Cross-Modal Matching Performance in Amnesia

Transcription

1 Neuropsychology 1993, Vol. 7, No. 3, In the public domain Cross-Modal Matching Performance in Amnesia Larry R. Squire, Carolyn Backer Cave, and Joyce A. Zouzounis A recent neuropsychological study of patients with postencephalitic amnesia raised the possibility that limbic regions within the medial temporal lobe might be important for forming cross-modal associations. The present study evaluated cross-modal matching in amnesic patients with damage to the hippocampal formation or the diencephalon. The patients with hippocampal formation damage were fully intact at cross-modal matching (both tactile to visual and visual to tactile). Two additional (non-korsakoff's) patients with diencephalic amnesia also performed normally at cross-modal matching. Finally, patients with Korsakoff's syndrome did not exhibit a cross-modal deficit, but they did perform poorly in some task conditions specifically, when a stimulus examined factually had to be matched to a visual stimulus. The results indicate that severe memory impairment can occur without affecting cross-modal performance. Bilateral damage to the hippocampus and related structures in the medial temporal lobe or damage to medial thalamic structures produces severe memory impairment (Damasio, 1984; Markowitsch, 1988; Victor, Adams, & Collins, 1989; Zola-Morgan & Squire, 1993). Although memory impairment is often found together with other disorders of intellectual function, as in depression or dementia, memory impairment can also occur as an isolated cognitive disorder. For example, severe memory impairment can occur against a background of intact short-term memory (Baddeley & Warrington, 197; Cave & Squire, 1992; Drachman & Arbit, 1966) and intact intellectual functions as mea- Larry R. Squire, Psychiatry Service, Department of Veterans Affairs Medical Center, San Diego, California and Department of Psychiatry, University of California, San Diego; Carolyn Backer Cave, Department of Psychology, Vanderbilt University; Joyce A. Zouzounis, Department of Psychiatry, University of California, San Diego. This research was supported by the Medical Research Service of the U.S. Department of Veterans Affairs, National Institute of Mental Health (NIMH) Grant MH 246, the Office of Naval Research, and an NIMH postdoctoral fellowship to Carolyn Backer Cave. We thank Kathleen Fuchs for research assistance and Mark Kritchevsky for assistance with the neurological examinations. Correspondence concerning this article should be addressed to Larry R- Squire, Psychiatry Service (116A), Department of Veterans Affairs Medical Center, 335 La Jolla Village Drive, San Diego, California sured by conventional IQ tests (Scoville & Milner, 1957; Talland, 1965). Severe memory impairment can also occur in the absence of abnormalities in emotional behavior. In nonhuman primates, a double dissociation was demonstrated between the effects of damage to the amygdaloid complex, which is important for emotional behavior, and damage to the hippocampus and related perirhinal, entorhinal, and parahippocampal cortices, which are important.for declarative memory (Squire & Zola-Morgan, 1991; Zola-Morgan, Squire, Alvarez- Royo, & Glower, 1991). Thus, damage to the amygdala impairs emotional behavior but, unless the hippocampus or related medial temporal cortex is also damaged, declarative memory is not affected. Conversely, damage to the hippocampus or related cortex impairs memory but, unless the amygdala is also damaged, emotional behavior is not affected. A similar question about the relation of memory to other cognitive functions arises in the context of recent interest in the capacity for cross-modal matching, the ability to identify a stimulus in one sensory modality on the basis of information presented in a different modality. In one study, amnesic patients with presumed medial temporal lobe damage caused by viral encephalitis were impaired in their ability to match stimuli across sensory modalities (Shaw, Kentridge, & Aggieton, 19). The question remains whether any association exists between cross-modal matching ability and memory function. There are two reasons to be cautious in attributing the impaired performance of the encephalitic patients to medial temporal lobe structures important for memory. First, encephalitis produces variable pathology 375

2 376 L. SQUIRE, C. CAVE, AND J. ZOUZOUNIS that can involve the hippocampus, the amygdala, underlying cortex within the medial temporal lobe, as well as lateral temporal cortex and orbitofrontal cortex (Damasio, Eslinger, Damasio, Van Hoesen, & Cornell, 1985; Damasio & Van Hoesen, 1985; Hierons, Janota, & Corsellis, 1978). Thus, impaired cross-modal matching might be related to neocortical damage or amygdala damage rather than damage to the medial temporal lobe memory system. Shaw et al. (19) emphasized the possible importance of the amygdala. Second, the small disadvantage in cross-modal matching exhibited by the postencephalitic patients when compared with their control subjects might be explained by differences in intellectual ability. The IQ scores of the amnesic patients ranged from 91 to 14 on the Wechsler Adult Intelligence Scale (WAIS), whereas scores on the National Adult Reading Test (NART), which were used to estimate IQ in the control group, ranged from 15 to 125. (Estimates of premorbid IQ for the patients, based on NART scores, ranged from 13 to 11.) In this same study, amnesic patients with Korsakoff's syndrome performed as well on cross-modal matching as did alcoholic control subjects (Shaw et al., 19). In contrast, when the cross-modal matching task included a memory requirement, patients with Korsakoff's syndrome performed worse than alcoholic subjects in some conditions (Oscar-Berman, Pulaski, Hutner, Weber, & Freedman, 19). Another study involving nonhuman primates led to the proposal that the amygdaloid complex is critical for performance on cross-modal tasks (Murray & Mishkin, 1985). In this case, lesions of the amygdala and underlying ventromedial cortex, but not lesions of the hippocampal formation and the parahippocampal gyrus, impaired cross-modal associative memory. This finding leaves open the question as to whether similar lesions also impair cross-modal matching performance per se, that is, when there is no memory requirement in the task. At the same time, because the effective lesion damaged not just the amygdala but also parts of entorhinal and perirhinal cortex (that is, structures that are components of the medial temporal lobe memory system), the findings raise again the issue of whether cross-modal capabilities and memory functions might be associated. Finally, in a single-case study, inspired by the study in monkeys, cross-modal recognition memory was tested in a patient before and after bilateral surgical damage to the amygdala (Lee, Meador, Smith, Loring, & Flanigin, 1988). The patient performed normally, but the test was so easy that performance was nearly perfect both before and after surgery, and ceiling effects could therefore have obscured a real deficit. The question addressed in the present study was whether medial temporal lobe amnesia is or is not independent of cross-modal matching capability. Because the diencephalic brain structures important for memory have anatomical connections to the medial temporal lobe, a parallel assessment of diencephalic amnesia was also carried out. Recent improvements in neuroimaging technology have made it possible to identify bilateral medial temporal lobe and diencephalic pathology in well-characterized amnesic patients with relatively isolated memory impairment (Press, Amaral, & Squire, 1989; Squire, Amaral, & Press, 19). Thus it was possible to assess performance on matching tasks within and between sensory modalities in amnesic patients with confirmed pathology in the medial temporal lobe or in the diencephalon. Amnesic Patients Method We studied 9 amnesic patients (Table 1) who have been tested in our laboratory on many occasions during the past several years. In neurological examination, all of the patients performed normally in tests of light touch, finger localization, and two-point discrimination on the forefinger. Patients with damage to the hippocampal formation. Four of the patients had confirmed (n = 3) or suspected (n = 1) damage to the hippocampal formation. Three patients (P.H., W.H., and J.L.) had participated in magnetic resonance imaging (MRI) studies that demonstrated marked reductions in the volume of the hippocampal formation bilaterally (Press et al., 1989; Squire et al., 19; for P.M., Polich & Squires, in press). The 4th patient (G.D.) was unable to participate in MRI studies. He became amnesic in 1983 following a period of hypotension that occurred during major surgery. Because damage to the hippocampal formation has been demonstrated in other amnesic study patients with a history of anoxia or ischemia (Squire et al., 19; Zola-Morgan, Squire, & Amaral, 1986), it is likely that the pathology in this patient also includes the hippocampus. All 4 patients have been described in detail elsewhere (Cave & Squire, 1991; Kritchevsky & Squire, 1993). These 4 patients averaged 64.8 years of age and had an average of 15.5 years of education. Their Full-Scale IQ (Wechsler Adult Intelligence Scale Revised [WAIS-R]) averaged 19.. Individual IQ and Wechsler Memory Scale Revised (WMS-R) index scores appear in Table 1. Immediate and delayed (12-min) recall of a short prose passage averaged 4.3 and segments, respectively (21 segments total). Scores for other memory tests appear in

3 CROSS-MODAL MATCHING 377 Table 1 Patient Characteristics Patient G.D. a P.H. W.H. J.L. M R.C. N.C. V.F. P.N. J.W. M Age WAIS-R IQ Attention Hippocampal Korsakoff's Verbal formation syndrome WMS-R Visual General Delay 6 57 < < Note. The Wechsler Adult Intelligence Scale-Revised (WAIS-R) and each of the five indices of the Wechsler Memory Scale-Revised (WMS-R) yield a mean score of 1 in the normal population with a standard deviation of 15. The WMS-R does not provide numerical scores for subjects who score below 5. Therefore, the values below 5 were scored as 5 for computing means. a Although the site of this patient's lesion has not been confirmed radiologically, the etiology of the amnesia (ischemia) suggests that damage has occurred to the hippocampal formation. Table 2. Note that the scores on the word recall test in Table 2 are above zero because on this test of immediate recall several items can be retrieved from immediate memory, which is intact in amnesia. The mean score on the Dementia Rating Scale (Mattis, 1976) was (maximum possible = 144; range = 1-136). Most of the points lost on this test were from the Memory subportion (M = 8.3 points lost). The average score on the Boston Naming Test was 55.3 (maximum possible = 6; range = 47-58). Scores for normal subjects on these same tests can be found elsewhere (Janowsky, Shimamura, Kritchevsky, & Squire, 1989; Squire et al., 19). Table 2 Memory Test Performance Patient or control group G.D. P.H. W.H. J.L. M R.C. N.C. V.F. P.N. J.W. M Healthy (n = 8) Alcoholic (n = 8) Diagram recall Paired associates Word recall (%) Hippocampal formation Korsakoff's syndrome Controls Word 5 recognition (%) words Note. The diagram recall score is based on delayed (12-min) reproduction of the Rey-Osterrieth figure (Osterrieth, 1944; maximum score = 36). The average score for copying the figure for all 9 patients was 28.1, a normal score (Kritchevsky, Squire, & Zouzounis, 1988). The paired associate score is the number of word pairs recalled on three successive trials (maximum score = 1 per trial). The word recall score is the percentage of words recalled out of 15 across five successive study-test trials (Rey, 1964). The word recognition score is the percentage of words identified correctly across five successive study-test trials (yes-no recognition of 15 new words and 15 old words). The score for words and faces is based on a 24-hr recognition test of 5 words or 5 faces (modified from Warrington, 1984; maximum score = 5, chance = 25). The mean scores for healthy control subjects and alcoholic control subjects shown for these tests are from Squire and Shimamura (1986) faces

4 378 L. SQUIRE, C. CAVE, AND J. ZOUZOUNIS Patients with damage to the diencephalon. Five of the patients had alcoholic Korsakoff's syndrome (3 men and 2 women). They had participated in either an MRI study (Squire et al., 19) or a quantitative computed tomography (CT) study (Shimamura, Jernigan, & Squire, 1988), which revealed marked reductions in the volume of the mammillary nuclei, reduced density of thalamic tissue, and frontal lobe atrophy. These 5 patients averaged 61.8 years of age and had 11.2 years of education. The average WAIS-R Full-Scale IQ was Individual IQ and WMS-R index scores appear in Table 1. Immediate and delayed (12-min) recall of a short prose passage averaged 5. and segments, respectively (21 segments total; Gilbert, Levee, & Catalano, 1968). Scores for other memory tests appear in Table 2. The mean score on the Dementia Rating Scale (Mattis, 1976) was (maximum possible = 144; range = ), with 5.6 points lost from the Memory subportion of the test and 3.6 points lost from the Initiation-Perseveration subportion. The average score on the Boston Naming Test was 54.8 (maximum possible = 6; range = 48-57). Scores for normal subjects on these same tests can be found elsewhere (Janowsky et al., 1989; Squire et al., 19). Healthy Control Subjects A total of 12 healthy normal subjects (9 men and 3 women) served as control subjects for the amnesic patients with damage to the hippocampal formation. They were either volunteers or employees at the Veterans Affairs Medical Center or were recruited from the University of California, San Diego retirement community and were selected to match the amnesic patients with respect to age (M = 64.8 years; range = 54-75), education (M = 14.8 years; range = 12-17), and two WAIS-R subtest scores: Information (control subjects, M = 22.; amnesic patients, M 21.8) and Vocabulary (control subjects, M = 54.8; amnesic patients, M = 53.). Immediate and delayed (12-min) recall of a short prose passage averaged 7.1 and 5.7 segments, respectively. Alcoholic Control Subjects A total of 6 alcoholic control subjects (4 men and 2 women) served as control subjects for the amnesic patients with Korsakoff's syndrome. They were recruited from alcoholic treatment programs in San Diego County and had abstained from alcohol for at least 2 weeks prior to testing. They were selected to match the patients with Korsakoff's syndrome with respect to age (M = 6.2 years; range = 54-71), education (M = 13. years; range = 11-16), and two WAIS-R subtest scores: Information (alcoholic control subjects, M = 2.2; Korsakoff's patients, M = 19.4) and Vocabulary (alcoholic control subjects, M = 52.; Korsakoff's patients, M = 51.6). Immediate and delayed (12-min) recall of a short prose passage averaged 7.2 and 5.5 segments, respectively. Materials The test was based on one described by Shaw et al. (19) and was developed originally by Nebes (1971). The test stimuli consisted of five black aluminum rings with inner diameters of 1.9, 2.5, 3.2, 3.8, or 4.4 cm and a cross-sectional area of 3 X 3 mm square, and 15 arcs (three different arc lengths 8, 12 C, and 28 and five different diameters 1.9, 2.5, 3.2, 3.8, and 4.4 cm). All rings and arcs were mounted on 1.2-cm square white cards. The tests described always involved three rings and one arc. A wood frame screen, 46 cm high and 49 cm wide, was placed on the table in front of the subject. The lower part of the screen (19.5 cm high) consisted of a black cloth, which contained two openings so that subjects could place either the right or left hand through the cloth without being able to see behind it. Stimulus cards could be displayed upright on the front of the screen at approximately eye level. In addition, a flat wooden tray provided a holder for the stimulus cards that the subject was to touch. The tray was divided into four side-by-side sections (11 X 15.5 cm) with 2.5-cm-high dividers separating each section. Procedure Subjects were tested in two within-modality conditions, tactile-tactile (T-T) and visual-visual (V-V), and two crossmodality conditions, tactile-visual (T-V) and visual-tactile (V-T). Three conditions, T-T, V-V, and T-V, were administered in one test session, with the order of the conditions counterbalanced across subjects. The V-T condition was tested in a second session several months later (M = 4.5 months). The 1.9-, 3.2-, and 4.4-cm rings and arcs were used in the T-T, T-V, and V-T conditions, and the 2.5-, 3.2-, and 3.8-cm rings and arcs were used in the V-V condition (Shaw et al., 19). In this way the difficulty of the V-V condition was made similar to the difficulty of the other conditions. In all four conditions (V-V, T-T, T-V, and V-T), 36 trials were administered. Specifically, each of the three arc lengths (8, 12, and 28 ) was tested four times in a pseudorandom order, using each of three diameters, for a total of 36 trials per condition (3 arc lengths X 3 diameters X 4 trials). T-T condition. Three different rings were placed side by side in the stimulus tray behind the cloth, with the 3.2-cm ring on the left, the 1.9-cm ring in the center, and the 4.4-cm ring on the right. The test arc was placed to the right of the three rings. In the T-T condition as well as in the other three conditions, the test arc had an arc length of 8, 12, or 28 and was constructed with one of the three diameters used to construct the three rings. Subjects placed their preferred hand through the opening in the cloth partition and were instructed to use the tip of their forefinger to touch the test arc and then each of the rings, one at a time. If necessary, the subject's finger was guided to the stimuli. Subjects were free to touch the rings and the test arc as often as they wished until they were prepared to point to the ring that was the same diameter

5 CROSS-MODAL MATCHING 379 as the test arc. No feedback was given about the accuracy of the response. V-V condition. Three rings were displayed side by side on the front of the screen, with the 3.2-cm ring on the left, the 2.5-cm ring in the center, and the 3.8-cm ring on the right. The test arc was placed in the stimulus tray on the subject's side of the screen. Subjects were asked to look at the test arc and the three rings, taking as long as necessary, and then to point to the ring that they judged to be the same diameter as the arc. Subjects were not permitted to touch the rings or the test arc. No feedback was given about their responses. T-V condition. Three rings were displayed side by side on the front of the screen, with the 3.2-cm ring on the left, the 1.9-cm ring in the center, and the 4.4-cm ring to the right. The test arc was placed in the stimulus tray, behind the cloth and out of view. Subjects were not permitted to touch the rings displayed on the front of the screen. Subjects were given as much time as necessary to touch the arc and to look at the rings before pointing to the ring that they thought was the same diameter as the test arc. In other respects, this condition was administered in the same way as the T-T condition. V-T condition. This condition was similar to the T-V condition except that the three different rings were placed side by side in the stimulus tray behind the cloth, with the 3.2-cm ring on the left, the 1.9-cm ring in the center, and the 4.4-cm ring on the right. The test arc was displayed on the front of the screen. Subjects placed their preferred hand behind the cloth to touch the rings and then pointed to the ring that they thought was the same diameter as the test arc. They were not permitted to touch the test arc on the front of the screen. Results The performance of amnesic patients with damage to the hippocampal formation and of normal control subjects was analyzed separately from the performance of patients with Korsakoff's syndrome and of alcoholic control subjects. Patients With Damage to the Hippocampal Formation and Normal Control Subjects The data were first submitted to an analysis of variance (ANOVA) that included the factors of subject group and test condition (four conditions: T-T, V-V, T-V, and V-T; see Figure 1, Panels A and B). Performance differed across the four test conditions, F(3,42) = 42.4, p <.1. However, there was no effect of subject group (F < 1) and no interaction of subject group and test condition (F < 1). In a second ANOVA, arc length was added as a factor (Figure 2). In addition to the effect of test condition, noted earlier, there was also an effect of arc length, F(2, 28) = 19.4, p <.1, and an interaction of test condition and arc length, F(6, 84) = 9.3, p <.1. The interaction reflected the fact that the four test conditions differed with respect to how task difficulty increased as a function of reduced arc length. It is important to note that there was no effect of subject group and no interactions involving the group factor (Fs < 1). Inspection of Figures 1 and 2 confirms that the patients performed normally on all the tests. Patients With Korsakoff's Syndrome and Alcoholic Control Subjects The data were first analyzed with an ANOVA that included subject group and test condition as factors (see Panels C and D of Figure 1). Performance differed across the four conditions F(3, 27) = 12.1, p <.1. Across all four tests, there was no effect of subject group (F < 1.), but there was a marginal interaction of subject group and test, F(3, 27) = 2.6, p =.7. In a second ANOVA, arc length was added as a factor (Figure 3). In addition to the effects just noted, there was also an effect of arc length, F(2, 18) = 47.9, p <.1, and an interaction of test condition and arc length, F(6,54) = 5.2, p <.1. This interaction reflected the same approximate relationship between task condition, arc length, and task difficulty that was observed in the data for the other two groups (Figure 2). There was also a three-way interaction involving subject group, test condition, and arc length, F(6, 54) = 2.9, p <.2. The marginal interaction between test condition and subject group suggested that there were differences between the two groups on one or more of the four conditions. Accordingly, we evaluated the performance of the two groups in each of the test conditions separately using t tests. There was no difference in the two groups in the T-T and V-V tests (Figure 1, Panel C, ts < 1.2). In addition, the groups performed similarly on one of the two cross-modal tasks, V-T (Figure 1, Panel D), t(9) =.35. However, on the other cross-modal task, T-V (Figure 1, Panel D), the patients with Korsakoff's syndrome performed more poorly than the alcoholic control subjects, r(9) = 3.81,/> <.1. This impairment in the T-V condition is also apparent in Figure 3 (lower left). The patients had difficulty at arc lengths of 12, t(9) = 4.22, p <.1, and 28, t(9) = 2.2, p <.6, but they had no difficulty at an arc length of 8 (t< 1.).

6 38 L. SQUIRE, C. CAVE, AND J. ZOUZOUNIS 1 TACTILE-TACTILE VISUAL-VISUAL 1 TACTILE-VISUAL VISUAL-TACTILE A. B. 8 8 o TACTILE-TACTILE VISUAL-VISUAL 1 TACTILE-VISUAL VISUAL-TACTILE C. D. 8 8 o O 7 7 ) o I Figure 1. Matching performance within and between sensory modalities for healthy control subjects (CON) and amnesic patients with damage to the hippocampal formation (AMN; Panels A and B) and matching performance for alcoholic control subjects (ALC) and diencephalic amnesic patients with alcoholic Korsakoff s syndrome (KOR; Panels C and D). Discussion The tests provided rather sensitive measures of cross-modal performance in that they were free of floor and ceiling effects (the performance of normal subjects was always above the chance level of 33% and below the maximum score of 1%). Nevertheless, amnesic patients with damage to the hippocampal formation performed entirely normally both within and between sensory modalities. Thus, bilateral damage to the hippocampal formation sufficient to impair memory had no measurable effect on cross-modal matching ability, either tactile to visual or visual to tactile. In the original study of cross-modal matching by Shaw et al. (19), patients with postencephalitic amnesia were tested only in tactile-visual matching. Those patients performed more poorly than normal control subjects across five test arc lengths and appeared to differ the most from normal subjects in tests involving complete circles (i.e., 36 arc lengths). Accordingly, in a separate session, we tested the 12 normal control subjects and the 4 patients with hippocampal formation damage on tactile-visual matching using 36 arc lengths. Just as with the testing that involved other arc lengths, 12 trials were given (3 ring diameters X 4 trials with each diameter). The results were that the patients scored 97.9% correct and the normal subjects scored 95.8% correct (t < 1.). Considering that these amnesic patients performed normally on all the cross-modal matching tasks, there seem to be two possible ways to understand the deficit in postencephalitic amnesic patients reported by Shaw et al. (19). One possibility is that the deficit was observed because the amnesic and control groups were not well matched in intellectual abilities (see Introduction) rather than because of any deficit in crossmodal matching performance. The second possibility

7 CROSS-MODAL MATCHING TACTILE-TACTILE VISUAL-VISUAL T - 8 I TACTILE-VISUAL VISUAL-TACTILE T 8 8 Figure 2. Matching performance within and between sensory modalities as a function of the arc size of the stimuli being matched for amnesic patients with damage to the hippocampal formation (AMN) and healthy control subjects (CON). is that there was a real impairment in postencephalitic amnesia but that this impairment was not due to damage to the hippocampal formation. The patients with Korsakoff s syndrome in the present study performed well on three of the matching tests, including the V-T matching task, but they were impaired in the T-V condition. This condition required that subjects select one of three rings presented visually that matched a test arc inspected tactually. Because the impairment in this group occurred on only one of the cross-matching test conditions (T-V) and not on the other condition (V-T), the impairment does not appear to constitute a general deficit in cross-matching ability. Moreover, the T-V impairment was observed only for arc lengths of 28 and 12, not for an arc length of 8. In separate tests involving 36 arc lengths, the 5 patients with Korsakoff's syndrome scored 75.% correct, and the 6 alcoholic control subjects scored 93.1% correct. This difference did not reach significance (p >.1). It cannot be the case that the patients simply failed the most difficult task, because they performed normally in several conditions that were at least as difficult as the T-V condition (e.g., the 8 test arcs in all four task conditions). Some task features specific to the T-V task that were not present in the V-T task must be responsible for the deficit. Perhaps the patients with Korsakoff's syndrome had difficulty establishing an effective representation of the test arc when it had to be inspected tactually rather than visually. If they had such a difficulty, it created problems only in the crossmodal task (the T-V condition), not in the withinmodality task (the T-T condition). It seems possible that whatever difficulty the patients with Korsakoff's syndrome had with these tests could have been due to frontal lobe pathology or to the com-

8 382 L. SQUIRE, C. CAVE, AND J. ZOUZOUNIS 1 TACTILE-TACTILE VISUAL-VISUAL - 8 i TACTILE-VISUAL VISUAL-TACTILE « ^ Figure 3. Matching performance within and between sensory modalities as a function of the arc size of the stimuli being matched for alcoholic Korsakoff's syndrome patients (KOR) and alcoholic control subjects (ALC). bination of diencephalic and frontal pathology. Patients with Korsakoff's syndrome have diencephalic damage together with frontal lobe atrophy (Jacobson & Lishman, 1987; Shimamura et al., 1988), and they exhibit a range of cognitive deficits in addition to memory impairment, including visuospatial deficits that can affect perception and the encoding of information (Butters & Cermak, 198; Cave & Squire, 1992; Shimamura, Janowsky, & Squire, 1991; Talland, 1965). Several of these deficits have been linked to frontal lobe pathology and dissociated from the core impairments of amnesia (Cave & Squire, 1992; Janowsky et al., 1989). Two other amnesic patients with radiologically confirmed damage in the medial thalamus (Patients N.A. and M.G.; see Cave & Squire, 1992) were also tested on the cross-modal tasks. These 2 patients averaged 83.3% correct in the T-V condition, 75% correct in the T-T condition, and 85% correct in the V-V condition. (N.A. and M.G. were not tested in the V-T condition). All these scores are normal or above normal. It should be emphasized, however, that these findings are based on only 2 patients, and a larger group of amnesic patients with circumscribed thalamic lesions would need to be tested to confirm that performance is indeed normal in this group. In any case, in the present study the patients with Korsakoff's syndrome were impaired in the T-V condition but not in the V-T condition. If this deficit does reflect something other than a general impairment in cross-modal matching ability, then the present results indicate that neither the hippocampal formation nor the medial thalamus plays a significant role in cross-modal matching and that severe memory impairment can occur without affecting cross-modal performance. It is unclear which brain structures are important for cross-modal matching ability. Reliable and severe deficits in cross-modal performance have not been associated with damage to specific polysensory regions in neocortex, either in humans (McNally, Ettlinger, &

9 CROSS-MODAL MATCHING 383 Smith, 1982) or in monkeys (Streicher & Ettlinger, 1987; Ettlinger & Garcha, 198; Aitken, 198; Horster et al., 1989). Two earlier studies did find cross-modal performance deficits following lesions in prefrontal cortex in the region of the arcuate sulcus (Petrides & Iversen, 1976), but more work is needed to understand fully the nature of this impairment (see Ettlinger & Garcha, 198). The possible importance of subcortical regions was suggested in a study of cross-modal performance in monkeys in which the 2-deoxy-[ 14 C]glucose (2-DG) technique was used (Horster et al., 1989), and the amygdala was implicated in an earlier study in monkeys (Murray & Mishkin, 1985). The latter work could be usefully extended by studying monkeys with lesions limited to the amygdala that spare underlying cortex and also by separating the possible difficulty of acquiring the cross-modal matching rule from the difficulty in cross-modal matching itself. In summary, the present results serve to emphasize the independence of memory functions and crossmodal matching capability and indicate that amnesic patients with damage to either the hippocampal formation or the diencephalon can be normal at crossmodal matching. References Aitken, P. G. (198). Lesions effects on tactual to visual cross-modal matching in the rhesus monkey. Neuropsychologia, 18, Baddeley, A. P., & Warrington, E. K. (197). Amnesia and the distinction between long and short-term memory. Journal of Verbal Learning and Verbal Behavior, 9, Butters, N., & Cermak, L. S. (198). Alcoholic Korsakoff's syndrome: An information processing approach to amnesia. San Diego, CA: Academic Press. Cave, C. B., & Squire, L. R. (1991). Equivalent impairment of spatial and nonspatial memory following damage to the human hippocampus. Hippocampus, J, Cave, C. B., & Squire, L. R. (1992). Intact verbal and spatial short-term memory following damage to the human hippocampus. Hippocampus, 2, Damasio, A. R. (1984). The anatomic basis of memory disorders. Seminars in Neurology, 4, Damasio, A. R., Eslinger, P. J., Damasio, H., Van Hoesen, G. W., & Cornell, S. (1985). Multimodal amnesia syndrome following bilateral temporal and basal forebrain damage. Archives of Neurology, 42, Damasio, A. R., & Van Hoesen, G. W. (1985). The limbic system and the localisation of herpes simplex encephalitis. Journal of Neurology, Neurosurgery and Psychiatry, 48, Drachman, D. A., & Arbit, J. (1966). Memory and the hippocampal complex. II. Is memory a multiple process? Archives of Neurology, J5, Ettlinger, G., & Garcha, H. S. (198). Cross-modal recognition by the monkey: The effects of cortical removals. Neuropsychologia, 18, Gilbert, J., Levee, R., & Catalano, K. (1968). A preliminary report on a new memory scale. Perceptual and Motor Skills, 27, Hierons, R., Janota, I., & Corsellis, J. A. N. (1978). The late effects of necrotizing encephalitis of the temporal lobes and limbic areas: A clinico-pathological study of 1 cases. Psychological Medicine, 8, Horster, W., Rivers, A., Schuster, B., Ettlinger, G., Skreczek, W., & Hesse, W. (1989). The neural structures involved in cross-modal recognition and tactile discrimination performance. An investigation using 2-DG. Behavioural Brain Research, 33, Jacobson, R. R., & Lishman, W. A. (1987). Selective memory loss and global intellectual deficits in alcoholic Korsakoff's syndrome. Psychological Medicine, 17, 649. Janowsky, J. S., Shimamura, A. P., Kritchevsky, M., & Squire, L. R. (1989). Cognitive impairment following frontal lobe damage and its relevance to human amnesia. Behavioral Neuroscience, 13, Kritchevsky, M., & Squire, L. R. (1993). Permanent global amnesia with unknown etiology. Neurology, 43, Kritchevsky, M., Squire, L. R., & Zouzounis, J. A. (1988). Transient global amnesia: Characterization of anterograde and retrograde amnesia. Neurology, 38, Lee, G. P., Meador, K. J., Smith, J. R., Loring, D. W., & Flanigin, H. R (1988). Clinical case report: Preserved crossmodal association following bilateral amygdalotomy in man. International Journal of Neuroscience, 4, Markowitsch, H. (1988). Diencephalic amnesia: A reorientation towards tracts? Brain Research Review, 13, Mattis, S. (1976). Dementia Rating Scale. In R. Bellack & B. Keraso (Eds.), Geriatric psychiatry, (pp ). New York: Grune & Stratton. McNally, C., Ettlinger, G., & Smith, A. M. R. (1982). Crossmodal performance in patients with cerebral lesions. Cortex, 18, Murray, E. A., & Mishkin, M. (1985). Amygdalectomy impairs crossmodal associations in monkeys. Science, 228, Nebes, R. D. (1971). Handedness and the perception of the part-whole relationship. Cortex, 7, Oscar-Berman, M., Pulaski, J. L., Hutner, N., Weber, D. A., & Freedman, M. (19). Cross-modal functions in alcoholism and aging. Neuropsychologia, 28, Osterrieth, P. A. (1944). Le test de copie d'une figure complexe [The test of copying a complex figure]. Archives de Psychologic,,

10 384 L. SQUIRE, C. CAVE, AND J. ZOUZOUNIS Petrides, M, & Iversen, S. D. (1976). Cross-modal matching and the primate frontal cortex. Science, 192, Polich, J., & Squire, L. R. (in press). P from amnesic patients with bilateral hippocampal lesions. Electroencephalography and Clinical Neurophysiology. Press, G. A., Amaral, D. G., & Squire, L. R. (1989). Hippocampal abnormalities in amnesic patients revealed by high-resolution magnetic resonance imaging. Nature, 341, Rey, A. (1964). L'examen clinique psychologic. Paris: Presses Universitaires de France. Scoville, W. B., & Milner, B. (1957). Loss of recent memory after bilateral hippocampal lesions. Journal of Neurology, Neurosurgery, and Psychiatry, 2, Shaw, C., Kentridge, R. W., & Aggleton, J. P. (19). Cross-modal matching by amnesic subjects. Neuropsychologia, 28, Shimamura, A. P., Janowsky, J. S., & Squire, L. R. (1991). What is the role of frontal lobe damage in memory disorders? In H. D. Levin, H. M. Eisenberg, & A. L. Benton (Eds.), Frontal lobe functioning and dysfunction (pp ). New York: Oxford University Press. Shimamura, A. P., Jemigan, T. L., & Squire, L. R. (1988). Korsakoff's syndrome: Radiological (CT) findings and neuropsychological correlates. Journal of Neuroscience, 8, 44-^41. Squire, L. R., Amaral, D. G., & Press, G. A. (19). Magnetic resonance measurements of hippocampal formation and mammillary nuclei distinguish medial temporal lobe and diencephalic amnesia. Journal of Neuroscience, 1, Squire, L. R., & Shimamura, A. P. (1986). Characterizing amnesic patients for neurobehavioral study. Behavioral Neuroscience, 1, Squire, L. R., & Zola-Morgan, M. (1991). The medial temporal lobe memory system. Science, 253, Streicher, M., & Ettlinger, G. (1987). Cross-modal recognition of familiar and unfamiliar objects by the monkey: The effects of ablation of polysensory neocortex of the amygdaloid complex. Behavioral Brain Research, 23, Talland, G. A. (1965). Deranged memory. San Diego, CA: Academic Press. Victor, M., Adams, R. D., & Collins, G. H. (1989). The Wernicke-Korsakoff syndrome and related neurological disorders due to alcoholism and malnutrition (2nd ed.). Philadelphia: F. A. Davis. Warrington, E. K. (1984). Recognition Memory Test. Windsor, England: NFER-Nelson. Zola-Morgan, S., & Squire, L. R. (1993). The neuroanatomy of amnesia. Annual Review of Neuroscience, 16, Zola-Morgan, S., Squire, L. R., Alvarez-Royo, P., & Glower, R. (1991). Independence of memory functions and emotional behavior: Separate contributions of the hippocampal formation and the amygdala. Hippocampus, 1, Zola-Morgan, S., Squire, L. R., & Amaral, D. G. (1986). Human amnesia and the medial temporal region: Enduring memory impairment following a bilateral lesion limited to field CA1 of the hippocampus. Journal of Neuroscience, 6, Received September 9, 1992 Revision received December 9, 1992 Accepted December 1, 1992