Epi-Care Center (BPH, ARW, GS), Baptist Memorial Hospital, Departments of Psychiatry (BPH), Neurosurgery (BPH, ARW), Biostatistics (GS), and Pathology (ADB, FCD), University of Tennessee, and Semmes-Murphey Clinic (BPH, ARW), Memphis, Tennessee Neurosurgery 31; 652-657, 1992 ABSTRACT: THIS INVESTIGATION TESTED the hypothesis that the degree of impairment to memory function caused by an anterior temporal lobectomy (ATL) is inversely related to the pathological status of the resected hippocampus. Specifically, the greatest risk to postoperative memory function should be to patients with no or minimal hippocampal sclerosis, i.e., those with a functional hippocampus. Forty patients who underwent a partial resection of the left (n = 21) or right (n = 19) anterior temporal lobe were administered tests of immediate and delayed verbal and figural memory, both preoperatively and 6 months postoperatively. The degree of postoperative impairment in memory function was then investigated as a function of the degree of hippocampal sclerosis, as determined by a standardized procedure. For a left ATL, an absence or mild degree of hippocampal sclerosis was associated with significantly greater postoperative impairment of both verbal and figural memory, compared with patients with moderate or marked sclerosis. No statistically significant relationship was noted for patients who underwent a right ATL, but the findings were in the same direction for five of the six memory measures. It may be possible to predict and avoid surgically induced impairment of memory function among patients who undergo left ATL through the use of preoperative hippocampal volumetric magnetic resonance imaging. Better clinical tests of right hippocampal function are needed to predict the outcome for patients who undergo a right ATL. KEY WORDS: Anterior temporal lobectomy; Epilepsy; Hippocampus; Memory The area of neuropsychological ability most likely to be adversely affected by anterior temporal lobectomy (ATL) is memory function (5,8). A variety of procedures are typically performed to reduce the possibility of causing either a surgically induced anterograde amnesia or otherwise clinically significant memory impairment. These techniques include preoperative intracarotid sodium Amytal memory testing (14) as well as intraoperative
PATIENTS AND METHODS A consecutive series of patients (n = 40) who underwent a partial resection of the left (dominant) (n = 21) or right (n = 19) anterior temporal lobe served as participants. The patients in this study met the following criteria: 1) they were nonretarded (WAIS-R Full Scale IQ > 69); 2) they were without magnetic resonance imaging evidence of foreign tissue structural lesions; and 3) they were left hemisphere dominant for speech, as determined by intracarotid sodium Amytal testing. In all cases, the patients' seizures were verified as being of unilateral temporal lobe origin by long-term electroencephalogram/video monitoring using bilateral subdural strip electrodes (26) . Table 1 provides information concerning the characteristics of the patients. There were no significant differences between the groups in any of the listed variables. Neuropsychological testing Memory was assessed using Russell's (20) revision of the Logical Memory and Visual Reproduction subtests of the Wechsler Memory Scale (25). This is a widely used procedure that assesses verbal and figural memory in both immediate and delayed recall conditions. The Logical Memory subtest consists of two complex prose paragraphs that are read aloud to the patient. The patient is asked to recall as much as possible after each presentation. Thirty minutes later, without a repeat presentation of the stories, the patient is asked to again recall as much as possible. Patients' responses are recorded verbatim and scored (per Russell's suggestion) according to the criteria of Schwartz and Ivnik (unpublished manuscript). Three dependent measures are obtained: 1) the total number of units recalled (of a possible 45) in the immediate recall condition; 2) the total number of units recalled in the delayed (30 minute) recall condition; and 3) a measure of forgetting called percentage retained, i.e.,
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
AUTHOR(S): Hermann, Bruce P., Ph.D.; Wyler, Allen R., M.D.; Somes, Grant, Ph.D.; Berry, Allen D., III, M.D.; Dohan, F. Curtis, Jr., M.D.
procedures such as cortical mapping of memory function (16). In spite of these precautions, outcomes vary as to the effects of ATL on postoperative memory function (5). Although there has been considerable speculation as to the cause of this variability in memory outcome, there remains substantial uncertainty (2). We developed a grading scale for hippocampal pathological status, and in the course of validating this scale, we searched for a relationship between memory outcome and the severity of the hippocampal abnormality. Recent reports have indicated a relationship between the severity of preoperative memory impairment and the degree of abnormality detected in the resected hippocampus (21). If the degree of abnormality is indeed an index of the functional status of the hippocampus (as reflected in the adequacy of memory performance), then the degree of postoperative memory loss should be inversely related to the amount of hippocampal abnormality. The purpose of this investigation was to test this hypothesis.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
Neurosurgery 1992-98 October 1992, Volume 31, Number 4 652 Pathological Status of the Mesial Temporal Lobe Predicts Memory Outcome from Left Anterior Temporal Lobectomy Clinical Study
Hippocampal grading system The hippocampal abnormality grading system was based on reviewing serial sections of hippocampus removed en bloc (see description below). After reviewing the entire hippocampal specimen, the sections showing the most severe involvement were used to grade the abnormality. Details regarding the development and refinement of the grading system are described in a companion publication (26). The grading system yielded the following categories: No hippocampal abnormality present. None of the conditions that describe the following grades was met. Grade 1. Mild mesial temporal damage (MTD): Mild damage (gliosis with slight [< 10%] or no neuronal dropout) involving sectors CA1, CA3, and/or CA4 of hippocampal pyramidal cell layer. Grade 2. Moderate MTD: Moderate damage (gliosis with moderate [10-50%] neuronal dropout) involving sectors CA1, CA3, and/or CA4 of hippocampal pyramidal cell layer. (If involvement is limited to CA3 and CA4, the lesion can be designated "end folium sclerosis.")
Data analyses Difference scores (postoperative score minus preoperative score) were computed for the six memory measures (i.e., immediate, delayed, and percentage retained indices for the Logical Memory and Visual Reproduction subtests). In this way, negative difference scores indicated postoperative impairment in memory function and positive difference scores indicated postoperative memory improvement. Patients were classified into two groups based on the neuropathological results-none/mild (Grades 0 and 1 combined) versus moderate/marked hippocampal sclerosis (Grades 2, 3, and 4 combined). Pre- to postoperative difference scores on the memory tests were examined as a function of the degree of hippocampal abnormality. This was done within the left and right ATL groups using two tailed t-tests. To rule out the possibility that the results were confounded by age or alteration in language function, analyses of covariance were also computed. Throughout the investigation, neuropathological analysis was performed while investigators were blinded to the neuropsychological results and vice versa. RESULTS Mean difference scores on the memory measures for the none/mild and moderate/marked hippocampal sclerosis groups are shown in Table 2. For patients who underwent left temporal lobe resection, greater memory loss for both verbal (Logical MemoryImmediate and Logical Memory-Delayed) and nonverbal material (Visual Reproduction-Percentage Retained) was significantly associated (P < 0.05) with the no/mild degree of hippocampal sclerosis. A similar trend (P < 0.10) was noted for Visual Reproduction-Delayed. For patients who underwent right temporal lobectomy, no significant relationship was noted between the degree of hippocampal sclerosis and postoperative memory loss, although it should be
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Surgical specimens Hippocampal specimens were obtained in the following way. All patients underwent ATL under general anesthesia. After routine exposure of the temporal lobe, the anterior lateral 4.5 cm of temporal neocortex was removed leaving the temporal horn of the lateral ventricle exposed along its anterior length. The resection included the superior through inferior temporal gyri, leaving the hippocampus, parahippocampus, and fusiform gyri. The fusiform gyrus was then resected, leaving only the hippocampal and parahippocampal gyri. A single silver clip was applied to the anterior edge of the hippocampus and then the hippocampus was removed en bloc to the posterior margin of the cerebral peduncle. Two silver clips were applied to the posterior margin of the hippocampus. By placing the single and double silver clips on the anterior and posterior margins, the specimen could be aligned after fixation to facilitate serial sectioning along its axis. The entire dissection of hippocampus was done using the operating microscope. After removal, the specimen was placed in formalin, fixed, and then serially sectioned along its long axis, which, although it is somewhat curved, was mainly oriented along the anteroposterior axis.
Grade 3. Moderate to marked MTD (hippocampal sclerosis or "classical" Ammon's horn sclerosis): Severe damage (gliosis with > 50% neuronal dropout) involving sectors CA1, CA3, and CA4 of hippocampal pyramidal cell layer but sparing CA2, the resistant sector. Grade 4. Marked MTD (hippocampal sclerosis or "total" Ammon's horn sclerosis): Severe damage (gliosis with > 50% neuronal dropout) involving all sectors of hippocampal pyramidal cell layer. The fascia dentata, subiculum, and parahippocampal gyrus may also be involved. Also specified in the surgical report are the sectors of the hippocampal pyramidal cell layer (CA1, CA2, CA3, CA4) that are present and identifiable, the number of sections in which they are present, and whether they are present in continuity. The presence of all the sectors in continuity provides the surgeon with a direct measure of how far posteriorly the hippocampal resection went, because not all the sectors are in continuity anteriorly at the level of the pes hippocampi.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
the number of units remembered at delayed (30 minutes) recall divided by the number of units remembered at immediate recall. The smaller the percentage, the greater the forgetting that took place over time. The Visual Reproduction subtest involves brief (10 second) presentations of three cards containing visual stimuli, after each of which the patient is asked to draw the items from memory. Thirty minutes later, the patient is again asked to draw the figures from memory. As with the Logical Memory subtest, three dependent measures are obtained: immediate recall, delayed (30 minute) recall, and percentage retained.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
DISCUSSION Investigation of patients after ATL has contributed greatly to models of memory function and memory abnormality (13). These previous investigations have demonstrated the important contributions of the medial temporal structures, especially the hippocampus, to a variety of learning and memory tasks (23). Because of this theoretical understanding, in association with clinical reports demonstrating that the greatest threat to neuropsychological status following ATL concerns memory function (5,8), special concern has surrounded the issue of postoperative memory loss. Particularly difficult to understand has been the variable effect of ATL on postoperative memory status. Although anterograde amnesias are rare, some patients clearly demonstrate clinically significant memory impairment, which represents an impediment in their day-to-day functioning. However, other patients report and demonstrate memory improvement, while others show no meaningful change (5,8). Explaining this variability in postoperative memory outcome, in the context of a standardized surgical approach, has proven difficult (2,18) . Previous investigations have been attempted to understand and predict postoperative memory outcome by considering patient characteristics, such as age, age at onset of epilepsy, and duration of epilepsy, or by considering the basis of operative variables, such as the effects of surgery on seizure frequency, or variations in the extent of hippocampal resection (2). This investigation was stimulated by the observation that patients vary as to the nature and severity of the abnormality evident in the resected hippocampus (12). If it is assumed that the functional integrity of the hippocampus (and hence the adequacy of memory function) is increasingly compromised by the severity of hippocampal sclerosis (21), then the degree of postoperative memory impairment experienced by a patient should be inversely related to the severity of the abnormality. Those patients with moderate or severe hippocampal sclerosis (i.e., with an essentially nonfunctional hippocampus), should show minimal or no memory loss. Conversely, patients without hippocampal sclerosis, or with only a mild degree, probably have a functional hippocampus, and its resection is more likely to cause postoperative memory difficulties. These predictions held true for patients who underwent a left ATL. Patients with no/mild hippocampal sclerosis (in the left hippocampus) showed significantly greater impairment in their ability to recall prose (both immediately and 30 minutes later) and forgot visual material at a faster
rate compared with patients with moderate/marked hippocampal sclerosis. These findings were not influenced by patient age or change in language function. No such relationship was found for patients who underwent a right ATL, although the same trend was seen for five of the six memory measures. In summary, our hypothesis held for a left but not a right ATL. Oxbury and Oxbury (17) were among the first to study the relationship between hippocampal pathology and memory outcome. They examined composite measures of immediate and delayed verbal memory among patients who underwent a left or right ATL, and they related changes in memory function to the pathological status of the resected hippocampus (Ammon's horn sclerosis vs. an intact hippocampus). The most precipitous decline across both immediate and delayed verbal memory was shown by patients who underwent a left ATL and whose resected hippocampus did not reveal Ammon's horn sclerosis (hippocampus intact group) (17). They also noted that "... the most bitter complaints that operation has precipitated memory impairment come from patients whose preoperative memory was good and whose lobectomy has included left hippocampus (+/amygdala) with only minor or no neuronal depletion" (17) (p. 147). Given this basic similarity in findings, it is important to note that there are differences between our investigation and the Oxbury and Oxbury study in several important characteristics, i.e., the age range of the patients, the indices of memory investigated, the definitions and methods used to assess hippocampal abnormality, and the degree to which potentially relevant confounding variables are controlled (e.g., pre- to postoperative language change), which are most likely responsible for those points on which there is not agreement. To understand our differential success in predicting memory outcome after a left versus right ATL, one needs to consider the current status of memory research and memory assessment and the place of the Wechsler Memory Scale in that literature. The predominant model of memory function at the present time is the so-called material-specific model (10) that posits that the left and right hemispheres in general, and medial regions of those hemispheres in particular, are specialized for the learning and recall of verbal and nonverbal visuospatial material, respectively. It is generally agreed (2) that the more robust finding has been the deficit in verbal learning and recall ability after a left ATL (3,4,9,11,15,16,18,22), an impairment we have been able to predict via analysis of hippocampal abnormality. It has proven considerably more difficult to detect visual/figural memory deficits after a right ATL (9,11,15,18,22). Our pattern of results clearly mirror this trend, but there is consensus that the Visual Reproduction subtest of the Wechsler Memory Scale is not especially sensitive to right hippocampal function. The ability to predict memory impairments after a right ATL probably needs to await a better understanding of right hippocampal function and the development of appropriate clinical tests. The one unusual finding in our results was that left
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
noted that for five of the six memory measures, the findings were in the same direction noted above. When chronological age and adequacy of expressive language function were taken into consideration via analyses of covariance, the identical statistically significant relationships between degree of hippocampal sclerosis and memory loss remained.
Received, November 18, 1991. Accepted, May 18, 1992. Reprint requests: Bruce P. Hermann, Ph.D., EpiCare Center, 910 Madison (Suite 906), Memphis, TN 38103. REFERENCES: (1-27) 1.
Cascino GD, Jack CR, Parisi JE, Sharbrough
3.
4.
5. 6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
ACKNOWLEDGMENT This work has been supported in part by Grant NS28712 awarded by the National Institutes of Health.
2.
FW, Hirschorn KA, Meyer FB, Marsh WR, O'Brien PC: Magnetic resonance imagingbased volume studies in temporal lobe epilepsy: Pathological correlations. Ann Neurol 30:31-36, 1991. Chelune GJ: Using neuropsychological data to forecast postsurgical cognitive outcome, in Luders H (ed): Epilepsy Surgery. New York, Raven Press, 1991, pp 477-485. Chelune GJ, Naugle R, Luders H, Iwad IA: Prediction of cognitive change as a function of preoperative ability status among temporal lobectomy patients seen at 6-month followup. Neurology 41:399- 404, 1990. Frisk V, Milner B: The relationship of working memory to the immediate recall of stories following unilateral temporal or frontal lobectomy. Neuropsychologia 28:121-135, 1990. Hermann BP, Wyler AR: Neuropsychological outcome of anterior temporal lobectomy. J Epilepsy 1:35-45, 1988. Hermann BP, Wyler AR, Steeman H, Richey ET: The interrelationship between language function and verbal learning/memory performance in patients with complex partial seizures. Cortex 24:245-253, 1988. Hermann BP, Seidenberg M, Haltiner A, Wyler AR: Adequacy of language function and verbal memory performance in unilateral temporal lobe epilepsy. Cortex (in press). Ivnik RJ, Sharbough FW, Laws ER: Effects of anterior temporal lobectomy on cognitive function. J Clin Psychol 43:128-137, 1987. Ivnik RJ, Sharbrough FW, Laws ER: Anterior temporal lobectomy for the control of partial complex seizures: Information for counseling patients. Mayo Clin Proc 63:783-793, 1988. Jones-Gotman M: Commentary: Psychological evaluation-testing hippocampal function, in Engel J (ed): Surgical Treatment of the Epilepsies. New York, Raven Press, 1987, pp 203-211. Lee GP, Loring DW, Thompson JL: Construct validity of material- specific memory measures following unilateral temporal lobectomy. Psychol Assess 1:192-197, 1989. Meldrum B, Corsellis JAN: Epilepsy, in Adams JH, Corsellis JAN, Duchen LW (eds): Greenfield's Neuropathology. New York, John Wiley, 1985, ed 4, pp 921-950. Milner B: Psychological aspects of focal epilepsy and its neurosurgical management, in Purpura DP, Penry JK, Walter RD (eds): Advances in Neurology. New York, Raven Press, 1975, pp 299-321. Milner B, Branch C, Rasmussen T: A study of short term memory after intracarotid injection of sodium Amytal. Trans Am Neurol Assoc 87:244-266, 1962. Novelly RA, Augustine EA, Mattson RH, Glaser G, Williamson P, Spencer D, Spencer S: Selective memory improvement and impairment in temporal lobectomy for
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
ATL patients with no/mild hippocampal sclerosis also showed significant postoperative decline in some aspects of figural memory. The reason is probably that the figures to be remembered can be encoded by the patient in a verbal fashion (e.g., a triangle with four crosses inside). Thus, more rapid forgetting of the figures may actually be another manifestation of a verbal memory deficit and are consistent with the results of prose recall. It has been demonstrated that volumetric magnetic resonance imaging analysis of the temporal lobes can predict the presence of hippocampal sclerosis (1). If the occurrence of postoperative memory deficits after a left ATL were found to be associated with normal hippocampal volumes (i.e., absence of hippocampal sclerosis), then it might be possible to avoid this serious complication of ATL. This approach appears promising as Trenerry et al. (24) recently presented data indicating that left temporal lobectomy patients with larger left hippocampi tended to have a greater postsurgical verbal memory decline (Wechsler Memory Scale-Revised). Another test used to predict postoperative memory patency is the intracarotid Amytal test. Previous findings raise the possibility that perfusion of the hemisphere contralateral to the epileptic temporal lobe, with assessment of the memory capability of the to-be-resected hippocampus, will yield information regarding its functional and hence pathological status (19) , perhaps yielding some idea as to the risk of postoperative memory loss. Finally, it has been suggested that the extent of verbal memory deficits may be correlated with the extent of lateral temporal lobe resection (16). Given the relationship between verbal memory performance and the adequacy of basic language function (6,7) and the probable increased risk of adversely affecting language ability that is related to a more extensive resection of dominant lateral temporal neocortex, alterations in language ability (e.g., postoperative dysnomia) need to be considered when examining the effects of lateral dominant temporal resection on memory outcome. To our knowledge, this has not been done. Analyses in the present investigation controlled for any surgically induced alterations in language ability and, therefore, indicate that the declines in verbal memory associated with resection of an intact left hippocampus reflect real memory impairment.
18. 19.
20. 21.
22.
23. 24.
25. 26. 27.
COMMENTS The authors present the interesting hypothesis that postoperative memory deficits after temporal lobectomy may occur in those patients who had the most normal preoperative mesial temporal lobe structures. This is a plausible hypothesis and is certainly supported by current theoretical constructs. In addition this article raises a number of other interesting questions. In this series, it appears that the degree of lateral removal was held constant. What added or differential effect do they propose would occur in response to differences in lateral resection? Second, the current concept is that the hippocampus should be removed whenever possible in patients
with temporal lobe epilepsy and that this structure is likely to be the source of the epilepsy in many cases. Would normal structural hippocampi occur in patients who have nonhippocampal temporal lobe epilepsy? Would seizure control be less adequate if this structure were left in place? Third, the authors note that all patients were left temporal lobe dominant on the intracarotid sodium amobarbital test. How did the results of memory testing during this examination correlate with pathological changes and with postoperative memory deficits? Ronald P. Lesser Baltimore, Maryland This is an interesting report that correlates the presence of hippocampal abnormality in the surgically excised temporal lobe with a neurocognitive outcome in patients undergoing anterior temporal lobectomy for intractable epilepsy. The authors clearly demonstrated that the patients with pathologically verified hippocampal neuronal cell loss experienced the least impairment in memory function postoperatively. This observation was statistically significant only for patients receiving left anterior temporal lobe surgery. Potentially, the preoperative studies assessing the presence of hippocampal abnormality may be of predictive value in assessing the neurocognitive outcome after temporal lobe surgery. Magnetic resonance imaging-based hippocampal volumetry has been demonstrated to be a reliable marker of moderate to severe mesial temporal sclerosis (1). One study has indicated that the presence of magnetic resonance imaging-identified hippocampal abnormality may correlate with memory before and after temporal lobe surgery (2). Patients with partial epilepsy of left temporal lobe origin appear to be at greater risk of experiencing a memory decline postoperatively if the preoperative volumetric study fails to reveal hippocampal formation atrophy (2). The present study by Hermann et al. provides additional evidence that preoperative assessments of hippocampal abnormality should be an intimate part of a comprehensive presurgical evaluation of patients with partial epilepsy of temporal lobe origin. Gregory D. Cascino, M.D. Rochester, Minnesota REFERENCES: (1,2) 1.
2.
Cascino GD, Jack CR, Parisi JE, Sharbrough FW, Hirschorn KA, Meyer FB, Marsh WR, O'Brien PC: Magnetic resonance imagingbased volume studies in temporal lobe epilepsy: Pathological correlations. Ann Neurol 30:31-36, 1991. Trenerry MR, Jack CR, Ivnik RJ, Sharbrough FW, Cascino GD, Hirschorn KA, Marsh WR, Kelly PJ, Meyer FB: Memory is correlated with presurgical magnetic resonance imaging: Hypocampal volumes before and after temporal lobectomy for intractable epilepsy.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
17.
epilepsy. Ann Neurol 15:64- 67, 1984. Ojemann G, Dodrill C: Verbal memory deficits after left temporal lobectomy for epilepsy: Mechanism and prediction. J Neurosurg 62:101-107, 1985. Oxbury J, Oxbury S: Neuropsychology: Memory and hippocampal pathology, in Reynolds EH, Trimble MR (eds): The Bridge between Neurology and Psychiatry. Edinburgh, Churchill Livingstone, 1989, pp 136-150. Rausch R: Effects of temporal lobe surgery on behavior. Adv Neurol 55:279-292, 1991. Rausch R, Babb TL, Engel J, Crandall PH: Memory following amobarbital injection contralateral to hippocampal damage. Arch Neurol 46:783-788, 1989. Russell EW: Renorming Russell's version of the Wechsler Memory Scale. J Clin Exp Neuropsychiatr 10:235-249, 1988. Sass KJ, Spencer DD, Kim JH, Westerveld M, Novelly RA, Lencz T: Verbal memory impairment correlates with hippocampal pryamidal cell density. Neurology 40:16941697, 1990. Saykin AJ, Gur RC, Sussman NM, O'Connor MJ, Gur RE: Memory deficits before and after temporal lobectomy: Effects of laterality and age of onset. Brain Cogn 9:191-200, 1989. Squire LR: Memory and Brain . New York, Oxford University Press, 1987. Trenerry MR, Jack CR, Ivnik RJ, Sharbrough FW, Cascino GD, Hirschorn, KA, Marsh WR, Kelly PJ, Meyer FB: Memory is correlated with presurgical magnetic resonance imaging hippocampal volumes before and after temporal lobectomy for intractable epilepsy. Epilepsia 32(Suppl 3):73, 1991 (abstr). Wechsler D, Stone CP: Wechsler Memory Scale. New York, Psychological Corporation, 1945. Wyler AR, Dohan FC, Schweitzer JB, Berry AD: A grading system for hippocampal sclerosis. (in press). Journal of Epilepsy. Wyler AR, Ojemann G, Lettich E, Ward AA, Jr: Subdural strip electrodes for localizing epileptogenic foci. J Neurosurg 60:11951200, 1984.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
16.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
Epilepsia 32(Suppl 3):73, 1991.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Table 2. Relationship of Postoperative Memory Decline with Degree of Hippocampal Sclerosis
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
Table 1. Patient Characteristics
PATIENTS AND METHODS A consecutive series of patients (n = 40) who underwent a partial resection of the left (dominant) (n = 21) or right (n = 19) anterior temporal lobe served as participants. The patients in this study met the following criteria: 1) they were nonretarded (WAIS-R Full Scale IQ > 69); 2) they were without magnetic resonance imaging evidence of foreign tissue structural lesions; and 3) they were left hemisphere dominant for speech, as determined by intracarotid sodium Amytal testing. In all cases, the patients' seizures were verified as being of unilateral temporal lobe origin by long-term electroencephalogram/video monitoring using bilateral subdural strip electrodes (26) . Table 1 provides information concerning the characteristics of the patients. There were no significant differences between the groups in any of the listed variables. Neuropsychological testing Memory was assessed using Russell's (20) revision of the Logical Memory and Visual Reproduction subtests of the Wechsler Memory Scale (25). This is a widely used procedure that assesses verbal and figural memory in both immediate and delayed recall conditions. The Logical Memory subtest consists of two complex prose paragraphs that are read aloud to the patient. The patient is asked to recall as much as possible after each presentation. Thirty minutes later, without a repeat presentation of the stories, the patient is asked to again recall as much as possible. Patients' responses are recorded verbatim and scored (per Russell's suggestion) according to the criteria of Schwartz and Ivnik (unpublished manuscript). Three dependent measures are obtained: 1) the total number of units recalled (of a possible 45) in the immediate recall condition; 2) the total number of units recalled in the delayed (30 minute) recall condition; and 3) a measure of forgetting called percentage retained, i.e.,
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
AUTHOR(S): Hermann, Bruce P., Ph.D.; Wyler, Allen R., M.D.; Somes, Grant, Ph.D.; Berry, Allen D., III, M.D.; Dohan, F. Curtis, Jr., M.D.
procedures such as cortical mapping of memory function (16). In spite of these precautions, outcomes vary as to the effects of ATL on postoperative memory function (5). Although there has been considerable speculation as to the cause of this variability in memory outcome, there remains substantial uncertainty (2). We developed a grading scale for hippocampal pathological status, and in the course of validating this scale, we searched for a relationship between memory outcome and the severity of the hippocampal abnormality. Recent reports have indicated a relationship between the severity of preoperative memory impairment and the degree of abnormality detected in the resected hippocampus (21). If the degree of abnormality is indeed an index of the functional status of the hippocampus (as reflected in the adequacy of memory performance), then the degree of postoperative memory loss should be inversely related to the amount of hippocampal abnormality. The purpose of this investigation was to test this hypothesis.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
Neurosurgery 1992-98 October 1992, Volume 31, Number 4 652 Pathological Status of the Mesial Temporal Lobe Predicts Memory Outcome from Left Anterior Temporal Lobectomy Clinical Study
Hippocampal grading system The hippocampal abnormality grading system was based on reviewing serial sections of hippocampus removed en bloc (see description below). After reviewing the entire hippocampal specimen, the sections showing the most severe involvement were used to grade the abnormality. Details regarding the development and refinement of the grading system are described in a companion publication (26). The grading system yielded the following categories: No hippocampal abnormality present. None of the conditions that describe the following grades was met. Grade 1. Mild mesial temporal damage (MTD): Mild damage (gliosis with slight [< 10%] or no neuronal dropout) involving sectors CA1, CA3, and/or CA4 of hippocampal pyramidal cell layer. Grade 2. Moderate MTD: Moderate damage (gliosis with moderate [10-50%] neuronal dropout) involving sectors CA1, CA3, and/or CA4 of hippocampal pyramidal cell layer. (If involvement is limited to CA3 and CA4, the lesion can be designated "end folium sclerosis.")
Data analyses Difference scores (postoperative score minus preoperative score) were computed for the six memory measures (i.e., immediate, delayed, and percentage retained indices for the Logical Memory and Visual Reproduction subtests). In this way, negative difference scores indicated postoperative impairment in memory function and positive difference scores indicated postoperative memory improvement. Patients were classified into two groups based on the neuropathological results-none/mild (Grades 0 and 1 combined) versus moderate/marked hippocampal sclerosis (Grades 2, 3, and 4 combined). Pre- to postoperative difference scores on the memory tests were examined as a function of the degree of hippocampal abnormality. This was done within the left and right ATL groups using two tailed t-tests. To rule out the possibility that the results were confounded by age or alteration in language function, analyses of covariance were also computed. Throughout the investigation, neuropathological analysis was performed while investigators were blinded to the neuropsychological results and vice versa. RESULTS Mean difference scores on the memory measures for the none/mild and moderate/marked hippocampal sclerosis groups are shown in Table 2. For patients who underwent left temporal lobe resection, greater memory loss for both verbal (Logical MemoryImmediate and Logical Memory-Delayed) and nonverbal material (Visual Reproduction-Percentage Retained) was significantly associated (P < 0.05) with the no/mild degree of hippocampal sclerosis. A similar trend (P < 0.10) was noted for Visual Reproduction-Delayed. For patients who underwent right temporal lobectomy, no significant relationship was noted between the degree of hippocampal sclerosis and postoperative memory loss, although it should be
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Surgical specimens Hippocampal specimens were obtained in the following way. All patients underwent ATL under general anesthesia. After routine exposure of the temporal lobe, the anterior lateral 4.5 cm of temporal neocortex was removed leaving the temporal horn of the lateral ventricle exposed along its anterior length. The resection included the superior through inferior temporal gyri, leaving the hippocampus, parahippocampus, and fusiform gyri. The fusiform gyrus was then resected, leaving only the hippocampal and parahippocampal gyri. A single silver clip was applied to the anterior edge of the hippocampus and then the hippocampus was removed en bloc to the posterior margin of the cerebral peduncle. Two silver clips were applied to the posterior margin of the hippocampus. By placing the single and double silver clips on the anterior and posterior margins, the specimen could be aligned after fixation to facilitate serial sectioning along its axis. The entire dissection of hippocampus was done using the operating microscope. After removal, the specimen was placed in formalin, fixed, and then serially sectioned along its long axis, which, although it is somewhat curved, was mainly oriented along the anteroposterior axis.
Grade 3. Moderate to marked MTD (hippocampal sclerosis or "classical" Ammon's horn sclerosis): Severe damage (gliosis with > 50% neuronal dropout) involving sectors CA1, CA3, and CA4 of hippocampal pyramidal cell layer but sparing CA2, the resistant sector. Grade 4. Marked MTD (hippocampal sclerosis or "total" Ammon's horn sclerosis): Severe damage (gliosis with > 50% neuronal dropout) involving all sectors of hippocampal pyramidal cell layer. The fascia dentata, subiculum, and parahippocampal gyrus may also be involved. Also specified in the surgical report are the sectors of the hippocampal pyramidal cell layer (CA1, CA2, CA3, CA4) that are present and identifiable, the number of sections in which they are present, and whether they are present in continuity. The presence of all the sectors in continuity provides the surgeon with a direct measure of how far posteriorly the hippocampal resection went, because not all the sectors are in continuity anteriorly at the level of the pes hippocampi.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
the number of units remembered at delayed (30 minutes) recall divided by the number of units remembered at immediate recall. The smaller the percentage, the greater the forgetting that took place over time. The Visual Reproduction subtest involves brief (10 second) presentations of three cards containing visual stimuli, after each of which the patient is asked to draw the items from memory. Thirty minutes later, the patient is again asked to draw the figures from memory. As with the Logical Memory subtest, three dependent measures are obtained: immediate recall, delayed (30 minute) recall, and percentage retained.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
DISCUSSION Investigation of patients after ATL has contributed greatly to models of memory function and memory abnormality (13). These previous investigations have demonstrated the important contributions of the medial temporal structures, especially the hippocampus, to a variety of learning and memory tasks (23). Because of this theoretical understanding, in association with clinical reports demonstrating that the greatest threat to neuropsychological status following ATL concerns memory function (5,8), special concern has surrounded the issue of postoperative memory loss. Particularly difficult to understand has been the variable effect of ATL on postoperative memory status. Although anterograde amnesias are rare, some patients clearly demonstrate clinically significant memory impairment, which represents an impediment in their day-to-day functioning. However, other patients report and demonstrate memory improvement, while others show no meaningful change (5,8). Explaining this variability in postoperative memory outcome, in the context of a standardized surgical approach, has proven difficult (2,18) . Previous investigations have been attempted to understand and predict postoperative memory outcome by considering patient characteristics, such as age, age at onset of epilepsy, and duration of epilepsy, or by considering the basis of operative variables, such as the effects of surgery on seizure frequency, or variations in the extent of hippocampal resection (2). This investigation was stimulated by the observation that patients vary as to the nature and severity of the abnormality evident in the resected hippocampus (12). If it is assumed that the functional integrity of the hippocampus (and hence the adequacy of memory function) is increasingly compromised by the severity of hippocampal sclerosis (21), then the degree of postoperative memory impairment experienced by a patient should be inversely related to the severity of the abnormality. Those patients with moderate or severe hippocampal sclerosis (i.e., with an essentially nonfunctional hippocampus), should show minimal or no memory loss. Conversely, patients without hippocampal sclerosis, or with only a mild degree, probably have a functional hippocampus, and its resection is more likely to cause postoperative memory difficulties. These predictions held true for patients who underwent a left ATL. Patients with no/mild hippocampal sclerosis (in the left hippocampus) showed significantly greater impairment in their ability to recall prose (both immediately and 30 minutes later) and forgot visual material at a faster
rate compared with patients with moderate/marked hippocampal sclerosis. These findings were not influenced by patient age or change in language function. No such relationship was found for patients who underwent a right ATL, although the same trend was seen for five of the six memory measures. In summary, our hypothesis held for a left but not a right ATL. Oxbury and Oxbury (17) were among the first to study the relationship between hippocampal pathology and memory outcome. They examined composite measures of immediate and delayed verbal memory among patients who underwent a left or right ATL, and they related changes in memory function to the pathological status of the resected hippocampus (Ammon's horn sclerosis vs. an intact hippocampus). The most precipitous decline across both immediate and delayed verbal memory was shown by patients who underwent a left ATL and whose resected hippocampus did not reveal Ammon's horn sclerosis (hippocampus intact group) (17). They also noted that "... the most bitter complaints that operation has precipitated memory impairment come from patients whose preoperative memory was good and whose lobectomy has included left hippocampus (+/amygdala) with only minor or no neuronal depletion" (17) (p. 147). Given this basic similarity in findings, it is important to note that there are differences between our investigation and the Oxbury and Oxbury study in several important characteristics, i.e., the age range of the patients, the indices of memory investigated, the definitions and methods used to assess hippocampal abnormality, and the degree to which potentially relevant confounding variables are controlled (e.g., pre- to postoperative language change), which are most likely responsible for those points on which there is not agreement. To understand our differential success in predicting memory outcome after a left versus right ATL, one needs to consider the current status of memory research and memory assessment and the place of the Wechsler Memory Scale in that literature. The predominant model of memory function at the present time is the so-called material-specific model (10) that posits that the left and right hemispheres in general, and medial regions of those hemispheres in particular, are specialized for the learning and recall of verbal and nonverbal visuospatial material, respectively. It is generally agreed (2) that the more robust finding has been the deficit in verbal learning and recall ability after a left ATL (3,4,9,11,15,16,18,22), an impairment we have been able to predict via analysis of hippocampal abnormality. It has proven considerably more difficult to detect visual/figural memory deficits after a right ATL (9,11,15,18,22). Our pattern of results clearly mirror this trend, but there is consensus that the Visual Reproduction subtest of the Wechsler Memory Scale is not especially sensitive to right hippocampal function. The ability to predict memory impairments after a right ATL probably needs to await a better understanding of right hippocampal function and the development of appropriate clinical tests. The one unusual finding in our results was that left
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
noted that for five of the six memory measures, the findings were in the same direction noted above. When chronological age and adequacy of expressive language function were taken into consideration via analyses of covariance, the identical statistically significant relationships between degree of hippocampal sclerosis and memory loss remained.
Received, November 18, 1991. Accepted, May 18, 1992. Reprint requests: Bruce P. Hermann, Ph.D., EpiCare Center, 910 Madison (Suite 906), Memphis, TN 38103. REFERENCES: (1-27) 1.
Cascino GD, Jack CR, Parisi JE, Sharbrough
3.
4.
5. 6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
ACKNOWLEDGMENT This work has been supported in part by Grant NS28712 awarded by the National Institutes of Health.
2.
FW, Hirschorn KA, Meyer FB, Marsh WR, O'Brien PC: Magnetic resonance imagingbased volume studies in temporal lobe epilepsy: Pathological correlations. Ann Neurol 30:31-36, 1991. Chelune GJ: Using neuropsychological data to forecast postsurgical cognitive outcome, in Luders H (ed): Epilepsy Surgery. New York, Raven Press, 1991, pp 477-485. Chelune GJ, Naugle R, Luders H, Iwad IA: Prediction of cognitive change as a function of preoperative ability status among temporal lobectomy patients seen at 6-month followup. Neurology 41:399- 404, 1990. Frisk V, Milner B: The relationship of working memory to the immediate recall of stories following unilateral temporal or frontal lobectomy. Neuropsychologia 28:121-135, 1990. Hermann BP, Wyler AR: Neuropsychological outcome of anterior temporal lobectomy. J Epilepsy 1:35-45, 1988. Hermann BP, Wyler AR, Steeman H, Richey ET: The interrelationship between language function and verbal learning/memory performance in patients with complex partial seizures. Cortex 24:245-253, 1988. Hermann BP, Seidenberg M, Haltiner A, Wyler AR: Adequacy of language function and verbal memory performance in unilateral temporal lobe epilepsy. Cortex (in press). Ivnik RJ, Sharbough FW, Laws ER: Effects of anterior temporal lobectomy on cognitive function. J Clin Psychol 43:128-137, 1987. Ivnik RJ, Sharbrough FW, Laws ER: Anterior temporal lobectomy for the control of partial complex seizures: Information for counseling patients. Mayo Clin Proc 63:783-793, 1988. Jones-Gotman M: Commentary: Psychological evaluation-testing hippocampal function, in Engel J (ed): Surgical Treatment of the Epilepsies. New York, Raven Press, 1987, pp 203-211. Lee GP, Loring DW, Thompson JL: Construct validity of material- specific memory measures following unilateral temporal lobectomy. Psychol Assess 1:192-197, 1989. Meldrum B, Corsellis JAN: Epilepsy, in Adams JH, Corsellis JAN, Duchen LW (eds): Greenfield's Neuropathology. New York, John Wiley, 1985, ed 4, pp 921-950. Milner B: Psychological aspects of focal epilepsy and its neurosurgical management, in Purpura DP, Penry JK, Walter RD (eds): Advances in Neurology. New York, Raven Press, 1975, pp 299-321. Milner B, Branch C, Rasmussen T: A study of short term memory after intracarotid injection of sodium Amytal. Trans Am Neurol Assoc 87:244-266, 1962. Novelly RA, Augustine EA, Mattson RH, Glaser G, Williamson P, Spencer D, Spencer S: Selective memory improvement and impairment in temporal lobectomy for
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
ATL patients with no/mild hippocampal sclerosis also showed significant postoperative decline in some aspects of figural memory. The reason is probably that the figures to be remembered can be encoded by the patient in a verbal fashion (e.g., a triangle with four crosses inside). Thus, more rapid forgetting of the figures may actually be another manifestation of a verbal memory deficit and are consistent with the results of prose recall. It has been demonstrated that volumetric magnetic resonance imaging analysis of the temporal lobes can predict the presence of hippocampal sclerosis (1). If the occurrence of postoperative memory deficits after a left ATL were found to be associated with normal hippocampal volumes (i.e., absence of hippocampal sclerosis), then it might be possible to avoid this serious complication of ATL. This approach appears promising as Trenerry et al. (24) recently presented data indicating that left temporal lobectomy patients with larger left hippocampi tended to have a greater postsurgical verbal memory decline (Wechsler Memory Scale-Revised). Another test used to predict postoperative memory patency is the intracarotid Amytal test. Previous findings raise the possibility that perfusion of the hemisphere contralateral to the epileptic temporal lobe, with assessment of the memory capability of the to-be-resected hippocampus, will yield information regarding its functional and hence pathological status (19) , perhaps yielding some idea as to the risk of postoperative memory loss. Finally, it has been suggested that the extent of verbal memory deficits may be correlated with the extent of lateral temporal lobe resection (16). Given the relationship between verbal memory performance and the adequacy of basic language function (6,7) and the probable increased risk of adversely affecting language ability that is related to a more extensive resection of dominant lateral temporal neocortex, alterations in language ability (e.g., postoperative dysnomia) need to be considered when examining the effects of lateral dominant temporal resection on memory outcome. To our knowledge, this has not been done. Analyses in the present investigation controlled for any surgically induced alterations in language ability and, therefore, indicate that the declines in verbal memory associated with resection of an intact left hippocampus reflect real memory impairment.
18. 19.
20. 21.
22.
23. 24.
25. 26. 27.
COMMENTS The authors present the interesting hypothesis that postoperative memory deficits after temporal lobectomy may occur in those patients who had the most normal preoperative mesial temporal lobe structures. This is a plausible hypothesis and is certainly supported by current theoretical constructs. In addition this article raises a number of other interesting questions. In this series, it appears that the degree of lateral removal was held constant. What added or differential effect do they propose would occur in response to differences in lateral resection? Second, the current concept is that the hippocampus should be removed whenever possible in patients
with temporal lobe epilepsy and that this structure is likely to be the source of the epilepsy in many cases. Would normal structural hippocampi occur in patients who have nonhippocampal temporal lobe epilepsy? Would seizure control be less adequate if this structure were left in place? Third, the authors note that all patients were left temporal lobe dominant on the intracarotid sodium amobarbital test. How did the results of memory testing during this examination correlate with pathological changes and with postoperative memory deficits? Ronald P. Lesser Baltimore, Maryland This is an interesting report that correlates the presence of hippocampal abnormality in the surgically excised temporal lobe with a neurocognitive outcome in patients undergoing anterior temporal lobectomy for intractable epilepsy. The authors clearly demonstrated that the patients with pathologically verified hippocampal neuronal cell loss experienced the least impairment in memory function postoperatively. This observation was statistically significant only for patients receiving left anterior temporal lobe surgery. Potentially, the preoperative studies assessing the presence of hippocampal abnormality may be of predictive value in assessing the neurocognitive outcome after temporal lobe surgery. Magnetic resonance imaging-based hippocampal volumetry has been demonstrated to be a reliable marker of moderate to severe mesial temporal sclerosis (1). One study has indicated that the presence of magnetic resonance imaging-identified hippocampal abnormality may correlate with memory before and after temporal lobe surgery (2). Patients with partial epilepsy of left temporal lobe origin appear to be at greater risk of experiencing a memory decline postoperatively if the preoperative volumetric study fails to reveal hippocampal formation atrophy (2). The present study by Hermann et al. provides additional evidence that preoperative assessments of hippocampal abnormality should be an intimate part of a comprehensive presurgical evaluation of patients with partial epilepsy of temporal lobe origin. Gregory D. Cascino, M.D. Rochester, Minnesota REFERENCES: (1,2) 1.
2.
Cascino GD, Jack CR, Parisi JE, Sharbrough FW, Hirschorn KA, Meyer FB, Marsh WR, O'Brien PC: Magnetic resonance imagingbased volume studies in temporal lobe epilepsy: Pathological correlations. Ann Neurol 30:31-36, 1991. Trenerry MR, Jack CR, Ivnik RJ, Sharbrough FW, Cascino GD, Hirschorn KA, Marsh WR, Kelly PJ, Meyer FB: Memory is correlated with presurgical magnetic resonance imaging: Hypocampal volumes before and after temporal lobectomy for intractable epilepsy.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
17.
epilepsy. Ann Neurol 15:64- 67, 1984. Ojemann G, Dodrill C: Verbal memory deficits after left temporal lobectomy for epilepsy: Mechanism and prediction. J Neurosurg 62:101-107, 1985. Oxbury J, Oxbury S: Neuropsychology: Memory and hippocampal pathology, in Reynolds EH, Trimble MR (eds): The Bridge between Neurology and Psychiatry. Edinburgh, Churchill Livingstone, 1989, pp 136-150. Rausch R: Effects of temporal lobe surgery on behavior. Adv Neurol 55:279-292, 1991. Rausch R, Babb TL, Engel J, Crandall PH: Memory following amobarbital injection contralateral to hippocampal damage. Arch Neurol 46:783-788, 1989. Russell EW: Renorming Russell's version of the Wechsler Memory Scale. J Clin Exp Neuropsychiatr 10:235-249, 1988. Sass KJ, Spencer DD, Kim JH, Westerveld M, Novelly RA, Lencz T: Verbal memory impairment correlates with hippocampal pryamidal cell density. Neurology 40:16941697, 1990. Saykin AJ, Gur RC, Sussman NM, O'Connor MJ, Gur RE: Memory deficits before and after temporal lobectomy: Effects of laterality and age of onset. Brain Cogn 9:191-200, 1989. Squire LR: Memory and Brain . New York, Oxford University Press, 1987. Trenerry MR, Jack CR, Ivnik RJ, Sharbrough FW, Cascino GD, Hirschorn, KA, Marsh WR, Kelly PJ, Meyer FB: Memory is correlated with presurgical magnetic resonance imaging hippocampal volumes before and after temporal lobectomy for intractable epilepsy. Epilepsia 32(Suppl 3):73, 1991 (abstr). Wechsler D, Stone CP: Wechsler Memory Scale. New York, Psychological Corporation, 1945. Wyler AR, Dohan FC, Schweitzer JB, Berry AD: A grading system for hippocampal sclerosis. (in press). Journal of Epilepsy. Wyler AR, Ojemann G, Lettich E, Ward AA, Jr: Subdural strip electrodes for localizing epileptogenic foci. J Neurosurg 60:11951200, 1984.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
16.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
Epilepsia 32(Suppl 3):73, 1991.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Table 2. Relationship of Postoperative Memory Decline with Degree of Hippocampal Sclerosis
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/4/652/2751916 by Mount Sinai Icahn School of Medicine user on 03 February 2019
Table 1. Patient Characteristics
