The Journal October 1986, of Neuroscience @IO): 2950-2967 Human Amnesia and the Medial Temporal Region: Enduring Memory Impairment Following a Bilateral Lesion Limited to Field CA1 of the Hippocampus Stuart Zola-Morgan,* Larry R. Squire,* and David G. Amaralt *Veterans Administration Medical Center, San Diego, California 92161, and Department of Psychiatry, University of California, San Diego, La Jolla, California 92093, and j-The Salk Institute and The Clayton Foundation for Research-California Division, San Diego, California 92138 During the past 100 years clinical studies of amnesia have linked memory impairment to damage of the hippocampus. Yet the damage in these cases has not usually been confined to the hippocampus, and the status of memory functions has often been based on incomplete neuropsychological information. Thus, the human cases have until now left some uncertainty as to whether lesions limited to the hippocampus are sufficient to cause amnesia. Here we report a case of amnesia in a patient (R.B.) who developed memory impairment following an ischemic episode. During the 5 years until his death, R.B. exhibited marked anterograde amnesia, little if any retrograde amnesia, and showed no signs of cognitive impairment other than memory. Thorough histological examination revealed a circumscribed bilateral lesion involving the entire CA1 field of the hippocampus. Minor pathology was found elsewhere in the brain (e.g., left globus pallidus, right postcentral gyrus, left internal capsule), but the only damage that could be reasonably associated with the memory defect was the lesion in the hippocampus. To our knowledge, this is the first reported case of amnesia following a lesion limited to the hippocampus in which extensive neuropsychological and neuropathological analyses have been carried out. Nearly a century ago, von Bechterew (1900) first suggested that memory impairment could occur following damage to the medial temporal lobe of the brain. He presented neuropathological findings from the brain of a 60-year-old man who had memory problems during the last 20 years of his life. The patient’s brain showed bilateral softening of the gyrus uncinatus (the piriform and periamygdaloid cortices and the underlying amygdaloid complex) and the gyrus comu ammonis (the hippocampal formation). A half century later, Grunthal (1947) and Glees and Griffith (1952) published separate case studies linking defects in memory to medial temporal damage. Both cases resulted from a vascular accident that bilaterally damaged the hippocampus and surReceived Dec. 9, 1985; revised Feb. 28, 1986; accepted Mar. 3, 1986. This research was supported by the Medical Research Service of the Veterans Administration by NIH Grants NS19063 (to S.Z.-M.) and NS20004 (to D.G.A.), and bv NIMH Grant MH246bO fto L.R.S.1. This work was conducted in nart bv the Clayton Foundation for Research, California Division (D.G.A. is a Clayto; Foundation Investigator). We thank Cecilia Le Clair, Carol Micheletti, and Michael Mesches for neurohistological preparation of the brain tissue; Kris Trulock and Marcia Lee Earnshaw for photographic assistance; Dr. Robert Terry for critical discussion and advice based on his analysis of the brain sections; Drs. Henry Powell and Thomas Kemper for examining sections of the brain; Dr. William Pulsinelli for his comments on this case; Drs. Leon Thai and Mark Kritchevsky for assistance in reviewing medical and hospital records; and Joyce Zouzounis and Jasmine Allen for research assistance. We are particularly indebted to Dr. Carlos Avendafio, who carried out a preliminary analysis of some of the histological material. Finally, we thank the patient’s family for sustained encouragement of this work. Correspondence should be addressed to Stuart Zola-Morgan at the above address. Copyright 0 1986 Society for Neuroscience 0270-6474/86/102950-18$02,00/O rounding cortex. Other brain regions, including the amygdala and the mamillary bodies, were reported to be normal. The evidence for impaired memory in these 2 cases was anecdotal and based on clinical impressions. Nevertheless, both patients seemed to have had considerable anterograde memory problems, and one seemed to have had retrograde amnesia covering several years (Glees and Griffith, 1952). In 1954, Scoville described a case of severe anterograde amnesia that followed bilateral medial temporal lobe resection in a patient with intractable seizures. The removal extended posteriorly along the medial surface of the temporal lobes for a distance of approximately 8 cm from the anterior poles, probably destroying the anterior two-thirds of the hippocampus and hippocampal gyrus bilaterally, as well as the amygdala and periamygdaloid cortex (Scoville, 1954; Scoville and Milner, 1957). This case (H.M.) has been studied extensively during the past 30 years (Corkin, 1984). Formal memory testing of H.M. and 9 other patients with less extensive bilateral medial temporal lobe removal led to the view that damage to the hippocampus was responsible for the amnesia (Scoville and Milner, 1957). Memory impairment was observed whenever the hippocampus and hippocampal gyrus were damaged bilaterally. A patient whose resection included only the amygdala and uncus (presumably the periamygdaloid cortex), while sparing the hippocampus, was not amnesic. The importance of hippocampal damage in memory impairment was further supported by the finding that patients with left or right unilateral medial temporal lobe resections exhibited verbal or nonverbal memory deficits, respectively. The severity of the deficits correlated tiith the extent of hippocampal damage (Milner, 1972). The hippocampus has also been linked to memory impairment in cases of viral encephalitis (Damasio et al., 1985a; Rose and Symonds, 1960), posterior cerebral artery occlusion (Benson et al., 1974), and hypoxic ischemia (Volpe and Hirst, 1983). In addition, impaired memory has been attributed to hippocampal lesions in several single case studies (Cummings et al., 1984; DeJong et al., 1968; Duyckaerts et al., 1975; Go1 and Faibish, 1967; Muramoto et al., 1979; Van Buren and Borke, 1972; Woods et al., 1982). However, the assessment of memory functions in these cases was often based on anecdotal reports (but see Damasio et al., 1985a). In addition, damage in many cases was not confined to the hippocampus but extended into the amygdala and parahippocampal gyrus. In other cases, only cursory neuropathological analysis was carried out. Thus the human cases have, until now, left some uncertainty as to whether lesions limited to the hippocampus are sufficient to cause amnesia. Alternatives to the view that hippocampal damage is sufficient to cause memory dysfunction have also been proposed. For example, Horel(1978) pointed out that none of the human cases available through the 1970s had well-documented lesions lim- 2950 The Journal of Neuroscience Memory Impairment Following Bilateral ited to the hippocampus.He proposedthat the “temporal stem,” a fiber systemthat lies superficial to the hippocampusand that links the temporal neocortex with subcortical regions, was the critical structure damagedin medial temporal lobe amnesia.He suggestedthat this fiber systemwould necessarilybe damaged by the surgical procedurescarried out during anterior temporal lobectomy. An alternative view was advanced in the sameyear by Mishkin (1978) who pointed out that all human surgical casesexhibiting amnesiahad damageto both the hippocampus and amygdala.Accordingly, the evidence from the surgicalcases is asconsistentwith the view that conjoint hippocampal-amygdaloid damageis neededto produce amnesiaas with the view that hippocampal damagealone is sufficient. Recent successes at developing an animal model of human amnesiain the monkey (Mahut and Moss, 1984;Mishkin et al., 1982; Squire and Zola-Morgan, 1983) have provided a way to addresstheseissuesexperimentally. Monkeys with lesionslimited to the temporal stem were not amnesic(Zola-Morgan and Squire, 1984; Zola-Morgan et al., 1982), and monkeys with conjoint hippocampal-amygdaloid lesionswere amnesiceven when the lesionscausedno damageto the temporal stem(Mahut et al., 1982; Moss et al., 1981;Zola-Morgan et al., 1982).Therefore, damageto the temporal stem doesnot seemto be a factor in memory impairment. The sameconclusion about the temporal stemwas reachedin a recent clinicopathologic casestudy of a patient who had become amnesic following an ischemic episode(Cummings et al., 1984). Studies in monkeys have also compared the amnesiceffects of separatehippocampal lesionswith the effects of combined hippocampal-amygdaloid lesions(Mahut et al., 1982; Mishkin, 1978; Squire and Zola-Morgan, 1985; Zola-Morgan and Squire, 1985). While hippocampal lesionsproduced a statistically significant impairment, the combined lesion produced an even larger deficit. Thus, the results with experimental animals suggestthat hippocampal lesionsalone produce a measurableimpairment in memory, though the severity of the deficit is increasedwhen the amygdala and surrounding cortical areasare included in the ablation. Theseresultsfrom monkeysleave somequestionasto whether a patient with damagelimited to the hippocampus would demonstrate a memory impairment. Would the deficit be substantial or only minor? Recently, we obtained extensive clinicopathological information from a patient who developed a clinically significant memory impairment as the result of an ischemic episode. This patient survived for 5 years following the episode,during which time he participated in our neuropsychological studiesof memory and amnesia.We report here a well-documentedcaseof amnesiawith neuropathologicalevidenceof bilateral damageto the hippocampus. Materials and Methods Case history: patient R.B. Pre-operative history. R.B. was a white male postal worker, retired in 1976 at 50 years of age because of medical disability. In 1960, he had experienced shortness of breath and chest pain, which was diagnosed as angina. In 1970, an electrocardiogram done at rest and during physical exercise was normal. In August 1976, he again developed shortness of breath with right chest pain and was hospitalized in an Intensive Care Unit for 6 d. On the basis of angiographic findings, coronary artery bypass surgery was performed 1 month later, with 1 graft to his circumflex artery. Angina1 episodes continued during the next 2 years. In July 1978, an angiogram showed that the graft was occluded, and a second bypass surgery was recommended. Up to this point there had been no history of neurologic problems except several transient episodes of hypoesthesia of the anterior right thigh. Surgery and perioperative event. In September 1978, R.B., now 52 years of age, was hospitalized, and on September 18 a second coronary artery bypass operation was performed, this time on the left anterior Hippocampal CA1 Lesion 130r n 2951 WA/S R.B AMN N.A. KORS R.B. AMN N.A KORS Figure 1. Top panel shows full-scale Wechsler Adult Intelligence Scale scores ( WAZS) and Wechsler Memory Scale scores ( WMS) for patient R.B., 3‘ patients with amnesia due to anoxia or ischemia (AMZV), case N.A.. and 8 natients with Korsakoff svndrome (KORS). In normal subjects,’ the W-MS is equivalent to the- WAIS. jottom panel shows the average WAIS-WMS difference score for the same patients. descending circumflex artery and the posterior descending artery. The surgery was uneventful, and the patient was transferred to the surgical Intensive Care Unit, where his condition remained stable throughout the early evening. He was noted to be awake and able to move all 4 limbs in response to commands. At about 1:OO AM on September 19, a sudden atria1 tear resulted in the loss of 800-1000 cc of blood in less than 2 min. A systolic pressure of 40 mm was recorded at this time. His chest was opened and internal cardiac massage initiated. Manual massage was continued while the patient was transferred to the operating room. His electrocardiogram was flat at this time, and his pupils were 2 mm in diameter and fixed. The atria1 tear was repaired, and cardiac rhythm and blood pressure were eventually brought to within normal limits with the help of a pacemaker and an intraaortic balloon pump. It was estimated that a total of 5000 cc of blood was lost. At about 9:00 AM, the patient was returned to the Intensive Care Unit. He was ventilated artificially and given a neuromuscular blocking agent (pancuronium bromide) intermittently for several days to control chest muscle fasciculation and shivering. During that period he was generally unresponsive, presumably because of the neuromuscular blockade, but there were times between drug administrations when he appeared to understand verbal stimuli and opened his eyes on request. Postoperative course. On September 20, R.B. was returned to the operating room after a routine chest x-ray suggested further bleeding. In the operating room he had 2 arrhythmias, and a systolic pressure of 90 mm was recorded. For the next 4 hr pressor drugs were adjusted, cardiac output eventually normalized, and the patient was returned to the Intensive Care Unit, where his blood pressure ranged between 80110/45-60 mm. At about 6:00 PM on the same evening, the patient had an episode of supraventricular tachycardia. His blood pressure dropped to 60/40 mm. He was cardioverted 9 x , and a heart rate of 110 beats/min was established. During this period, the patient was noted to have fixed pupils, 5-6 mm in diameter. On September 21, an electroencephalogram (EEG) showed diminished amplitude and continuous slow delta activity. He was able to open and close his eyes to command, but there was no correlated change in the EEG. Over the next 2 d, the patient’s need for pressor agents decreased. On September 23, lidocaine and pancuronium bromide were discontinued and the patient was able to breath without mechanical assistance. Later that day, he opened his eyes when his name was called and appeared to understand what was being said to him. That evening he attempted to talk and was able to move all extremeties. He could move 2952 Zola-Morgan PAIRED ASSOCIATES 25 R.B AMN N.A. STORY 7r KORS I ALC - CON RECALL 6 - 25 DIAGRAM r RB. AMN N.A. RECALL KORS ALC CON Figure 2. Performance on 3 tests of new learning ability by patient R.B., 3 other amnesic groups (see Fig. l), alcoholic control subjects (ALC; N = 6), and healthy control subjects (CON, N = 6). For the patients, the score is the average for 3 separate test administrations. R.B. was tested 6 months, 9 months, and 23 months after the onset of amnesia. his fingers but could not grasp. On the following day he recognized his son. At this time, reduced grip strength was noted in the left hand, and paresthesia was noted in the left hand and forearm. On September 24, the balloon pump, the right atria1 and intracardiac lines, and the chest tubes were removed. He could respond readily to verbal stimuli and could nod appropriately. On September 25, he asked for a drink ofwater and asked where he was. On the next day, it was noted that he appeared to forget things easily. On September 29, the patient suffered an apparent respiratory arrest and was intubated and respirated artificially. During this episode, blood pressure ranged between 60/40 mm and 1 lo/60 mm. He required mechanical ventilation over the course of the next month, although he was reported to be alert and responsive during this time. On October 30, he was successfully weaned from the respirator. At this point 2 chart entries indicated that he was disoriented as to time and place. On November 11, he was moved from Intensive Care to a medical ward for 2 weeks of convalescence. A neurological exam on November 12 noted impaired memory: he could not recall or recognize any of 3 words presented a few minutes previously. He had poor recall of current events as well as of the surgical events of the previous weeks. He was less impaired on questions about the distant past. Spontaneous speech, naming, reading aloud, and comprehension were judged to be normal. He was discharged on November 2 1, 1978. et al. Vol. 6, No. 10, Oct. 1986 Patient R.B. was hospitalized 5x during the following 5 years. He suffered shortness of breath, edema, decreased exercise tolerance, as well as atria1 flutter. In each case, myocardial infarction was ruled out. His blood pressure was never lower than 90/70 mm during any of these short hospital stays. On September 22,1983, R.B. suffered a fatal cardiac arrest. Cause of death was reported as end-stage congestive heart failure secondary to severe and widespread atherosclerosis. Neuropsychological evaluations. In March 1979, about 6 months after his surgery, R.B. was given a battery of standard neuropsychological tests as part of his follow-up evaluation. He obtained a full-scale WAIS IQ of 1l 1 (verbal: 108; pe;fonnance: 114), and he was considered to have severe impairment on tests of verbal and nonverbal memory functions (prose passage: immediate recall, 5 segments; delayed [20 min] recall, 0 segments; Rey-Osterreith figure: copy, 30136 segments; delayed [20 min] recall, l/36 segments). During the testing session, he stated that he had severe memory problems. He explained that he needed to ask his wife repeatedly to tell him what had gone on and, if he talked to his children on the phone, he did not remember anything about it the following day. He said that he did notice some improvement in his memory and was able to recall a few things from television or the newspaper. At this time, through discussions with staff of the Neurology and Psychology Services, we became aware of the patient and his memory problem. During the next 4 years, we evaluated R.B.‘s performance on a variety of tests of anterograde and retrograde amnesia, and compared his performance to that of other amnesic study patients and control subjects. To facilitate interpretation, quantitative data for patient R.B. are presented here, together with comparable data for 3 other kinds of amnesic patients: (1) Patients with Korsakoff syndrome: These 3 women and 5 men with amnesia have been tested as a group for several years. They have 12.4 years of forma1 education (range = 12-14 years) and averaged 50.5-52.3 years of age at the time of the testing summarized here. (For additional information about this group, see Shimamura and Squire, 1984.) (2) Case N.A.: This man has been amnesic for verbal material since 1960 when, at the age of 22, he sustained a stab wound to the brain from a miniature fencing foil (Kaushall et al., 198 1; Teuber et al., 1968). Radiographic evidence identified a lesion in the left dorsal thalamus (Squire and Moore, 1979). He has had 13 years of forma1 education. The testing summarized here was done when he was 40-47 years of age. (3) Ischemia and anoxia: These 3 males have amnesia as the result of ischemia or anoxia. Case 1: Cardiac arrest and anoxia in 1976. Testing occurred 5-8 years after the episode. His age at completion of testing was 46 years, and he had received 20 years of education (Ph.D. in clinical psychology). Case 2: Hypotensive episode during surgery in 1983. Testing occurred l-2.5 years after the episode. His age at completion of testing was 44 years, and he had received 13 years of education. Case 3: Respiratory arrest and anoxia during epileptic seizure in 1984. Testing occurred 7-13 months after the episode. His age at completion of testing was 55 years, and he had received 15 years of education. Figure 1 shows the full-scale Wechsler Adult Intelligence Scale (WAIS) IQ and the Wechsler Memory Scale (WMS) score for each kind of amnesic patient. The difference between the WAIS and the WMS provides 1 index of the severity of memory impairment. In normal subjects, the WMS score is equivalent to the WAIS IQ. In March 1979,6 months after his episode, R.B.‘s WAIS was 111, and his WMS was 91. Figure 2 shows the scores for 3 tests of new learning ability that assess the severity of anterograde amnesia. The data for the amnesic patients are average scores for 3 separate administrations of each test (except for N.A.‘s diagram recall score, which is based on a single test). R.B. was tested at 6 months, 9 months, and 23 months after the onset of his amnesia. For paired associate learning, subjects were given 3 consecutive trials to learn 10 unrelated word pairs (Jones, 1974). R.B. recalled 3.7 pairs during the 3 trials (maximum = 30). For story recall, subjects were read a short prose passage (Gilbert et al., 1968) and asked to recall it immediately after hearing it and again, without forewarning, 1O-l 2 min later. R.B. recalled an average of 5 segments of the passage immediately but only 0.7 segments after the delay. For diagram recall, subjects copied a complex figure and then, without forewarning, were asked to reproduce it from memory 10-20 min later. We used the Rey-Osterreith figure (Osterreith, 1944), the Taylor figure (Milner and Teuber, 1968), and a similar complex figure developed in our laboratory. The maximum score for each figure was 36. R.B. averaged 29 points for his copy and 3 for his reproduction. Note that patient N.A., who has a left diencephalic lesion, performed much better on diagram recall than on the other two The Journal of Neuroscience Memory Impairment Following Bilateral Hippocampal CA1 Lesion Control Figure 3. Performance by patient R.B. on 2 separate administrations of the Rey-Osterreith complex figure test. R.B. was first asked to copy the figure illustrated in the small box to the left. Then, without forewarning, he was asked to reproduce the figure from memory 10-20 min later. Left panel, R.B.‘s copy (top) and reproduction (bottom) 6 months after the onset of his amnesia. Middle R.B.‘s copy and reproduction 23 months after the onset of amnesia. Right, Copy and reproduction from 1 of the 6 healthy control subjects matched to R.B. for age and education. verbal memory tests. Figure 3 shows R.B.‘s copy and delayed reproduction of the Rey-Osterreith figure, done at 6 and 23 months after the onset of his amnesia. Figure 2 also presents scores for 2 control groups, who were given 1 administration of each test. Alcoholic subjects (2 females and 4 males) were current or former enrollees in an alcoholic treatment program. They averaged 52 years of age (range = 40-58) and had 12.6 years of education (range = 1 l-l 5). They report having had a drinking problem for an average of 9.5 years and having abstained from alcohol for an average of 2 years prior to testing. Healthy control subjects (5 females and 1 male) averaged 50.2 years of age (range = 44-55) and had 15 vears of education (range = 13-18). Thev had WAIS-R subtest scores of 22.5 for information(R.B.‘s WAIS-R-score = 18) and 52.8 for vocabulary (R.B.‘s WAIS-R score = 52). Figure 4 shows that R.B.‘s memory impairment was still present 4 years after his injury. Subjects were read a list of 10 words twice in succession, each time in a different order, and were then asked to write down as many of the words as they could remember. Five filler words (3 at the beginning of each list and 2 at the end) were included to prevent primacy and recency effects. Descriptions of each group have been published previously (shaded bar in Graf et al., 1984, Experiment 1 -liking condition; open bar in Graf et al., 1985, Experiment 1-visual presentation). For patient R.B. and the control group (CONT), the shaded bar shows average free-recall performance for 4 separate lists of 10 words. All the other data show free-recall performance for 2 lists of 10 words. R.B.‘s anterograde amnesia was apparent on tests ofrecognition memory, not only on tests of free recall. In 1 test (Squire et al., 1978), subjects were presented 36 words and, after a 20 min delay, were tested for 2-choice recognition ( 12 target words and 12 novel words), yes-no recognition (12 target words and 12 novel words), and cued recall using the initial 3 letters of 12 target words. R.B. was given this test on 8 different occasions with 8 different forms of the test. He averaged 9.5 correct out of 36 (corrected for guessing). Six control subjects (mean age = 49.7 years, range = 43-60) averaged 19.3 correct (range = 14.923.5) on 3 different forms of the same test. Figure 5 presents scores for R.B. and a control group on 6 tests that assess retrograde amnesia, i.e., memory for events that occurred before the onset of amnesia. Data for other amnesic patients on these tests, and descriptions of all the tests, have been published previously (Cohen and Squire, 198 1; Zola-Morgan et al., 1983). R.B. was first given a test that asked about 88 news events that had occurred in 1 of the 4 decades from 1940-1979. The test was first given in a form (recall) that asked for a short answer to each question, e.g.,“What was the S.L.A.?” The same test was then given as a 4-choice, multiple-choice test (recognition). Finally, R.B. was asked to recall as much as he could remember about the same events, and his answers were scored according to how many facts he could recall (detailed recall). For these 3 tests, R.B. was compared to 8 male medical inpatients matched to him with respect to age (mean = 54.6 years, education = 12.1 years; information subtest of WAIS: mean = 18.6 [R.B.‘s WAIS score = 161; vocabulary subtest: mean = 52.1 [R.B.‘s WAIS score = 511). R.B. performed better than his control group on all 3 tests, and only 1 of his scores was noticeably lower: detailed recall of facts about the 1970s. For the 197Os, R.B. scored lower than 6 of 8 control subjects. R.B. was next given the Boston Famous Faces test (Albert et al.. 5(l- WORD RECALL 4( I- 5al 0: F z & a 313- 2( 3- 3- 3- .::...:.. >>>;,. :.>>> : .,i;:;;::‘i:; ...:.: .:.: .:.: ..I .:.:.:.:.>> >> ::.+\y d:::::::,:::,:: R.B. KORS (7) (8) ALC (14) (8) CONT (6) (8) Figure 4. Recall of lo-word lists in 2 separate studies by patient R.B., patients with Korsakoff syndrome (KORS), an alcoholic control group (ALC), and a control group of medical inpatients (CONT). The numbers under the bars show the number of subjects in each group. R.B. was tested 43 months (shaded bars; data from Graf et al., 1984) and 51 months (open bars; data from Graf et al., 1985) after the onset of his amnesia. 2954 Zola-Morgan et al. Vol. 6, No. 10, Oct. 1986 PUBLIC EVENTS RECALL 60 PUBLIC EVENTS RECOGNITION r 30 1 t 1970s 1960s 1950s 1940s PUBLIC DETAILED EVENTS RECALL 1970s 90l- I I 1960s 1950s FAMOUS FACES I 1940s 60r 60 - 40 ,t , 1970s T.V. PROGRAMS RECOGNITION I 1960s 1 1950s I 1940s I 1930s AUTOBIOGRAPHICAL RECALL 7Or Figure 5. Performance by R.B. and matched control groups on 6 tests of retrograde amnesia. The first 5 tests were given in 1979,7-10 months after the onset of his amnesia. Autobiographical recall was tested 2 years after the onset of amnesia. R.B. had little, if any, retrograde amnesia, except perhaps for a few years in the late 1970s. 77-75 74-72 71.69 Year d 963- 1979) which asked for identification of 130 photographs of people who had come into prominence at various times between 1930 and 1979. The test had been updated to include faces from the latter part of the 1970s (Cohen and Squire, 198 1). He was also given a 4-choice multiplechoice test, which asks subjects to recognize 74 television programs that were broadcast for a single season between 1963 and 1977 (approximately 5 for each year; Squire and Fox, 1980). For these 2 tests, R.B. was com’pared to a group of 29 nonalcoholic medical inpatients or hospital volunteers (23 males, 6 females; mean age = 55.3 years, mean education = 12.9 years). Sixteen were given the Famous Faces test, and 13 were given the television test (Cohen and Squire, 1981). R.B. performed normally on the Famous Faces test. On the television test, he performed close to the normal level, but achieved a noticeably lower score than the control subjects for the period 1975-1977. Finally, R.B. and 6 male control subjects (mean age = 52.8 years, mean education = 11.7 years; information subtest of WAIS: mean = 16.7 [R.B.‘s WAIS score = 161; vocabulary subtest: mean = 46.0 [R.B.‘s WAIS score = 5 11) were given a test of autobiographical recall (Crovitz and Schiffman, 1974). Ten concrete nouns were read one at a time (e.g., flag, bird, window) together with the instruction to recall a specific event from any time in the past that involved the target word. Transcripts were scored from 0 (a generic response) to 3 (for an episodic memory that occurred at a particular time and place). R.B. was as good as the control subjects at producing past episodic memories. In addition, the time periods from which the memories were drawn were similar for R.B. and his control subjects. Together, these retrograde amnesia tests indicate that R.B. had little 66-66 65-63 1977) if any loss of information about premorbid events. The possibility remains that he could have suffered some retrograde amnesia for a period of a few years prior to his coronary surgery in 1978. Although some of his test performance is consistent with this possibility (detailed recall of public events; television test), the tests are too coarsely grained and performance is too variable to reliably detect such a deficit in a single subject. In addition to formal memory tests, anecdotal evidence derived from R.B. himself and from people who knew him also attest to his anterograde memory impairment. As described earlier, R.B. often depended on his wife to recount details of recent events or conversations. During his contacts with our research group, he frequently told the same stories and asked the same questions within a short period. The physician who treated R.B. on an outpatient basis recognized the memory impairment in the context of the physical exam. Thus he noted that, even 4 years after R.B.‘s injury, he was often unable to recount how he was feeling the day before his visit. R.B. showed no signs of significant cognitive impairment other than loss of memory. His WAIS IQ of 111 (verbal, 108; performance, 114), was above the normal level, and with 1 exception his subtest scaled scores were average or above average (Information: 10; Comprehension: 13; Arithmetic: 11; Similarities: 10; Digit Span: 11; Vocabulary: 11; Digit Symbol: 8; Picture Completion: 12; Block Design: 10; Picture Arrangement: 12; Object Assembly: 10). The digit-symbol substitution subtest presumably depends on memory to some extent and has been reported previously to be difficult for patients with memory impairment (Kapur and Butters, 1977). The Journal of Neuroscience Memory Impairment Following Bilateral HippocampalCA1 Lesion 2956 Zola-Morgan et al. Vol. 6, No. 10, Oct. 1986 Figure 7. Photograph of a 1%-cm-thick unstained coronal section of R.B.‘s brain at a level through the anterior hippocampal formation. The hippocampus (I? 011 left, arrow on right) has a relatively normal appearance except for a region of thinning (between the smaller arrows) in the CA1 region. The mamillary nuclei (M) and the two fomices cf) are also indicated. Six months after his episode of cardiac dysfunction, R.B. was given the Boston Naming Test, Aphasia Screening Test, the Parietal Lobe Battery, and the Apraxia Test (Goodglass and Kaplan, 1972). On the Boston Naming Test, R.B. obtained a score of 88% correct (mean of 8 control subjects = 92%). On the Aphasia Screening Test, R.B. demonstrated 1 spelling error (“traingle” for “triangle”), but otherwise performed perfectly. On the Parietal Lobe Battery, R.B. obtained a score of 208 out of 249 possible points (the Diagram-Copy subtest was not given). Eight control subjects obtained a mean of 242 points. R.B. lost 26 points on this test because of his poor performance on the tactilevisual finger subtest. Here, the examiner lightly touches a finger on the subject’s hand while it is hidden from view; the subject then points to the touched finger on a drawing of a hand. R.B. had no experience of being touched on any fingers of his left hand (score: 0 out of 16). and he correctly matched only touches of the thumb and little finger of his right hand (score: 6 out of 16). On the Anraxia Test. R.B. was able to carry out correctly all movements to oral commands and obtained a perfect score. Reports from R.B.‘s family were in agreement with these neuropsychological findings. The memory impairment was readily apparent to them, but there was never any mention of personality change or of any other cognitive deficit. Results: Neuropathological Findings Acquisition and preparation of tissue The brain was removed at autopsy approximately 4 hr after death and placed in 30 liters of 10%formalin in 0.1 M phosphate buffer (pH 7.2). Approximately 2 weeks later, the brain was suspendedupside down in the samesolution with strings attached to the basalvasculature. The brain wasfixed for 7 months prior to any histological processing,and during this period the formalin solution was changedevery 2 weeks. Before the brain was sectioned, photographs were taken of the dorsal, ventral, and lateral surfaces.Thereafter, the brain was cut transversely into 6 blocks approximately 1% cm thick (much of the visual cortex was not further sectioned),and the surfacesof these blocks were photographed. The blocks were then subdivided to a sizeappropriate for sectioningon a freezing sledgemicrotome. Initially, the tissue blocks were cut to approximately 36 cm*, and the sectionsthrough the diencephalon and medial temporal lobe were of this size. However, with the introduction of a modified freezing stage, much larger blocks (up to 120 cm*) could be sectioned, and much of the cortical tissuewas processedin this way. The blocks were placed in a cryoprotectant solution of 20%glycerol in 0.1 M phosphatebuffer at least 1 week prior to sectioning. Continuous 50 Mmsectionswere cut through each block, and every fifth section (through the diencephalon, the medial temporal lobe, and much of the neocortex) or every tenth section (through portions of the neocortex and cerebellum)was mounted on gelatin-coated slidesand stained with 0.25% thionin. In addition, selected sectionsthrough the medial temporal lobe The Journal of Neuroscience Memory Impairment Following Bilateral Hippocampal CA1 Lesion 2957 Figure 8. Photomicrographs of thionin-stained coronal sections through the hippocampal formation of R.B.‘s brain. Sections are from the left (L) and right (R) sides of the rostra1 extreme (top), the midportion (middle), and the caudal extreme (bottom) of the hippocampal formation. The only pathology evident in these sections is the complete loss of pyramidal cells confined to the CA1 field of the hippocampus (asterisks). CA3, Field CA3 of the hippocampus; DC, dentate gyrus; EC, entorhinal cortex; PrS, presubiculum; S, subiculum. Holes in tissue in top panels (small stars) are artifactual. Calibration bar, 2 mm. 2958 Zola-Morgan et al. Vol. 6, No. 10, Oct. 1986 Figure 9. Higher-magnification photomicrograph of a coronal section through the left hippocampal formation. The region of cell loss in field CA1 (asterisk) is delimited by 2 bold arrows. The sharp boundary with field CA3 is apparent in this section. All other fields of the hippocampal formation have normal cell densities and appropriate laminar organization. The rectangle outlines the region shown at higher magnification in Figure 10. Additional abbreviations: PUS, parasubiculum; PRC, perirhinal cortex. Calibration bar, 2 mm. The Journal of Neuroscience Memory Impairment Following Bilateral and diencephalon were stained by a recent modification of the Heidenhain procedure for the demonstration of myelinated fibers(Hutchins and Weber, 1983).The unmounted sectionsMvere stored in 10% formalin solution. Sections were analyzed with a Wild stereomicroscopeor a Leitz Dialux 20 microscopeequipped for 35 mm photomicrography. Low-magnification photographswere madewith a pointsourcephotographic enlargerthrough which sectionswere projected onto 4 x 5 negatives. Gross appearance of the brain The grossappearanceof the brain was essentiallynormal (Fig. 6). There was no indication of general atrophy of the cortex, and only one infarct was visible in the depths of the posterior bank of the central sulcusof the right hemisphere(asterisk,Fig. 6A). This infarct had a yellow/brown appearance,and extended laterally from a point 2 cm lateral to the midline to a point 3.5 cm lateral to the midline; the rostrocaudal extent of the lesion was not visible from the surface. No other pathology was apparent on the surfaceof the brain. In particular, the mamillary nuclei were of normal size and shape(Figs. 60, 7). Microscopic appearanceof the brain Medial temporal lobe. The primary lesionin this brain occurred in field CA1 of the hippocampus (Figs. 8-10). There was a complete and bilateral lossof cellsthrough the full mediolateral and rostrocaudal extent of this field (Figs. 8, 9). The radial dimension of field CA1 wasmarkedly shrunken, and there was an apparent increasein the density of stained glial cells above and below the empty pyramidal cell layer (Fig. 10, A, B). This increasedglial density did not extend into the pyramidal cell layer and waspresumablya consequenceofthe abnormally close apposition of cells related to the fibers of the alveus and perforant path, rather than an indication of a reactive process.The lesion had an old appearance,asjudged by the absenceof reactive gliosis,and presumablyoriginated at the time of the initial ischemic event. Throughout most of the hippocampus,the lesionwasconfined to the CA1 field and did not invade field CA3 or the subiculum. Figure 9 demonstratesthe strikingly abrupt transition from normal cells in field CA3 to a complete lossof cells in field CAl. In the caudal one-third of the hippocampus,however, the area of cell lossextended slightly into the subiculum and CA3. This was observed only in 2 sections (representing a rostrocaudal distanceof 0.5 mm), and caudal to thesesectionsthe lesion was again restricted to field CAl. Fiber preparations through the hippocampusdemonstratedthat neither the alveus (which runs subjacentto the pyramidal cell layer of CAl) nor the perforant path (which originates in the entorhinal cortex and runs in the superficial layers of CAl) were affected by the pathology (Fig. 10,A, B). Fibers of the perforant path were seento travel through CA 1to reach their normal termination in the dentate gyrus and field CA3 of the hippocampus. The fimbria had a normal appearance,though there was a small patch of gliosislocated ventromedially in the region that was presumably occupied by the degeneratedfibers originating in CA 1. No cell losswas detectable in any field of the hippocampal formation other than CA1 . The granule cell layer of the dentate gyrus appearednormal bilaterally along its rostrocaudal extent, and the molecular layer appeared to be of normal thickness. While damageto cells in the hilar region of the dentate gyrus is a common finding in experimental ischemia(Plum, 1983) a normal density of cells was seenthere in this case.Other than the focal cell lossin CA3 describedabove, this field had a normal appearanceat all other levels of the hippocampus. The subiculum, presubiculum, and parasubiculum all had a normal appearance. The entorhinal cortex demonstrated its distinctive Hippocampal CA1 Lesion 2959 laminated appearance(Figs. 9, 1l), and there were no obvious patches of cell loss like those observed in the neocortex (see below). Sectionswere alsoexamined through the rostrocaudal extent of the amygdaloid complex. All subdivisions of the amygdala were presentand all appearedto be populated by an appropriate density of neurons(Fig. 11). There was a small area containing patchy cell loss bilaterally at the junction of the amygdaloid complex, with the substantia innominata in what appearedto be the anterior amygdaloid area. The cell loss,which occurred in foci of about 1 x 0.5 x 0.5 mm, involved only a small portion of this latter area. Diencephalon. Damage to the mamillary nuclei (Barbizet, 1973;Brierley, 1977)and the midline and mediodorsalthalamic nuclei (Talland, 1965; Victor et al., 1971) have long been associatedwith memory impairment. We therefore paid particular attention to the analysisof thesestructures. In accordancewith the apparently normal grossappearanceof the medial mamillary nuclei, there wasno indication of cell lossin any component of the mamillary complex (Fig. 12, A, B). Nor was there any increasein glial staining within the fomix or the principal mamillothalamic tract. Both of these fiber bundles also appeared normal in the myelin preparations. The mediodorsal nucleus had a generally normal appearance,except for a small perivascular patch of gliosis(< 1 mm in width) just within the internal medullary lamina of the left thalamus. There wasno evidence, however, of cell lossin this nucleus(Fig. 12, C, D). There were also no pathological changesin any other subdivision of the thalamus. Cortex. Tissuesectionswere examined from all regionsof the cortical mantle except for primary visual cortex. The only significant focus of cortical damage was an infarct in the right postcentral gyrus (primary somatosensorycortex). At its caudal border, the infarct alsoinvolved a portion of area 5 but did not appear to extend into area 7 of the posterior parietal lobe. The infarct extended for approximately 1.5 cm mediolaterally and 1 cm rostrocaudally and was located in the region of the hand and forelimb somaticrepresentation.Damagein this region was heaviest in the white matter and involved the cortex itself only at the center of the infarct where there was a cavity that was partially filled with fibrous processes.Adjacent regions of somatosensorycortex had an entirely normal appearance. There were no other infarcts of this size or type in any other cortical region. There were, however, small foci of cell lossdistributed throughout the cortex. Thesefoci (Fig. 10, C, D) tended to be small; they generally involved a volume of 0.125 mm3or less,though the largestoccupied a volume of 1 mm3.They were confined mainly to only 2 or 3 cellular layers of the cortex and were seen as often in a supragranular as in an infragranular position. They rarely occupied all cell layers (Fig. lOD), and in thesecasesformed a column of cell loss.There wasno indication that the white matter subjacent to the foci was involved in the pathology. The foci were characterized by an irregular region of complete cell losswith a slightly increaseddensity of glial cells and occasionalpatchesof apparent neuronal debris. There was often an increasedvascular&y in the focus, and at its periphery there appearedto be an increasein the density of stained neurons. We have not found descriptionsof this type of damagein the neuropathologicalliterature. Thesefoci of cell lossareclearly smaller than the lacunaedescribedby Fisher (1965) which are generally found in subcortical structures. To estimate the number and distribution of these foci, we counted them from sectionsthrough the temporal lobe at levels adjacent to the amygdaloid complex and hippocampal formation. In the sectionsstained in this region, we noted 64 such foci. While this would appear to representa sizable amount of damage, it can be seenin Figure 1OC(in which the largest of Zola-Morgan et al. Vol. 6, No. 10, Oct. 1986 Figure IO. A and B, Photomicrographs of adjacent coronal sections from the CA1 field of the hippocampus (see rectangle Fig. 9) stained for cell bodies (A) or for myelinated fibers (B). No pyramidal cells were observed in the pyramidal cell layer (PL). The granular staining above and below the pyramidal cell layer is associated with the fibers of the perforant path (pp) and alveus (al), respectively; neither pathway was affected by the damage. Fibers leave the perforant path (arrow in B) to enter the molecular layer of the dentate gyrus (MI,). C, Low-power photomicrograph of the right temporal lobe. An acellular focus located along the superior temporal sulcus is indicated with a bold arrow. D, Higher-magnification photomicrograph of a second acellular focus (open arrow) located in the perirhinal cortex. These foci were characterized by lack of neuronal staining, The Journal of Neuroscience Memory Impairment Following Bilateral Hippocampal CA1 Lesion 2961 Figure 1 I. Photomicrograph of a coronal section through the left amygdaloid complex. All subdivisions of the amygdala have a normal appearance. AB, Accessory basal nucleus; B, basal nucleus; C, central nucleus; EC, entorhinal cortex; L, lateral nucleus; PAC, periamygdaloid cortex. Calibration bar, 2 mm. slightly increased glial staining, a prominent vascularization, and increased density of stained neurons around their periphery. While many foci were confined to 2 or 3 layers of the cortex, this focus extended throughout all 6 layers. Calibration bar in B (applies to A and D as well), 500 pm. Calibration bar in C, 2 mm. 2962 Zola-Morgan et al. Vol. 6, No. 10, Oct. 1986 Figure 12. Photomicrographs of coronal sections through rostral (A) and caudal (B) levels of the mamillary complex. Both the medial (MM) and lateral (LM) nuclei had a normal density of neurons, with little indication of gliosis. Neither the fomix (f) nor the principal mamillothalamic tract (mt) demonstrated abnormal glial staining. C, Low-power photomicrograph of the medial, magnocellular division of the mediodorsal nucleus of the thalamus. The normally patchy appearance of the nucleus is apparent at this magnification. 0, Higher-magnification photomicrograph of the same region that shows normal density of neurons. TB, Tuberomamillary nucleus. Calibration bar in B, 2 mm (also applies to A). Calibration bar in C, 1 mm; in D, 500 pm. The Journal of Neuroscience Memory Impairment Following Bilateral these foci has been illustrated) that these foci of cell loss representa very small portion of the cortical surface.The foci were not bilaterally symmetrical and appearedto be randomly distributed throughout the cortex. Some regions, such as primary motor cortex, appearedto have a smaller number of such foci. It shouldbe noted that the neocortex generally appearednormal and had an appropriate laminar organization. There was no indication of laminar necrosis,asis seenin human casesof prolongedischemia(Garcia and Conger, 1981; Plum, 1983)and in experimental models of ischemic damage (Pulsinelli and Brierley, 1979; Siesjo, 1981; White et al., 1984). Moreover, the density of neuronsin the various cortical regionsappearednormal. It is possible,however, that a small amount of cell loss (< 10%)would not be detected by our qualitative analysis. Other regions.There wasa prominent lesionin the left globus pallidus. The lesionwasfocusedin the lateral medullary lamina but involved both the medial and lateral segmentsof the globus pallidus. The lesion consistedof a dorsoventrally oriented strip of densegliosiswith bandsof gliosisassociatedwith fiber bundlesextending laterally into the putamen. The right globuspallidus appearednormal. There wasalsoa small (~250 pm-wide) focus of gliosisin the left internal capsulelocated dorsomedially between the caudate nucleusand nucleusventralis anterior of the thalamus. The only other region in which any cell losswas detectedwasthe cerebellarcortex. There appearedto be a patchy lossof Purkinje cells, with an apparent increaseof glial staining within the Purkinje cell layer. Finally, although damagein the basalforebrain haspreviously beenlinked to memory impairment (Damasioet al., 1985b),no cell losswas detected in this region in the presentcase.In particular, the large cells of the basalnucleusof Meynert appeared to be presentin normal density, and they were not involved in the globus pallidus lesion. In summary, the major finding wasa bilateral lesioninvolving the entire CA1 field of the hippocampus.Minor pathology was observed in a few areas: the right postcentral gyrus, the left globus pallidus, the left internal medullary lamina of the thalamus, small foci (lessthan 1 mm3) in the dorsomedialportion of the anterior amygdaloid area, a small (lessthan 250 pm in width) focus of gliosis in the left internal capsule,patchy loss of cerebellarPurkinje cells, and scatteredsmall foci of cell loss (0.125 mm3 or less)in the neocortex. Discussion This caseshowsthat circumscribed damageto field CA1 of the hippocampal formation resultsin a clinically meaningful memory impairment. To our knowledge,this is the first reported case of amnesiafollowing a lesion limited to the hippocampus in which extensive neuropsychological and neuropathological analyseshave been carried out. R.B. exhibited normal performanceon a variety of neuropsychologicaltestssensitiveto cognitive impairments other than memory. These findings were consistentwith the neuropathologicalfindings that R.B.‘s brain exhibited no grosscortical abnormalities. The one exception wasan infarct involving the forearm representationof the right somatosensorycortex. This infarct wasconsistentwith the finding that R.B. had hypoesthesiaof the dorsal surfaceof his left arm and hand; and of his fingers.This lesion also explains, in part, his impaired performanceon the tactual-visual finger subtest of the Parietal Lobe Battery. The possibility should be consideredthat R.B.‘s memory impairment resulted from the combination of the CA1 lesion and other damage.For example, it might be suggestedthat cortical damage,i.e., the small acellularfoci, contributed to R.B.‘s memory impairment. This seemsunlikely. First, R.B.‘s impairment was limited to the domain of memory; his performance was normal on several tests of other cognitive abilities. Thesefindings suggestthat the cortical damagewas of little clinical rele- Hippocampal CA1 Lesion 2963 Vance.Second,work with nonhuman primates hasdemonstrated that bilateral damageto the hippocampus is sufficient to produce amnesia (Mahut et al., 1981, 1982; Mishkin, 1978; Zola-Morgan and Squire, 1986a). Third, damage observed in other regions of R.B.‘s brain, such as the globus pallidus or cerebellum, has never been associatedwith the syndrome of amnesia. The severity of the memory impairment. The severity of amnesiacan vary widely among patients. For example, H.M., the noted patient who becameseverely amnesicfollowing bilateral medial temporal lobe resection (Scoville and Milner, 1957) is so impaired that he is capable of almost no new learning and forgets the events of daily life as quickly as they occur. His WAIS-WMS difference score of 43.7 (average of 6 tests from 1955 to 1983; Corkin, 1984) reflects the severe nature of his memory impairment. R.B.‘s memory impairment was not as severe as that of H.M.; his WAIS-WMS difference score, for example, was only 20 points. Yet his deficit was readily detectable by formal memory tests and apparent to casualobservation. His performance on tests of anterograde memory (e.g., paired associatelearning, story recall, and diagram recall) was similar to that of other amnesic patients studied in our laboratory (Fig. 2). The neuropsychologicalfindings from this caseare consistent with the resultsfrom recent studiesin monkeys (Mahut et al., 1981; Mishkin, 1978; Squire and Zola-Morgan, 1983; ZolaMorgan and Squire, 1986a) and rats (Jarrard, 1978; Olton et al., 1979) in which lesions limited to the hippocampal formation impaired learning and memory. Particularly relevant to the presentcaseis the recent finding that experimentally induced ischemia in rats damagesthe CA1 region of the hippocampus and resultsin a long-lastingdeficit in new learning ability (Davis et al., 1986;Volpe et al., 1984).The behavioral deficits following hippocampal lesionsin animals, though statistically significant, are not alwaysquantitatively impressive.The information from the presentcaseindicatesthat the deficit following hippocampal lesionsis nonethelessclinically important; it reflectsa degreeof memory impairment that in humanswould be expectedto have a significant impact on daily learning ability and quality of life. Behavioral tests used in studies of experimental animals will need to be sensitive to the full range of impairment that is of clinical relevance. The persistenceof the memory impairment. The presentcase alsoshowsthat the memory impairment following limited damage to the hippocampal formation is stable and long lasting. Figure 4 showsthat a marked deficit in verbal recall was still presentmore than 4 years after the injury. At the sametime, it is possiblethat somedegreeof recovery did occur. R.B.‘s wife noted that, while R.B. never fully regained his memory after the injury, his memory impairment seemedmore severeduring the first year than in subsequentyears. In addition, Figure 3 showsthat R.B.‘s memory performance was slightly better, albeit still very impaired, after 23 months (middle), compared to after only 6 months (left). Recent studies of monkeys that have been operated on support the finding that memory impairment following hippocampal damagecan be long lasting. Monkeys with conjoint removal of hippocampus and amygdala were just as impaired 1.5 to 2 years after surgery as during the first 2 months after surgery (Zola-Morgan and Squire, 1985). A similar result hasbeen observed after lesionslimited to the hippocampal formation (Mahut et al., 1982; Zola-Morgan and Squire, 1986b). The hippocampusand the neuropsychologyof memory. The neuropsychological findings from this caseare entirely compatible with recent descriptionsof the humanamnesicsyndrome (seeBaddeley, 1982;Hirst, 1982; Squire, 1987a;Squire and Cohen, 1984; Weiskrantz, 1982). Amnesia has long been recognized as the combination of intact generalintellectual capacity, 2964 Zola-Morgan et al. Vol. 6, No. 70, Oct. 1986 EC Figure 13. Schematicdrawingof the primatehippocampal formation.The numbers andthe solid lines connectingthemshowthe unidirectional flow of information(from entorhinalcortex I, to dentategyms2, to CA3 3, to CA1 4, to the subiculum5). In caseR.B., a lesionof the CA1 field (represented by cell 4 andthe dashed lines) disruptedthe sequence of informationflow in the hippocampus. EC, Entorhinalcortex; PUS,parasubiculum;PrS, presubiculum; S, subiculum;DG, dentategyrus;CA1 and CA3, fieldsof the hippocampus. intact digit span (immediate memory), intact personality, language,and social skills, together with impaired new learning capacity for verbal and nonverbal material and variable retrogradeamnesia.More recently, it has alsobeen appreciatedthat amnesiasparessome memory functions, including motor and cognitive skill learning, and priming. The amnesic syndrome has been useful in revealing information about the organization of memory and about the normal function of the damagedneural system. The finding that R.B. could recall events that occurred prior to his ischemic episode is consistent with the view that the hippocampus is not a repository for storageof remote memories(Damasioet al., 1985a; Milner, 1972). Nonetheless,R.B. may have had somedifficulty recalling events from 1 to 2 yearsprior to hisepisode;temporally limited retrograde amnesiascovering a few years have been observed in other amnesic patients. It has been suggested,in fact, that the medial temporal region, including the hippocampus, hasan essentialrole in memory functions for a considerable time after learning (Squire, 1987b; Squire et al., 1984). R.B., like other amnesicpatients, exhibited intact word priming; that is, like normal subjects, his performance on wordcompletion tests (MOT-) could be biased for a time by prior exposure to test words (MOTEL) (Graf et al., 1984, 1985; Experiment 1, Case 1). Priming, like skill learning, occurs in amnesic patients without consciousawareness,and without recognition of the previously presented stimulus material. These and other findingshave suggesteda distinction betweenmemory for facts and episodes(declarative memory), which is impaired in amnesia,and memory for skills and procedures(procedural memory), which is intact (Cohen, 1984;Cohenand Squire, 1980; Squire and Cohen, 1984). The findings from the present case support the idea that declarative memory is impaired in amnesia,and that the hippocampus is an essentialcomponent of the damagedneural system.Procedural memory is independent of this system. Field CAl: neural circuitry. The hippocampal formation consistsof several cytoarchitectonically distinct subdivisions, including the dentate gyrus, the hippocampus proper (which is subdivided into fieldsCAl, CA2, and CA3 after Lorente de No, 1934) the subicularcomplex (the subiculum,presubiculum,and parasubiculum), and the entorhinal cortex. One distinguishing characteristic of these fields is that they are linked by a largely unidirectional flow of information (Fig. 13). The entorhinal cortex provides the major input, via the fibers of the so-called perforant path, to the dentate gyrus. While the dentate gyrus doesnot reciprocate the projection of the entorhinal cortex, it doesprovide a major excitatory input to the pyramidal cellsof CA3 of the hippocampusvia the axons (the mossyfibers) of the dentate granulecells. Field CA3, in turn, gives rise to a powerful associationalprojection (the Schaffer collaterals) to field CAl, which then projects to the subicular complex. The subicular complex completes the circuit by projecting to the entorhinal cortex, among other regions. Within this circuit, a lesion confined to field CA1 of the hippocampusessentiallybreaks the chain of information-processing that beginsat the dentate gyrus and ends in the subicular complex and entorhinal cortex. This lesion, while spatially limited, would be expected to have a profound influence on the functioning of the hippocampal formation becausethe subicular complex and entorhinal cortex are the main sourcesof output from the hippocampus to subcortical structures such as the mamillary nuclei and the anterior thalamus, and to limbic and cortical structures such as the amygdaloid complex, the perirhinal and parahippocampal gyri, and the orbitofrontal cortex (Amaral, 1987; Van Hoesen, 1982). [In each hemisphere,the CA1 field of the human hippocampuscontains approximately The Journal of Neuroscience Memory Impairment Following Bilateral 4.63 million pyramidal cells. Fields CA2 and CA3 contain an additional 2.35 million pyramidal cells(Brown and Cassell,1980; M. D. Cassell,personalcommunication, 1985).The CA1 region of the human hippocampushasundergonethe greatestenlargement in area relative to the other fields of the hippocampal formation (Stephan, 1983; Stephan and Manolescu, 1980)]. Field CAI: mechanismsof damage.The question arisesasto why the brain damagein this casewas so circumscribed and involved only field CA1 of the hippocampus. Sommer (1880) noted the selective vulnerability of this region more than a century ago in patients who had sufferedfrom epilepsy. In one case,in which sectionsthrough the hippocampuswere studied microscopically, Sommer noted a discreteregion of marked cell loss.Since there was no establishednomenclature for the hippocampusor its subfieldsat that time, Sommer describedthe area of destruction as a sector of an ellipse superimposedover the coronal section of the hippocampus. This area of the hippocampus closely approximates field CA1 of Lorente de No (1934), and it has come to bear the name “Sommer’s sector.” The pattern of brain damage(including CA1 cell loss) seen in the presentcaseis consistentwith the idea that the responsible event wasan ischemicepisode.Depletion of pyramidal cellsin the CA1 field of hippocampus and in the Purkinje cell layer of cerebellum have been identified as neuropathological markers of ischemiain both humans(Brierley and Cooper, 1962; Plum, 1983; Siesjo, 1981) and animals (Levine, 1960; Pulsinelli and Brierley, 1979). Ischemiarefers to a reduction in normal blood flow to a level that is insufficient to meet metabolic demands (Garcia and Conger, 1981). During an ischemicepisode,oxygen delivery may fall below critical levels, producing an anoxic state. In some cases,oxygen pressurecan be maintained artificially during hypotensive states,and tissue requirements for oxygen may be reducedin responseto hypotension (Garcia and Conger, 1981). During R.B.‘s acute episode,he lost approximately 5000 cc of blood and was presumably ischemic during this episode. He was provided with oxygen at the time of the ischemia, and thus it is difficult to know whether he was ever anoxic. The mechanism of selective neuronal death following ischemia is not well understood(for reviews, seePlum, 1983; Siesjo, 1981; White et al., 1984). One possibility is that the CA1 region of the hippocampus is selectively vulnerable becauseof the architecture or mechanics of the blood supply to this region. However, such an explanation does not seemaltogether satisfactory. First, the blood supply to field CA1 also suppliesthe subicular complex. It seemsunlikely, therefore, that compression of this vasculature could lead to damageof the CA1 field while at the sametime sparingthe subiculum. In addition, the sharpboundary of the pathology at the CA3-CA1 border cannot easily be explained by the known vascular pattern of the hippocampus(Coyle, 1978; Lindenberg, 1963; Nilges, 1944). Second, in studiesof barbiturate-induced systemichypotension in experimental animals,the hippocampusdid not exhibit a greater reduction in blood flow than other brain areas(Gamache and Dold, 1975; Gamache et al., 1976). Third, damageto the CA1 field of hippocampusis a common finding in the brains of epileptic patients (Dam, 1980; Margerison and Corsellis, 1966; Sommer, 1880). Yet blood flow in the hippocampus during experimentally induced seizures is increased rather than decreased(Sloviter, 1983). It has recently been proposed that the selective damageof CA1 cells during ischemia is due to the neurotoxic properties of glutamate or a related excitatory amino acid (Fagg, 1985; Meldrum, 1985; Olney, 1978; Siesjo, 1981). A number of independent observations support this idea. First, there is a high density of glutamate receptors in the CA1 field of the hippocampus; these are primarily of the N-methyl-D-aspartate (NMDA) type (Greenamyreet al., 1984;Monaghan et al., 1983). Second, the major excitatory input to the CA1 pyramidal cells Hippocampal CA1 Lesion 2965 (the Schaffercollateralsfrom field CA3) appearsto useglutamate asa transmitter (Storm-Mathisen, 1981, 1984;Wolf et al., 1984). Third, injection of an excitatory amino acid antagonist,2-amino 7-phosphonoheptanicacid (AHP), into the hippocampusprior to ischemia can protect against pyramidal cell loss(Meldrum, 1985). Fourth, decreasingthe synaptic activation of the CA1 field by eliminating the main excitatory input to the hippocampus (the perforant path) has also protected the CA1 cells from ischemicdamage(Wieloch et al., 1985). In addition, the administration of antiepileptic drugs suchasbarbiturates has reduced the extent of ischemic cell loss (Siesjo, 1981). And fifth, sustained electrical stimulation of the perforant path produces a pattern of hippocampal damagethat includesextensive cell loss in the CA1 subfield (Olney, 1983; Sloviter, 1983). Overview. Over the last 100 years, amnesiahas been associated in clinicopathological studies with damageto several regions of the brain. Among these are the amygdala and hippocampalformation in the medialtemporal lobe, andthe mamillary bodies and mediodorsal nucleus of the diencephalon. It has remained unclear, however, whether damageconfined to any one of these structures would produce a clinically significant memory impairment. The present case shows that a circumscribed bilateral lesion limited to the CA1 field of the hippocampal formation is sufficient to produce amnesia. References Albert, M. S., N. Butters, and J. Levin (1979) Temporal gradients in retrograde amnesia of patients with alcoholic KorsakolI’s disease. Arch: Neurol. 36: 2 1 l-2 16. Amaral. D. G. (1987) Memorv: The anatomical organization of candidatk brain regions. In Ha&book of Physiology: ?he Nervous System, J. M. Brookhart and V. B. Mountcastle, eds., American Physiological Society, Bethesda, MD (in press). Baddeley, A. (1982) Amnesia: A minimal model and an interpretation. In Human Memory and Amnesia, L. Cermak, ed., pp. 305-336, Erlbaum, Hillsdale, NJ. Barbizet, J. (1973) Human Memory and its Pathology, pp. 1-198, Freeman, San Francisco. von Bechterew, W. (1900) Demonstration eines Gehims mit Zerstorung der vorderen und inneren Theile der Himrinde beider Schlafenlappen. Neurologisch. Zentralbl. 19: 990-99 1. Benson, D. F., C. D. Marsden, and J. C. Meadows (1974) The amnesic syndrome of posterior cerebral artery occlusion. Acta Neurol. Stand. 50: 133-145. Brierley, J. B. (1977) Neuropathology of amnesic states. In Amnesia, 2nd Ed., C. W. M. Witty and 0. L. Zangwill, eds., pp. 199-223, Butterworths, London. Brierley, J. B., and J. E. Cooper (1962) Cerebral complications of hypotensive anaesthesia in healthy adult. J. Neurol. Neurosurg. Psychiatry 25: 24-30. Brown, M. W., and M. D. Cassell (1980) Estimates of the number of neurones in the human hippocampus. J. Physiol. (Lond.) 301: 58P59P. Cohen, N. J. (1984) Preserved learning capacity in amnesia: Evidence for multiple memory systems. In The Neuropsychology ofMemory, L. Squire and N. Butters, eds., pp. 83-103, Guilford, New York. Cohen, N. J., and L. R. Squire (1980) Preserved learning and retention of pattern analyzing skill in amnesia: Dissociation of knowing how and knowina that. Science 210: 207-209. Cohen, N. J., and L. R. Squire (198 1) Retrograde amnesia and remote memory impairment. Neuropsychologia 19: 337-356. Corkin, S. (1984) Lasting consequences of bilateral medial temporal lobectomy: Clinical course and experimental findings in H.M. Semin. Neurol. 4: 249-259. Coyle, P. (1978) Spatial features of the rat hippocampal vascular system. Exp. Neurol. 58: 549-56 1. Crovitz, H. F., and H. Schiffman (1974) Frequency of episodic memories as a function of their age. Bull. Psychonom. Sot. 4: 5 17-5 18. Cummings, J. L., U. Tomiyasu, S. Read, and D. F. Benson (1984) Amnesia with hippocampal lesions after cardiopulmonary arrest. Neurology 34: 679-68 1. 2966 Zola-Morgan et al. Dam, A. M. (1980) Epilepsy and neuron loss in the hippocampus. Epilepsia 21: 6 17-629. Damasio, A. R., P. J. Eslinger, H. Damasio, G. W. Van Hoesen, and S. Cornell (1985a) Multimodal amnesic syndrome following bilateral temporal and basal forebrain damage. Arch. Neurol. 42: 252259. Damasio, A. R., N. R. Graff-Radford, P. J. Eslinger, H. Damasio, and N. Kassell (1985b) Amnesia following basal forebrain lesions. Arch. Neurol. 42:‘263-27 1. Davis, H. P., J. Tribuna, W. A. Pulsinelli, and B. T. Volpe (1986) Reference and working memory of rats following hippocampal damage induced by transient forebrain ischemia. Physiol. Behav. 37: 387392. DeJong, R. N., H. H. Itabashi, and J. R. Olson (1968) “Pure” memory loss with hippocampal lesions: A case report. Trans. Am. Neurol. Assoc. 93: 3 l-34. Duyckaerts, C., C. Derouesne, J. L. Signoret, F. Gray, R. Escourolle, and P. Castaigne (1985) Bilateral and limited amygdalohippocampal lesions causing a pure amnesic syndrome. Ann. Neurol. 18: 3 14-3 19. Fagg, G. E. (1985) L-glutamate, excitatory amino acid receptors and brain function. Trends Neurosci. May: 207-2 10. Fisher, C. M. (1965) Lacunes: Small, deep cerebral infarcts. Neurology 15: 774-784. Gamache, F. W., and G. M. Dold (1975) Alterations in cerebral blood flow produced by hypotension: A comparison of methods. J. Neurol. Neurosurg. Psychiatry 39: 765-770. Gamache, F. W., R. E. Myers, and E. Monell (1976) Changes in local cerebral blood flow following profound systemic hypotension. J. Neurosurg. 44: 2 15-225. Garcia, J. H., and K. A. Conger (1981) Ischemic brain injuries: Structural and biochemical effects. In Brain Failure and Resuscitation, A. Grenvik and P. Safar, eds., pp. 35-54, Churchill Livingstone, New York. Gilbert. J. G.. R. F. Levee, and F. L. Catalan0 (1968) A preliminary report on a’new memory scale. Percep. Mot. Skills 27: 277-278. Glees. P.. and H. B. Griffith (1952) Bilateral destruction of the hinpocampus (comu ammonis) in a case of dementia. Psychiatr. Neural. Med. Psychol. (Leipz.) 123: 193-204. Gol, A., and G. M. Faibish (1967) Effects of human hippocampal ablation. J. Neurosurg. 26: 390-398. Goodglass, H., and E. Kaplan (1972) The Assessment ofAphasia and Related Disorders, pp. l-80, Lea and Febiger, Philadelphia, PA. Graf, P., L. R. Squire, and G. Mandler (1984) The information that amnesic patients do not forget. J. Exp. Psychol. (Hum. Learn.) IO: 164-178. Graf, P., A. P. Shimamura, and L. R. Squire (1985) Priming across modalities and priming across category levels: Extending the domain of preserved function in amnesia. J. Exp. Psychol. (Hum. Learn.) 11: 386-396. Greenamyre, J. T., A. B. Young, and J. B. Penny (1984) Quantitative autoradiographic distribution of L-3H-glutamate-binding sites in rat central nervous system. J. Neurosci. 4: 2 133-2 144. Grunthal, E. (1947) Uber das klinische Bild nach umschriebenem beiderseitigem Ausfall der Ammonshomrinde. Monatsschr. Psychiat. Neurol. 113: 1-16. Hirst, W. (1982) The amnesic syndrome: Descriptions and explanations. Psychol. Bull. 91: 435-460. Horel, J. A. (1978) The neuroanatomy of amnesia. Brain 101: 403445. Hutchins, B., and J. T. Weber (1983) A rapid myelin stain for frozen sections: Modification of the Heidenhain procedure. J. Neurosci. Meth. 7: 289-294. Jarrard, L. E. (1978) Selective hippocampal lesions: Differential effects on performance by rats of a spatial task with preoperative versus postoperative training. J. Comp. Physiol. Psychol. 92: 1119-l 127. Jones, M. K. (1974) Imagery as a mnemonic aid after left temporal lobectomy: Contrast between material-specific and generalized memory disorders. Neuropsychologia 12: 21-30. Kapur, N., and N. Butters (1977) Visuoperceptive deficits in longterm alcoholics and alcoholics with Korsakoffs psychosis. J. Stud. Alcohol 38: 2025-2035. Kaushall, P. I., M. Zetin, and L. R. Squire (198 1) A psychosocial study of chronic, circumscribed amnesia. J. Nerv. Ment. Dis. 169: 383389. Levine, S. (1960) Anoxic-ischemic encephalopathy in rats. Am. J. Pathol. 36: 1-17. Vol. 6, No. 10, Oct. 1986 Lindenberg, R. (1963) Patterns of CNS vulnerability in acute hypoxaemia, including anaesthesia accidents. In Selective Vulnerability of the Brain in Hypoxaemia, J. P. Schade and W. H. McMenemey, eds., pp. 189-209, Blackwell, Oxford. Lorente de No, R. ( 1934) Studies on the structure of the cerebral cortex II. Continuation of the study of the ammonic system. J. Psychol. Neurol. 46: 113-l 17. Mahut, H., and M. Moss (1984) Consolidation of memory: The hippocampus revisited. In The Neuropsychology of Memory, L. R. Squire and N. Butters, eds., pp. 297-3 15, Guilford, New York. Mahut, H., M. Moss, and S. Zola-Morgan (1981) Retention deficits after combined amygdalo-hippocampal and selective hippocampal resections in the monkey. Neuropsychologia 19: 201-225. Mahut, H., S. Zola-Morgan, and M. Moss (1982) Hippocampal resections impair associative learning and recognition memory in the monkey. J. Neurosci. 2: 1214-1229. Margerison, J. H., and J. A. N. Corsellis (1966) Epilepsy and the temporal lobes. Brain 89: 499-530. Meldmm, B. (1985) Excitatory amino acids and anoxic/ischaemic brain damage. Trends Neurosci. February: 47-48. Milner, B. (1972) Disorders of learning and memory after temporal lobe lesions in man. Clin. Neurosurg. 19: 421-446. Milner. B.. and H.-L. Teuber (1968) Alteration ofnercention and memory in man: Reflections on methods. In Analyses if Behavioral Change, L. Weiskrantz, ed., pp. 268-375, Harper and Row, New York. Mishkin, M. (1978) Memory in monkeys severely impaired by combined but not by separate removal of amvgdala and hippocampus. _Nature 273: 297-298. Mishkin. M.. B. .I. Sniealer. R. C. Saunders. and B. J. Malamut (1982) An animal modelbf>obal amnesia. In ibward a Treatment if Alz: heimer’s Disease, S. Corkin, K. L. Davis, J. H. Growdon, E. Usdin, and R. J. Wurtman, eds., pp. 235-247, Raven, New York. Monaghan, D. T., V. R. Holets, D. W. Toy, and C. W. Cotman (1983) Anatomical distributions of four pharmacologically distinct 3H-Lglutamate binding sites. Nature 306: 176-l 79. Moss, M., H. Mahut, and S. Zola-Morgan (198 1) Concurrent discrimination learning in monkeys after hippocampal, entorhinal, or fomix lesions. J. Neurosci. I: 227-240. Muramoto, O., Y. Kum, M. Sugishita, and Y. Toyokura (1979) Pure memory loss with hippocampal lesions: A pneumoencephalographic study. Arch. Neurol. 36: 54-56. Nilges, R. G. (1944) The arteries of the mammalian comu ammonis. J. Comp. Neurol. 80: 177-190. Olney, J. W. (1978) Neurotoxicity of excitatory amino acids. In Kainic Acid as a Tool in Neurobiology, E. G. McGreer, J. W. Olney, and P. L. McGreer. eds.. DD. 95-121. Raven. New York. Olney, J. W. ‘( 1983) -Excitotoxms: An’overview. In Excitotoxins, K. Fuxe, P. Roberts, and R. Schwartz, eds., pp. 82-96, Macmillan, London. Olton, D. S., J. T. Becker, and G. E. Handelmann (1979) Hippocampus, space, and memory. Behav. Brain Sci. 2: 3 13-365. Osterreith, P. (1944) Le test de copie dune figure complexe. Arch. Psychol. 30: 306-356. Plum, F. (1983) What causes infarction in ischemic brain?: The Robert Wartenberg Lecture. Neurology 33: 222-233. Pulsinelli, W. A., and J. B. Brierley (1979) A new model of bilateral hemispheric ischemia in the unanesthetized rat. Stroke 10: 267-272. Rose, F. C., and C. P. Symonds (1960) Persistent memory defect following encephalitis. Brain 83: 195-2 12. Scoville, W. B. (1954) The limbic lobe and memory in man. J. Neurosurg. 11: 64. Scoville, W. B., and B. Milner (1957) Loss of recent memory after bilateral hippocampal lesions. J. Neurol. Neurosurg. Psychiatry 20: 11-21. Shimamura, A. P., and L. R. Squire (1984) Paired associate learning and priming effects in amnesia: A neuropsychological study. J. Exp. Psychol. (Gen.) 113: 556-570. Siesjo, B. K. (198 1) Cell damage in the brain: A speculative synthesis. J. Cereb. Blood Flow Metabol. 1: 155-185. Sloviter, R. S. (1983) “Epileptic” brain damage in rats induced by sustained electrical stimulation of the perforant path. I. Acute electrophysiological and light microscopic studies. Brain Res. Bull. 10: 675-697. Sommer, W. (1880) Erkrankung des Ammonshoms als aetiologisches Moment der Epilepsie. Arch. Psychiatr. Nervenkr. 10: 631-675. Squire, L. R. (1987a) Memory: Neural organization and behavior. In The Journal of Neuroscience Memory Impairment Following Handbook of Physiology: Higher Functions of the Nervous System, J. M. Brookhart and V. B. Mountcastle, eds., American Psychological Society, Bethesda, MD (in press). Squire, L. R. (1987b) Memory and Brain, Oxford U.P., New York (in press). Squire, L. R., and N. J. Cohen (1984) Human memory and amnesia. In Neurobiology of Learning and Memory, G. Lynch, J. L. McGaugh, and N. M. Weinberger, eds., pp. 3-64, Guilford, New York. Squire, L. R., and M. M. Fox (1980) Assessment of remote memory: Validation of the television test by repeated testing during a seven year period. Behav. Res. Meth. Instrum. 12: 583-586. Squire, L. R., and R. Y. Moore (1979) Dorsal thalamic lesion in a noted case of human memory dysfunction. Ann. Neurol. 6: 503-506. Squire, L. R., and S. Zola-Morgan (1983) The neurology of memory: The case for correspondence between the findings for human and nonhuman primate. In The Physiological Basis of Memory, 2nd Ed., J. A. Deutsch, ed., pp. 199-268, Academic, New York. Squire, L. R., and S. Zola-Morgan (1985) Neuropsychology of memory: New links between humans and experimental animals. In Conference on Memory Dysfunctions, D. Olton, S. Corkin, and E. Gamzu, eds., pp. 137-149, New York Academy of Sciences, New York. Squire, L. R., C. D. Wetzel, and P. C. Slater (1978) Anterograde amnesia following ECT: An analysis of the beneficial effect of partial information. Neuropsychologia 16: 339-347. Squire, L. R., N. J. Cohen, and L. Nadel (1984) The medial temporal region and memory consolidation: A new hypothesis. In Memory Consolidation, H. Weingartner and E. Parker, eds., pp. 185-210, Erlbaum, Hillsdale, NJ. Stephan, H. (1983) Evolutionary trends in limbic structures. Neurosci. Biobehav. Rev. 7: 367-374. Stephan, H., and J. Manolescu (1980) Comparative investigations on hippocampus in insectivores and primates. Z. Mikrosk. Anat. Forsch. 94: 1025-1050. Storm-Mathisen, J. (198 1) Glutamate in hippocampal pathways. Adv. Biochem. Psychopharmacol. 27: 43-55. Storm-Mathisen, J. (1984) Uptake of D-aspartate and L-glutamate in excitatory axon terminals in hippocampus: Autoradiographic and biochemical comparison with gamma-aminobutyrate and other amino acids in normal rats and rats with lesions. Neuroscience I1 I: 79-l 00. Talland, G. A. (1965) DerangedMemory, pp. l-356, Academic, New York. Teuber, H.-L., B. Milner, and H. C. Vaughan (1968) Persistent anterograde amnesia after stab wound of the basal brain. Neuropsychology 6: 267-282. Van Buren, J. M., and R. C. Borke (1972) The mesial temporal substratum of memory: Anatomical studies in three individuals. Brain 95: 599-632. Bilateral Hippocampal CA1 Lesion 2967 Van Hoesen, G. (1982) The parahippocampal gyrus: New observations regarding its cortical connections in the monkey. Trends Neurosci. 5: 345-350. Victor, M., R. D. Adams, and G. H. Collins (1971) The WernickeKorsakoffSyndrome, pp. l-206, Blackwell, Oxford. Volpe, B. T., and W. Hirst (1983) The characterization of an amnesic syndrome following hypoxic ischemic injury. Arch. Neurol. 40: 436440. Volpe, B. T., W. A. Pulsinelli, J. Tribuna, and H. P. Davis (1984) Behavioral performance of rats following transient forebrain ischemia. Stroke 15: 558-562. Weiskrantz, L. (1982) Comparative aspects of studies of amnesia. In Philosophical Transactions of the Royal Society of London, Vol. 298, D. E. Broadbent and L. Weiskrantz, eds., pp. 97-109, The Royal Society, London. White, B. C., J. G. Wiegenstein, and C. D. Winegar (1984) Brain ischemic anoxia. JAMA 251: 1586-l 590. Wieloch, T., 0. Lindvall, P. Blomqvist, and F. H. Gage (1985) Evidence for amelioration of ischaemic neuronal damage in the hippocampal formation by lesions of the perforant path. Neurol. Res. 7: 24-26. Wolf, G., G. Schunzel, and J. Storm-Mathisen (1984) Lesions of Schaffer’s collaterals in the rat hippocampus affecting glutamate dehydrogenase and succinate dehydrogenase activity in the stratum radiatum of CAl. A study with special reference to the glutamate transmitter metabolism. J. Himforsch. 25: 249-253. Woods. B. T.. W. Schoene. and L. Kneislev (1982) Are hiPDocamDa1 lesions sufficient to cause lasting amnesia? J: Neural. Neurosurg. Psychiatry 45: 243-247. Zola-Morgan, S., and L. R. Squire (1984) Preserved learning in monkeys with medial temporal lesions: Sparing of motor and cognitive skills. J. Neurosci. 4: 1072-1085. Zola-Moraan. S.. and L. R. Sauire (1985) Medial temnoral lesions in monkey& impair memory in a variety of tasks sensitive to amnesia. Behav. Neurosci. 99: 22-34. Zola-Morgan, S., and L. R. Squire (1986a) Memory impairment in monkeys following lesions limited to the hippocampus. Behav. Neurosci. 100: 155-160. Zola-Morgan, S., and L. R. Squire (1986b) Recovery of memory function following diencephalic damage but not medial temporal damage in monkeys. J. Neurosci. Abstr. (in press). Zola-Morgan, S., L. R. Squire, and M. Mishkin (1982) The neuroanatomy of amnesia: Amygdala-hippocampus versus temporal stem. Science 218: 1337-1339. Zola-Morgan, S., N. J. Cohen, and L. R. Squire (1983) Recall ofremote episodic memory in amnesia. Neuropsychologia 21: 487-500.