:808 Brandt SA, Meyer BU, Ploner A, et al. Repetitive transcranial magnetic stimulation over the prefrontal lobe disturbs mernoryguided saccades. SOCNeurosci Abstr 1994;20:1402 Miiri RM, Rwaud S, Timsit S, et al. Role of the medial temporal lobe in the control of memory-guided saccades. Exp Brain Res 1994;101:165-168 Arikuni T , Sako H, Murata A. Prefrontal, premoror, motor and hippocampal connections of area 24 in the macaque monkey. SOCNeurosci Abstr 1991;17:1019

Neurological Progress: Pierrot-Deseilligny et al: Cortical Control of Saccades 567