ERP Measures of Multiple Attention Deficits Following Prefrontal Damage

INO056 10/18/04 6:34 PM Page 339 CHAPTER 56 ERP Measures of Multiple Attention Deficits Following Prefrontal Damage Leon Y. Deouell and Robert T. Knight ABSTRACT Maintaining a goal-directed behavior requires selectively attending to a subset of the sensory input at the expense of the rest of the input. At the same time, a surveillance mechanism must be in operation, so that deviant or novel events may bring about reorientation of attention and avoidance of potential hazards. Eventrelated potential (ERP) studies in patients with lateral prefrontal damage, due mainly to stroke, reveal deficits in all of these components of the attentional system. These deficits may explain some pervasive neuropsychological impairment in these patients. I. INTRODUCTION Maintaining a goal-directed behavior requires attention, that is, biased processing of a subset of sensory input, motor planning, and thought streams. Multiple goals (from survival to hitting a speeding tennis ball) are present at any one time, requiring a control system that will dynamically prioritize attention allocation based on changing internal goals. While enhancing one stream of input and suppressing another ( voluntary or endogenous attention), the control system needs to allow for changes in the environment, as well as changes in internal drives to interfere with ongoing behavior as needed in a rapid and flexible manner ( involuntary or reflexive attention). Consequently, this control system needs to interact with multiple sensory regions of the brain, as well as with motor output regions involved in orientation, locomotion, and speech. Extensive research using event-related brain potentials (ERPs) established scalp-recorded signatures of voluntary and involuntary attentional processes (see Chapter 84), which may serve as probes into mechanisms of attention in braindamaged patients. Together with behavioral studies, these measures support critical involvement of the lateral prefrontal (LPFC) region in multiple components of attention control. II. SELECTIVE ATTENTION A. Inhibition Selectively attending to an auditory stream, to a visual location or a visual object, and maintaining the goal of behavior requires not only enhancement of processing of some sensory stimuli or motor plans, but also suppression, or inhibition, of competitive stimuli or plans. A biased competition model suggests that excitatory signals to neurons result in inhibition of nearby task-irrelevant neurons, resulting in a sharpening of the attentional focus (see Chapter 50). Patients with lateral prefrontal (LPFC) damage are frequently unable to suppress inappropriate reaction to objects and events in their environment (so-called environmental dependency syndrome (Lhermitte, 1986) or to suppress prepotent responses (Barcelo and Knight, 2002). Whereas some of these effects may be at the level of response selection, ERP studies of these patients suggests that LPFC is essential for normal suppression of information processing at the level of unimodal sensory cortices. Knight and co-workers delivered auditory stimuli (clicks or tone bursts) or brief electric shocks to the median nerve to patients with damage to LPFC and to patients with comparably sized lesions in the temporoparietal junction or the lateral parietal cortex, as well as to age-matched controls. Both types of stimuli were distractors, not relevant to the task. Lesions in the Neurobiology of Attention 339 Copyright 2005, Elsevier, Inc. All rights reserved.

INO056 10/18/04 6:34 PM Page 340 340 CHAPTER 56. ERP MEASURES OF MULTIPLE ATTENTION DEFICITS FOLLOWING PREFRONTAL DAMAGE posterior association cortex, sparing primary sensory regions, had no effects on the amplitudes or latencies of the primary cortical evoked responses. Lesions invading either the primary auditory or somatosensory cortex reduced early latency (20 to 40 ms) evoked responses generated in these regions (Knight, Scabini, Woods, and Clayworth, 1988). In a distinct contrast, LPFC damage resulted in enhanced amplitude of both the primary auditory and somatosensory evoked responses generated from 20 to 40 ms poststimulation (Chao and Knight, 1998; Knight, Scabini, and Woods, 1989; Yamaguchi and Knight, 1990). Spinal cord and brainstem potentials were unaffected by prefrontal damage, indicating that amplitude enhancement of primary cortical responses was due to abnormalities in either prefrontal-thalamic connectivity or in direct interactions between LPFC and sensory cortex. The failure to inhibit irrelevant sensory input indeed affects the patients performance. Chao and Knight (Chao and Knight, 1995) showed that patients with LPFC lesions are impaired in an auditory delayed-match-to-sample task when the delay period is filled with irrelevant tone pips. In the same task, the irrelevant tones elicited abnormally augmented primary auditory potentials (Na, Pa) in the patients, and the performance decrement directly correlated with the Pa enhancement (Chao and Knight, 1998). Similarly, in patients with right prefrontal damage, sounds presented to an unattended ear in a dichotic paradigm abnormally reduced the attentional enhancement (see next section) of subsequent to-beattended sounds (Woods and Knight, 1986). In agreement with these findings, patients with prefrontal lesions show reduced or even reversed negative priming effects (Metzler and Parkin, 2000). B. Facilitation The LPFC seems to be essential not only for suppression of irrelevant information, but also for facilitation of relevant inputs. A series of ERP studies in patients with LPFC damage (centered on Brodmann s areas 9 and 46) suggests that human LPFC regulates extrastriate neural activity through three distinct mechanisms: (1) through a tonic excitatory influence on ipsilateral posterior areas, affecting attended and nonattended sensory inputs alike; (2) by enhancement of extrastriate cortex response to attended information; (3) by a phasic excitatory influence on ipsilateral posterior areas response to correctly perceived taskrelevant stimuli (targets). In these experiments, the patients were asked to detect targets (inverted triangles) in a series of distractors (upright triangles). We refer to this paradigm below as the triangles task. 1. Tonic Excitatory Modulation When the stimuli in the triangles task were presented at fixation, the patients showed a diminished extrastriate N1 (at 170 ms) component relative to controls (Knight, 1997). A similar reduction of N1 was observed in another experiment using visual word stimuli (Swick, 1998). Even more revealing were conditions in which targets and nontargets were lateralized and could appear in either visual field. In one of these conditions, the patients were instructed to attend to all stimuli regardless of their location (Barcelo, Suwazono, and Knight, 2000; Yago and Knight, 2000). The extrastriate P1, immediately preceding the N1 component, was reduced for all stimuli presented to the contralesional hemifield. Moreover, directing the patients attention to one of the fields at a time, revealed that this reduction of P1 over ipsilesional extrastriate cortex was attention independent; it was observed for attended and not attended contralateral stimuli alike (Yago and Knight, 2000). Thus, lateral prefrontal cortex seems to exert a modulatory tonic facilitation over processing at ipsilateral extrastriate cortex. A similar effect of reduced association cortex activity ipsilateral to LPFC damage has been reported for auditory-induced N1 potentials (Chao and Knight, 1998). 2. Attentional Enhancement of a Selected Channel Directing attention to one hemifield in a target detection task normally elicits a sustained negativity starting around 100 ms for all stimuli presented at this hemifield ( Selection Negativity, or Negative Difference ; see Chapter 84). When attention was directed to the contralesional visual field of LPFC patients, this attention effect on extrastriate cortex was normal in the first 200 milliseconds but significantly disrupted thereafter (Yago and Knight, 2000). This finding suggests that LPFC facilitation effect begins around 200 ms after stimulus onset, whereas other cortical areas are responsible for attention-dependent regulation of extrastriate cortex in the first 200 milliseconds. It is conceivable that inferior parietal cortex is responsible for the early reflexive component of attention, whereas LPFC is responsible for more controlled and sustained aspects of visual attention beginning later on. The effect of LPFC lesion on the selection negativity is not limited to visual areas. In an auditory selective attention task, prefrontal patients generated reduced selection negativity as well (Knight, Hillyard, Woods, and Neville, 1981). Notably, this diminished attention effect depended on the side of the lesion. Patients with left hemisphere lesions showed a mildly reduced attention effect regardless of which ear was attended. In con-

INO056 10/18/04 6:34 PM Page 341 III. NOVELTY AND DEVIANCE DETECTION AND INVOLUNTARY ATTENTION SHIFT 341 FIGURE 56.1 Primary cortical auditory and somatosensory event-related potentials for controls (solid line) and patients (dashed line) with focal damage in the lateral parietal cortex (top, n = 8), temporal-parietal junction (middle, n = 13), or dorsolateral prefrontal cortex (bottom, n = 13). Reconstructions of the center of damage in each patient group are shown in the middle. Somatosensory stimuli were square-wave pulses of 0.15ms duration delivered to the median nerve at the wrist. Auditory stimuli were clicks delivered at a rate of 13/s at intensity levels of 50dB HL. Prefrontal damage resulted in a selective increase in the amplitude of the P27 primary somatosensory response and P30 primary auditory response (hatched area). Data from Knight et al., 1988. trast, patients with right prefrontal damage showed reduced effect mainly when the contralateral ear was to be attended. Posterior association cortex lesions in the temporoparietal junction had comparable effects on the selection negativity regardless of ear of stimulation (Woods, Knight, and Scabini, 1993). 3. Target-Related Phasic Activation Target detection in the visual oddball paradigm described above is normally associated with a prominent late response, starting at 200 ms and continuing for the next 500 ms, including the N2-P3b complex. Since this component is elicited only by detected targets, it is considered to be a manifestation of topdown effects. Following LPFC lesions, the N2, a component that likely reflects postselection processing of the target in the inferior temporal lobe, was not observable over the lesioned hemisphere, in response to targets in either visual field (Barcelo et al., 2000). The P3b was reduced over the temporo-occipital electrodes, but not at parietal sites, attesting to the fact that the P3b most likely reflects multiple distinct cortical processes related to target detection (Soltani and Knight, 2000). The patient s performance in this study, producing more errors, was concordant with this electrophysiological evidence of impaired top-down effects following LPFC lesion. A spatially limited reduction of target P3 was recently reported also by Daffner et al. (Daffner et al., 2003), although in their group of LPFC patients the main reduction of target P3 was reported to be in anterior electrodes. Further investigations will be needed to clarify the source of these differences. III. NOVELTY AND DEVIANCE DETECTION AND INVOLUNTARY ATTENTION SHIFT Being able to selectively attend to a subset of the sensory stream and suppress another subset of the input relies on the concomitant operation of a surveil-

INO056 10/18/04 6:34 PM Page 342 342 CHAPTER 56. ERP MEASURES OF MULTIPLE ATTENTION DEFICITS FOLLOWING PREFRONTAL DAMAGE (a) ` (b) ` - (c) - FIGURE 56.2 Visual ERPs in healthy controls and in patients with unilateral right and left lateral prefrontal damage. Patients fixated centrally and were required to detect inverted triangles in a series of upright triangles and novel stimuli presented to the left and right of fixation. In separate blocks, subjects directed attention either to the ipsilesional hemifield or to the contralesional hemifield. All lesions are presented on the left hemisphere. POc/POi Parieto-Occipital recording sites contralateral or ipsilateral to the lesion, respectively. TOc/TOi Temporo-Occipital recording sites contralateral or ipsilateral to the lesion, respectively. (a) Responses to attended and unattended nontargets. Note reduction in P1 amplitude for contralesional stimuli regardless of attention. (b) Significant differences in attention effects (attended minus unattended) between patients and controls start around 200ms post-onset. (c) Responses to targets. Reduced amplitude of N2-P3b. Data from Barcelo and Knight, 2000, and Yago et al., 2001). lance mechanism allowing for salient events to penetrate and shift attention, so that the organism may avoid hazards or take advantage of opportunities. Such events may be slight perturbations of an established sensory regularity (such an intensity decrement in background noise) or grossly unexpected ( novel ) events such as screeching brakes sound. Automatic responses to novel sounds and visual stimuli are typified by a frontoparietal response consisting of the novelty P3a component, peaking around 300 ms fol-

INO056 10/18/04 6:34 PM Page 343 III. NOVELTY AND DEVIANCE DETECTION AND INVOLUNTARY ATTENTION SHIFT 343 lowing the onset of the triggering stimulus. The P3a is considered a marker of attention orienting, elicited by a distributed multimodal corticolimbic system. Automatic response to slight changes in acoustic regularity outside the focus of attention bears the electrophysiological signature of the mismatch negativity (MMN) (Näätänen 1990), a frontal negativity accompanied by lower temporal positivity with a peak latency of 100 to 250 ms following the deviation. The MMN presumably reflects an error signal, generated automatically in the secondary auditory cortex by a neural mechanism comparing a perceived stimulus to a sensory memory trace formed by the regular stimuli, or the process of updating of the existing model of the environment. This signal may be a trigger for ensuing shift of attention toward the deviant event. Both the MMN and the novelty P3a components were shown to be reduced in patients with PFC lesions. A. Deviance Detection In paradigms designed to study the MMN, patient and control subjects watched silent movies or were engaged with a visual reaction time paradigm, and were instructed to ignore series of repetitive sounds. Infrequently, the regularity of the sound stream was broken by the occurrence of a deviant stimulus. The MMN in response to pitch and pattern changes was reduced in patients with PFC lesions, indicating an early deficit in automatic detection of deviance outside the focus of attention (Alain, Woods, and Knight, 1998; Alho, Woods, Algazi, Knight, and Näätänen, 1994). Whereas temporoparietal lesions caused an MMN reduction mainly to contralesional stimuli, LPFC lesions elicited comparable deficits regardless of stimulus side in one study (Alain et al., 1998), and more to ipsilesional sounds in another (Alho et al., 1994). In addition, there was a tendency for right prefrontal lesions to be associated with larger MMN reductions than left frontal lesions, while no such asymmetry was found for the temporoparietal lesions (Alain et al., 1998). These results suggest different contributions of the temporoparietal region and the prefrontal region to the MMN (cf. Deouell, Bentin, and Giard, 1998). Further research will be needed to determine whether the reduced MMN following prefrontal damage reflects a weakened memory trace of the previous regularity due to disinhibition of irrelevant information (see above), reduced frontal facilitation of a comparator mechanism in the secondary auditory cortex which is the main generator of scalp MMN, a failure to initiate an attention switch following the detection of the change, or damage to a postulated frontal generator of MMN. In addition, it is still unclear whether the effect of LPFC damage is independent of the dimension of the regularity that is disturbed (e.g., spatial location, pitch or more abstract dimensions of the sound; cf. Deouell, Bentin, and Soroker, 2000). B. Novelty Detection Novelty P3a responses generated over prefrontal scalp sites to unexpected novel stimuli are reduced by prefrontal lesions, with reductions observed throughout the lesioned hemisphere (Knight, 1997). Comparable P3a decrements have been observed in the auditory (Knight, 1984), visual (Daffner et al., 2000; Knight, 1997), and somatosensory (Yamaguchi and Knight, 1991) modalities in humans with prefrontal damage. Reductions appear to be more severe after right prefrontal damage. In most studies of novelty P3a, the novel stimuli are completely task irrelevant, and the response is considered to reflect an involuntary orienting response toward the salient event. Accordingly, galvanic skin response (GSR), a peripheral marker of the orienting response, is also reduced by damage to the prefrontal as well as posterior association cortex (Tranel and Damasio, 1994). Nevertheless, LPFC damage caused a dramatic reduction in P3a amplitude even in a paradigm in which the novels where relevant to one of two tasks the patients were involved with (Daffner et al., 2003). In this paradigm, patients had to self-pace a succession of visual stimuli, while also looking for a designated (nonnovel) target. LPFC patients not only showed a significantly attenuated P3a response to novel visual stimuli, but also spent less time than controls looking at novels, suggesting a more general decrement of novelty seeking. Taken together, these findings converge with both clinical observations and animal experimentation supporting a critical role of prefrontal structures in the processing of novel stimuli, probably as part as a prefrontalhipocampal network (Kimble, Bagshaw, and Pribram, 1965; Knight and Nakada, 1998). C. Conclusion The behavior of patients with LPFC lesions suggests an interruption of attentional control at multiple levels. Thus, in everyday life patients may inadequately respond to environmental stimuli, get disproportionally distracted by irrelevant information, and are unable to focus on a task on the one hand and unable to flexibly shift from one goal to another (perseveration) on the other hand. The data from electrophysiology reveals the role of LPFC in automatic processes such as maintaining an adequate level of responsive-

INO056 10/18/04 6:34 PM Page 344 344 CHAPTER 56. ERP MEASURES OF MULTIPLE ATTENTION DEFICITS FOLLOWING PREFRONTAL DAMAGE ness to all stimuli, change detection, and novelty processing, as well as controlled processes such as inhibition of irrelevant information, enhancement of processing of relevant information, and phasic responses to targets. Moreover, the electrophysiological data reveals that LPFC implements some of these functions through interaction with remote regions of the brain, including unimodal sensory cortices and the hippocampus. These effects are evident early in the course of processing. The fact that LPFC lesions impair the response to novel stimuli and to unattended changes in the sensory environment seems to be at odds with the finding that processing of irrelevant information is disinhibited at primary sensory cortices. One might have predicted that disinhibition and failure to maintain attentional focus would enhance the response to irrelevant novels and deviants making the patient reorient attention inappropriately. One reason for the reduction of responses to novels and deviants may be a reduction of the signal-to-noise ratio in the system, owing to elevated noise levels. In other words, the relative salience, or novelty, of a given stimulus is reduced when all stimuli are processed and perhaps attract some attentional resources owing to disinhibition of early sensory flow. Alternatively, or concomitantly, this combination of results suggest that deviance detection (as exhibited by the MMN) and novelty processing (as reflected by the P3a) are active processes requiring input from the LPFC, rather than purely local processes within sensory areas. Thus, lesions to the LPFC can at the same time cause disinhibition but still prevent adequate responses to novelty. References Alain, C., Woods, D. L., and Knight, R. T. (1998). A distributed cortical network for auditory sensory memory in humans. Brain Res. 812, 23 37. Alho, K., Woods, D. L., Algazi, A., Knight, R. T., and Näätänen, R. (1994). Lesions of frontal cortex diminish the auditory mismatch negativity. Electroencephalography and Clinical Neurophysiology 91, 353 362. Barcelo, F., and Knight, R. T. (2002). Both random and perseverative errors underlie WCST deficits in prefrontal patients. Neuropsychologia 40, 349 356. Barcelo, F., Suwazono, S., and Knight, R. T. (2000). Prefrontal modulation of visual processing in humans. Nat. Neurosci. 3, 399 403. Chao, L. L., and Knight, R. T. (1995). Human prefrontal lesions increase distractibility to irrelevant sensory inputs. Neuroreport 6, 1605 1610. Chao, L. L., and Knight, R. T. (1998). Contribution of human prefrontal cortex to delay performance. J. Cogn. Neurosci. 10, 167 177. Daffner, K. R., Mesulam, M. M., Scinto, L. F. M., Acar, D., Calvo, V., Faust, R., et al. (2000). The central role of the prefrontal cortex in directing attention to novel events. Brain 123, 927 939. Daffner, K. R., Scinto, L. F. M., Weitzman, A. M., Faust, R., Rentz, D. M., Budson, A. E., et al. (2003). Frontal and parietal components of a cerebral network mediating voluntary attention to novel events. J. Cogn. Neurosci. 15, 294 313. Deouell, L. Y., Bentin, S., and Giard, M. H. (1998). Mismatch negativity in dichotic listening: evidence for interhemispheric differences and multiple generators. Psychophysiology 35, 355 365. Deouell, L. Y., Bentin, S., and Soroker, N. (2000). Electrophysiological evidence for an early (pre-attentive) information processing deficit in patients with right hemisphere damage and unilateral neglect. Brain 123, 353 365. Kimble, D. P., Bagshaw, M. H., and Pribram, K. H. (1965). The GSR of monkeys during orienting and habituation after selective partial ablations of the cingulate and frontal cortex. Neuropsychology 3, 121 128. Knight, R. T. (1984). Decreased response to novel stimuli after prefrontal lesions in man. Electroencephalogr. Clin. Neurophysiol. 59, 9 20. Knight, R. T. (1997). Distributed cortical network for visual attention. J. Cogn. Neurosci. 9, 75 91. Knight, R. T., Hillyard, S. A., Woods, D. L., and Neville, H. J. (1981). The effects of frontal cortex lesions on event-related potentials during auditory selective attention. Electroencephalogr. Clin. Neurophysiol. 52, 571 582. Knight, R. T., and Nakada, T. (1998). Cortico-limbic circuits and novelty: a review of EEG and blood flow data. Rev. Neurosci. 9, 57 70. Knight, R. T., Scabini, D., and Woods, D. L. (1989). Prefrontal cortex gating of auditory transmission in humans. Brain Res. 504, 338 342. Knight, R. T., Scabini, D., Woods, D. L., and Clayworth, C. (1988). The effects of lesions of superior temporal gyrus and inferior parietal lobe on temporal and vertex components of the human AEP. Electroencephalogr. Clin. Neurophysiol. 70, 499 509. Lhermitte, F. (1986). Human autonomy and the frontal lobes. Part II: Patient behavior in complex and social situations: the environmental dependency syndrome. Ann. Neurol. 19, 335 343. Metzler, C., and Parkin, A. J. (2000). Reversed negative priming following frontal lobe lesions. Neuropsychologia 38, 363 379. Näätänen, R. (1990). The role of attention in auditory information processing as revealed by event-related potentials and other brain measures of cognitive function. Behav. and Brain Sci. 13, 201 288. Soltani, M., and Knight, R. T. (2000). Neural origins of the P300. Crit. Rev. Neurobiol. 14, 199 224. Swick, D. (1998). Effects of prefrontal lesions on lexical processing and repetition priming: an ERP study. Cogn. Brain Res. 7, 143 157. Tranel, D., and Damasio, H. (1994). Neuroanatomical correlates of electrodermal skin conductance responses. Psychophysiology 31, 427 438. Woods, D. L., and Knight, R. T. (1986). Electrophysiologic evidence of increased distractibility after dorsolateral prefrontal lesions. Neurology 36, 212 216. Woods, D. L., Knight, R. T., and Scabini, D. (1993). Anatomical substrates of auditory selective attention: behavioral and electrophysiological effects of posterior association cortex lesions. Brain Res. Cogn. Brain Res. 1, 227 240. Yago, E., and Knight, R. T. (2000). Tonic and phasic prefrontal modulation of extrastriate processing during visual attention. Paper presented at the Society for Neuroscience. Yamaguchi, S., and Knight, R. T. (1990). Gating of somatosensory input by human prefrontal cortex. Brain Res. 521, 281 288. Yamaguchi, S., and Knight, R. T. (1991). Anterior and posterior association cortex contributions to the somatosensory P300. J. Neurosci. 11, 2039 2054.