From the Departments of Neurology and Neurobiology and the Brain Re-search Institute, UCLA School of Medicine, Los Angeles. Address reprint re-quests to Dr. Engel at the Reed Neurological Research Ctr., 710 Westwood Plaza,Los Angeles, CA 90095-1769.
Supported in part by grants (NS-02808, NS-15654, NS-33310, and GM-24839) from the National Institutes of Health and a contract (DE-AC03-76-SF00012) with the Department of Energy.
1996, Massachusetts Medical Society.
O
F the approximately 2 million Americans with adiagnosis of epilepsy who are treated with antiep-
ileptic drugs, 20 percent continue to have seizures
1
; thisgroup of patients accounts for over 75 percent of thecost of epilepsy in the United States.
2
For many of thosewith medically refractory epilepsy, their disability canbe completely eliminated by surgical intervention. Onlya small percentage of potential surgical candidates, how-ever, are currently referred to epilepsy-surgery centers.
3
O
VERVIEW
The classic 1886 paper of Victor Horsley
4
heraldedthe modern era of epilepsy surgery, and the introduc-tion of electroencephalography (EEG) in the first halfof this century provided a practical means for localizingepileptogenic abnormalities for resection.
5,6
Neverthe-less, only a handful of epilepsy-surgery centers were cre-ated, treating relatively few patients — and only onebook on the subject appeared
7
— before 1986, when aseries of international conferences and textbooks beganto reflect an explosion of interest in the field.
8-18
By1992, over 100 epilepsy-surgery centers throughout theworld offered a wide selection of surgical procedures(Table 1) to an increasing number of patients, rangingfrom infants to senior citizens, for the treatment of dis-abling partial, and even generalized, seizures refractoryto medical therapy.
17
Modern epilepsy surgery, like heart-transplant sur-gery, requires a multidisciplinary team of highly trainedand experienced specialists working together in an ep-ilepsy center. A variety of surgical interventions are nowperformed, usually with the patient under general anes-thesia, according to the location and nature of the epi-leptogenic abnormality. The majority of procedures re-quire only a few hours in the operating room and a fewdays of postoperative hospital care. The most commonsurgery consists of removal of the amygdala and ante-rior part of the hippocampus and entorhinal cortex, aswell as a small portion of the temporal pole, leaving thelateral temporal neocortex intact. New techniques forhemispherectomy and multilobar resection involve the
partial removal and partial disconnection of affectedtissue; these and related techniques are designed to re-duce movement of the remaining portions of the brainwithin the cranial vault and to ensure resorption of cer-ebrospinal fluid. Corpuscallosotomies now usually in-volve only the anterior two thirds of the corpus callo-sum unless the patient has severe retardation. For somelocalized cortical resections, however, intraoperative test-ing may be necessary, which prolongs the operation andoccasionally requires the patient to be briefly awakenedfrom anesthesia. New techniques for treating epilep-togenic regions within primary cortical areas, such asthose controlling language and motor function, includethe removal of a discrete lesion without disturbing theadjacent cortex (lesionectomy) and multiple subpial tran-sections, which sever intracortical connections in a waythat prevents the spread of epilepsy and still preservesthe columnar structure necessary to maintain normalcortical function.
19
The current resurgence of interest in surgery for epi-lepsy can be attributed largely to technical advances invideo EEG monitoring and neuroimaging, improvementsin surgical technique, and a better understanding of theanatomical and pathophysiologic bases of the sympto-matic epilepsies. Another factor is the correction of a va-riety of misconceptions that have discouraged primarycare physicians from referring patients for surgery in thepast. Finally, a clearer delineation of the natural historyof certain catastrophic epileptic disorders of infants andyoung children and a new understanding of the plasticityof the developing brain and the damage that seizures doto it, as well as the improvements in diagnostic and sur-gical technique, have created a major new field, pediatricepilepsy surgery.
20
P
RESURGICAL
E
VALUATION
The optimal surgical intervention for epilepsy shoulddestroy just enough neuronal tissue to eliminate seizuresand no more. Therefore, the objective of presurgical eval-uation is to identify the area of brain most responsiblefor generating habitual seizures and to demonstrate thatit can be removed without causing additional unaccept-able neurologic or cognitive deficits. There is no simpletest to delineate the epileptogenic zone, defined as thevolume of brain tissue necessary and sufficient for thegeneration of seizures. The boundaries of the epilepto-genic zone can only be approximated by identifying ar-eas of the brain marked by persistent dysfunction, bothepileptic and nonepileptic. A variety of diagnostic testsare used for this purpose (Table 2), but there is no con-sensus on how much information is actually needed be-fore a particular surgical intervention can be recom-mended.
21
In most cases, presurgical evaluation involvestests that localize epileptic excitability with interictalEEG as well as long-term video EEG monitoring de-signed to capture and characterize ictal electrical activ-ity and clinical symptoms; imaging studies, usually mag-netic resonance imaging (MRI), that indicate structural
648 THE NEW ENGLAND JOURNAL OF MEDICINE March 7, 1996
abnormalities; tests for nonepileptic dysfunction, includ-ing positron-emission tomography to reveal areas of ab-normal glucose use, single-photon-emission computed to-mography to reveal areas of abnormal blood flow, andneuropsychological testing; and studies of normal corti-cal function to determine areas that must be preservedduring surgery. This last category includes cortical map-ping and intracarotid injection of amobarbital (the Wadatest) to identify the language-dominant hemisphere andthe laterality of memory function.
Diagnostic strategy is currently tailored to the specif-ic surgical intervention to be used.
21
For standardizedtemporal-lobe resections, the presurgical evaluation needonly determine that habitual seizures are originatingwithin the boundaries of the intended excision and thatthe structures of the contralateral mesial temporal lobecan support memory. For specific neocortical resectionsand multiple subpial cortical transections, more detailedinvestigation is required to identify the boundaries ofthe epileptogenic zone, as well as of adjacent areas ofessential primary cortex. For hemispherectomies andlarge multilobar resections, the goal of the presurgicalevaluation is to determine the extent of the functionaland structural disturbance of the involved hemisphereand whether the contralateral hemisphere is reasonablyintact. If section of the corpus callosum is contemplat-ed, there must be a documented history of disabling dropattacks as the principal type of seizure; it is also impor-tant to determine that the patients are not candidatesfor a more definitive resection. Before lesionectomy isperformed, it is necessary only to demonstrate that sei-zures are originating at the site of the structural lesionand that the lesion is in an essential cortical area thatcannot be resected.
Long-term video EEG monitoring is generally per-
formed on an inpatient basis at most centers before anysurgical treatment for epilepsy, in order to verify thatthe events are epileptic, characterize the seizure semi-ology, and if possible, identify electrographically the siteof ictal onset.
22
Several days of continuous monitoringis often required in order to record a sufficient numberof seizures, thus making this the most expensive partof presurgical evaluation. Because noninvasive monitor-ing with scalp and sphenoidal electrodes can providefalse information about the origin of the seizures, in thepast this procedure was often repeated with stereotacti-cally implanted depth electrodes, or subdural electrodes,placed so as to record information directly from the pre-sumed epileptogenic region.
21
The use of structural andfunctional neuroimaging to confirm location by nonin-vasive ictal EEG has now eliminated the need for in-tracranial recording in all but a few patients. Interictalpositron-emission tomography was the first method offunctional neuroimaging found to be of value for this pur-pose,
23
but interictal — and particularly ictal — single-photon-emission computed tomography
24
also providesimportant information that permits safe and effectivesurgery without invasive monitoring.
Identifying the epileptogenic region is most difficultin patients with so-called cryptogenic partial epilepsy,in which no structural lesion has been found preopera-tively. Marked improvements in high-resolution MRI,however, now permit visualization of hippocampal at-rophy in most patients who previously were given a di-agnosis of cryptogenic temporal-lobe epilepsy
25
and of
*Data were obtained from Engel et al.
3
Dashes indicate that no data are available.
†Results reported for 39 epilepsy-surgery centers participating in the first InternationalPalm Desert Conference in 1986.
‡Results reported for 107 epilepsy surgery centers participating in the second InternationalPalm Desert Conference in 1992.
Table 1. Surgical Procedures Commonly Performed to Treat Epilepsy.
focal, dysplastic, cortical lesions in many patients pre-viously given a diagnosis of cryptogenic neocortical epi-lepsy.
26
Positron-emission tomography
27
and MRI
28
havealso helped to identify localized, resectable, cortical ab-normalities in infants and young children with crypto-genic forms of catastrophic secondary generalized epi-lepsy, such as infantile spasms, who otherwise wouldnot have been considered for surgery. Presurgical eval-uation may become cheaper because of new techniquesfor outpatient ictal EEG recording with digital home-monitoring systems,
29
advances in the identification ofictal, as well as interictal, spike sources by magnetoen-cephalography,
30
and the more widespread use of mag-netic resonance spectroscopy and functional MRI.
25
E
ARLY
I
NTERVENTION
Timely identification of potential candidates for sur-gery has suffered from the imprecise definition of med-
ically refractory epilepsy. In practice, most patientsreferred to epilepsy-surgery centers still have several sei-zures a month — and sometimes several a day — despitetreatment with standard antiepileptic drugs, alone and incombination, at adequate doses. However, patients withdisabling but infrequent seizures can also benefit greatlyfrom surgery, as can those for whom a doctor’s insist-ence on yet another drug regimen only delays a defini-tive surgical procedure and creates a risk that irrevers-ible psychosocial consequences of prolonged illness willdevelop. Surgical intervention need not be consideredonly as a last resort. There are surgically remediable syn-dromes (Fig. 1) — conditions with a known pathophys-iology and natural history that have a poor prognosiswith purely medical treatment, but that respond well tosurgical treatment.
3
Because patients with these condi-tions can be readily identified by noninvasive studies,and because these disorders can have progressive fea-
Figure 1. Surgically Remediable Epileptic SyndromesImportant diagnostic features of three surgically remediable syndromes are shown. Mesial temporal-lobe epilepsy (top panel) is char-acterized on EEG by focal interictal sphenoidal spikes (small arrows) and ictal onset (large arrow). Unilateral hippocampal atrophy(arrow) is apparent on a T
In the middle panel, EEG in a 23-year-old woman with a low-grade glioma (a discrete neocortical lesion) in the right inferior temporal–occipital junction showed focal interictal spikes (small arrow) and ictal onset (large arrow) in the right posterior temporal area. Thestructural lesion is evident on a T
2
-weighted coronal MRI through the temporal–occipital junction (arrow). Functional MRI of the re-
FunctionalImaging
MesialTemporal-Lobe
Epilepsy
DiscreteNeocortical
Lesion
DiffuseHemisphericDisturbance
Adolescentor
young adult
Any age
Infant orvery young
child
Age Groupat Surgery
EEGLeads
Electroencephalogram StructuralImaging
MRI
MRI
MRI PET
MRI
PET
sponse to visual stimulation indicated that the lesion did not encroach on primary visual cortex (an axial image is shown).In the bottom panel, EEG in a two-year-old child with catastrophic secondarily generalized seizures and unilateral seizures due to aleft-sided hemimegaloencephaly (diffuse hemispheric disturbance) revealed attenuation over the left hemisphere (lower channels),widespread interictal spikes most prominent on the left side (small arrows), and a variety of ictal discharges emanating from the leftside (a tonic seizure is shown, beginning at the large arrow). The patient also had a markedly dysplastic left hemisphere on MRI (theright side of the T
1
-weighted axial image) and, on functional PET, profound hypometabolism of glucose in the left hemisphere (rightside of axial image), but an apparently normal pattern of glucose metabolism in the right hemisphere (left side of image). Illustration
by Lynne Olson from material provided by Drs. John Curran, John Mazziotta, Michael Phelps, Raman Sankar, and Arthur Toga.
650 THE NEW ENGLAND JOURNAL OF MEDICINE March 7, 1996
tures, referral to an epilepsy-surgery center ought to beconsidered as soon as first-line antiepileptic medicationsfail to be effective. For most conditions, this means high-dose carbamazepine and phenytoin. Continued attemptsto treat patients with second-line drugs, or combinationsof drugs, may not be in the best interest of those withsurgically remediable syndromes.
The prototype of a surgically remediable syndrome ismesial temporal-lobe epilepsy,
31
which has a character-istic presentation and a specific pathophysiologic basis:hippocampal sclerosis (Table 3). This disorder is possi-bly the most common form of epilepsy and one of themost refractory to medical treatment. Seizures usuallybegin in the first decade of life and characteristically be-come intractable as early as adolescence. The risk of ir-reversible psychosocial consequences for patients withintractable seizures is great. Most patients with this con-dition, however, can be easily identified as likely candi-dates for surgery by anterior temporal interictal spikeson EEG, hippocampal atrophy on high-resolution MRI,and temporal-lobe hypometabolism noted on interictalpositron-emission tomography. Ictal EEG, neuropsycho-logical tests, and if necessary, ictal single-photon-emis-sion computed tomography can confirm the diagnosis,and anterior mesial temporal-lobe resection offers a 70to 80 percent chance of cure.
Patients with medically refractory partial seizuresthat are due to discrete structural lesions, such as glial tu-mors or congenital malformations, also have a surgicallyremediable syndrome. Caution should be exercised, how-ever, because some structural lesions are not clinicallyimportant and others are part of a multifocal process inwhich another lesion that cannot be visualized is actu-ally responsible for the epileptic condition.
33
Consequent-ly, surgical treatment should not be undertaken on thebasis of structural imaging alone; confirmation of epi-leptogenicity is necessary, and this usually requires ictalEEG. Surgical outcomes in properly evaluated patientswith discrete epileptogenic lesions are equivalent to thosein patients treated for mesial temporal-lobe epilepsy.
Catastrophic seizures, either generalized or unilater-al, in infants and young children can result from a num-ber of brain disturbances that are confined to one, or partof one, hemisphere; these include hemimegaloenceph-aly and other diffuse cortical dysplasias, Sturge–Webersyndrome, large porencephalic cysts, and the usuallyunilateral inflammatory process of Rasmussen’s enceph-alitis.
34
Medically refractory seizures in these conditionsoften occur many times a day, are associated with pro-found developmental delay, and can be life-threatening.The pathologic region is easily identified with MRI or,in some cases, positron-emission tomography. Ictal EEGcan demonstrate that the epileptogenic abnormalitiesare restricted to the structurally abnormal hemisphere,and both EEG and positron-emission tomography canbe useful in confirming that the contralateral hemi-sphere is functionally intact. In these situations, hemi-spherectomy or a large multilobar resection can end ha-bitual seizures and reverse the inevitable developmentaldelay.
34
Because removal of the perirolandic area is usu-ally considered only for patients who already have hem-
iparesis with a useless hand, this surgical procedure in-troduces no new motor deficit; in fact, function of theaffected limbs often improves. Without surgery suchchildren might be condemned to life in an institution, butwith appropriate surgical intervention they have a 60 to80 percent chance of living a nearly normal life.
Patients with secondary generalized epilepsy, such asthe Lennox–Gastaut syndrome, have diffuse brain dam-age and often have disabling drop attacks that cause fre-quent severe injury. Antiepileptic drugs are usually inef-fective against such seizures, and patients must thereforewear protective helmets and greatly limit their activities.If drop attacks are the most disabling type of seizureexperienced by a patient with secondary generalized ep-ilepsy, corpuscallosotomy should be considered.
35
Cor-puscallosotomy can completely end drop attacks for alarge proportion of patients, but it is a palliative, not acurative, procedure; it is not likely to affect other typesof seizures or to alter the mental retardation or otherneurologic abnormalities usually associated with second-ary generalized epilepsy. Nevertheless, the tremendouspositive effect of this surgical intervention on the qual-ity of life of patients with disabling drop attacks justifiesregarding this condition as surgically remediable.
Advances in diagnostic technology and surgical pro-cedures will undoubtedly result in the identification ofmore surgically remediable syndromes in the future. For
*Adapted from Engel.
32
Table 3. The Syndrome of Mesial Temporal-Lobe Epilepsy.
*
History
Higher incidence of complicated febrile convulsions than in other types of epi-lepsy.
Family history of epilepsy common.Onset in latter half of first decade of life.Auras that often occur in isolation common.Infrequent secondarily generalized seizures.Seizures that often remit for several years until adolescence or early adulthood.Seizures that often become medically intractable.Interictal behavioral disturbances can develop, most commonly depression.
Clinical features of seizures
An aura is usually present. The most common is epigastric, often with other au-tonomic or psychic symptoms, including emotion (e.g., fear). Olfactory or gus-tatory sensations can occur. Auras usually last several seconds.
Complex partial seizures often begin with arrest and stare; oroalimentary autom-atisms and complex automatisms are common. Posturing of one arm may oc-cur contralateral to the ictal discharge. The seizure usually lasts one to twominutes.
The postictal phase usually includes disorientation, recent-memory deficit, am-nesia for the event, and dysphasia if seizures begin in the language-dominanthemisphere. This phase lasts several minutes.
Neurologic and laboratory features
Neurologic examination usually normal except for memory deficit.Unilateral or bilateral independent anterior temporal EEG spikes with maximal
amplitude in basal electrodes.Extracranial ictal EEG activity only with symptoms of complex partial seizure;
usually initial or delayed focal rhythmic onset pattern of 5 to 7 per second,maximal amplitude in one basal temporal derivation.
Usually temporal-lobe hypometabolism on interictal positron-emission tomogra-phy with fluorodeoxyglucose, often involving ipsilateral portion of the thala-mus and basal ganglia.
Usually temporal-lobe hypoperfusion on interictal single-photon-emission com-puted tomography and characteristic pattern of hyperperfusion and hypoper-fusion on ictal single-photon-emission computed tomography.
Usually memory dysfunction specific to the involved temporal lobe on neuropsy-chological testing and amnesia with contralateral intracarotid injection ofamobarbital.
instance, studies are under way to determine whetherthe progressive verbal agnosia that develops in childrenwith the Landau–Kleffner syndrome,
36
presumably dueto cryptogenic epileptic activity involving language cor-tex, can be reversed by multiple subpial transection thateliminates epileptogenic activity in that area without in-troducing additional language disturbances.
Patients with medically refractory seizures who clearlydo not have one of the surgically remediable syndromesmentioned here should be given more aggressive ther-apy with antiepileptic drugs, alone or in combination.They should not, however, be discounted as possiblecandidates for surgery, and at some point referral to anepilepsy-surgery center is appropriate. Although thesepatients often require prolonged and expensive invasivemonitoring with depth electrodes or subdural electrodes,and although fewer than 50 percent of patients who donot have a surgically remediable syndrome become sei-zure-free postoperatively, most obtain some benefit. It isimportant to make sure that patients do not have a sur-gically remediable syndrome before delaying referral forsurgery and proceeding with numerous, prolonged ma-nipulations of medical therapy.
S
URGICAL
O
UTCOME
Table 4 shows data on the results of surgical treat-ment for epileptic seizures, during the period 1986 to1990, as obtained from an international survey of 100epilepsy-surgery centers.
37
These data do not fully re-flect the success of current surgical techniques for tworeasons: some reporting centers were in developing coun-tries that did not have access to the most modern ap-proaches, and results in general have improved consid-erably in the past five years. Although no comparableworldwide data on outcomes are available for surgicalprocedures performed since 1990, the results reported inthe more recent literature from individual centers,
38
aswell as data presented at professional meetings, indicatesteady progress. In attempting to gauge the cost effec-tiveness of surgery, however, the translation of the suc-cessful elimination of seizures into psychosocial reha-bilitation, elimination of disability, and improved qualityof life becomes important.
39,40
Although patients who nolonger have seizures represent an important savings indirect costs for medical care, they may still remain de-pendent on family and the social-welfare system andhave many indirect costs associated with their disabili-ty.
41
Patients are most likely to be able to work and tolive relatively normal, productive lives if surgical inter-vention takes place early in the course of their epilepticdisorders.
Operative complications of surgery for epileptic sei-zures are rare
42
and account for minimal disability. Inlocalized resective surgery, less than 5 percent of pa-tients have some postoperative neurologic deficit dueto unintended vascular compromise or other accidentaldamage to essential neural tissue; the great majority ofthese disturbances are transient and resolve within aperiod of months. Mesial temporal-lobe resections areoften associated with defects in the contralateral supe-rior quadrant of the visual field that are identifiable by
formal testing but almost never noticed by patientsthemselves. Because memory function specific to theinvolved temporal lobe is usually depressed preopera-tively, hippocampal resection is unlikely to introduce anew deficit, and in fact, often results in an improvementin memory function specific to the contralateral tem-poral lobe. However, anteromesial temporal lobectomyin the dominant hemisphere of patients with normalmemory will produce a deficit in verbal memory thatcould pose a problem for those who need to functionat a high intellectual level. Functional mapping tech-niques, including the intracarotid amobarbital proce-dure, can be used to predict when surgical interventionis likely to cause further language, memory, or otherneurologic disturbances and can enable surgical strate-gy to be altered in order to avoid unacceptable conse-quences. In some circumstances, however, new neurolog-ic deficits are unavoidable and must be accepted bypatient and physician as a tolerable trade-off beforeany surgery is undertaken.
Microsurgical techniques and other improvements insurgical methods have not only increased the safetyand efficacy of routine surgical procedures for epilepticseizures, but also made even hemispherectomy and cor-puscallosotomy more attractive alternatives. Modifica-tions of hemispherectomy have almost eliminated thedevastating delayed complications previously associat-ed with the procedure.
43
The ability to sever the corpuscallosum without entering the third ventricle has great-ly improved the early postoperative course of patientstreated with that technique; section of only the anteriortwo thirds of the corpus callosum can avert the some-times disturbing symptoms that arise from disconnect-ing the two hemispheres.
44
F
UTURE
D
IRECTIONS
New antiepileptic drugs will undoubtedly benefit somepatients who now have medically refractory epilepsy,
*Results reported for 100 epilepsy-surgery centers at the second International Palm DesertConference in 1992. Data were obtained from Engel et al.
37
†The patients had to be free of disabling seizures for at least two years. Some patients maystill have occasional auras. Patients take antiepileptic medication for at least two years post-operatively and may elect to continue thereafter.
‡This was defined as more than a 90 percent reduction in the frequency of seizures. Thiscategory includes patients who may have had only one or two seizures since surgery.
Table 4. Results of Surgical Treatment for Epilepsy, Worldwide,1986–1990.
652 THE NEW ENGLAND JOURNAL OF MEDICINE March 7, 1996
but pharmacologic advances are unlikely to decreasethe large number of potential candidates for surgery inthe near future. Stimulation of the vagus nerve
45
and thethalamus
46
may reduce the frequency and severity ofsome forms of epileptic seizures, but these techniquesremain experimental and the indications for their useare uncertain. The accurate identification of surgicallyremediable syndromes, the application of advanced di-agnostic tools that eliminate the need for invasive mon-itoring, and the potential for early intervention alreadymake safe and effective surgical treatment possible fora great many patients who now suffer from disablingepileptic seizures. Surgery for epilepsy, however, will beconsidered for only a relatively small proportion of thepatients who could benefit from such treatment, unlessit gains more widespread acceptance. We have a moralobligation to make this potentially curative therapy avail-able to people disabled by epilepsy. To take a globalperspective, 90 percent of the world’s population livesin the developing countries; they bear the brunt of theoverwhelming burden of epilepsy.
47
Supporting and dis-seminating advances that make epilepsy surgery morecost effective not only will eventually help tens of thou-sands of patients in the United States, but also will aidmillions of people in developing countries who need notsuffer the consequences of medically refractory epilep-tic seizures.
2. Begley CE, Annegers JF, Lairson DR, Reynolds TF, Hauser WA. Cost of ep-ilepsy in the United States: a model based on incidence and prognosis. Ep-ilepsia 1994;35:1230-43.
3. Engel J Jr, Shewmon DA. Who should be considered a surgical candidate?In: Engel J Jr, ed. Surgical treatment of the epilepsies. 2nd ed. New York:Raven Press, 1993:23-34.
4. Horsley V. Brain-surgery. BMJ 1886;2:670-5.5. Bailey P, Gibbs FA. The surgical treatment of psychomotor epilepsy. JAMA
1951;145:365-70.6. Jasper H, Pertuisset B, Flanigin H. EEG and cortical electrograms in pa-
tients with temporal lobe seizures. Arch Neurol Psychiatr 1951;65:272-90.
7. Purpura DP, Penry JK, Walter RD, eds. Neurosurgical management of theepilepsies. Vol. 8 of Advances in neurology. New York: Raven Press, 1975.
8. Engel J Jr, ed. Surgical treatment of the epilepsies. New York: Raven Press,1987.
10. Dam M, Andersen AR, á Rogvi-Hansen B, Jennum P, eds. Epilepsy surgery:non-invasive versus invasive focus localization. Acta Neurol Scand Suppl1994;89(152):1-218.
11. Duchowny M, Resnick R, Alvarez L, eds. Pediatric epilepsy surgery. J Ep-ilepsy 1990;3:Suppl 1.
12. Apuzzo MLJ, ed. Neurosurgical aspects of epilepsy. Park Ridge, Ill.: Amer-ican Association of Neurological Surgeons, 1991.
13. Spencer SS, Spencer DD, eds. Surgery for epilepsy. Boston: Blackwell Sci-entific, 1991.
14. Theodore W, ed. Surgical treatment of epilepsy. Epilepsy Res 1992;Suppl 5.15. Lüders HO, ed. Epilepsy surgery. New York: Raven Press, 1992.16. Silbergeld DL, Ojemann GA, eds. Epilepsy surgery. Neurosurg Clin North
Am 1993;4.17. Engel J Jr, ed. Surgical treatment of the epilepsies. 2nd ed. New York: Raven
Press, 1993.18. Wyler AR, Hermann BP, eds. The surgical management of epilepsy. Boston:
Butterworth–Heinemann, 1994.
19. Morrell F, Whisler WW, Bleck TP. Multiple subpial transection: a new ap-proach to the surgical treatment of focal epilepsy. J Neurosurg 1989;70:231-9.
20. Tuxhorn I, Holthausen H, Boenigk HE, eds. Paediatric epilepsy syndromesand other surgical treatment. London: John Libbey (in press).
21. Lüders HO, Engel J Jr, Munari C. General principles. In: Engel J Jr, ed. Sur-gical treatment of the epilepsies. 2nd ed. New York: Raven Press, 1993:137-53.
22. Engel J Jr, Burchfiel J, Ebersole J, et al. Long-term monitoring for epilepsy:report of an IFCN committee. Electroencephalogr Clin Neurophysiol 1993;87:437-58.
23. Engel J Jr, Henry TR, Risinger MW, et al. Presurgical evaluation for partialepilepsy: relative contributions of chronic depth-electrode recordings ver-sus FDG-PET and scalp-sphenoidal ictal EEG. Neurology 1990;40:1670-7.
24. Newton MR, Berkovic SF, Austin MC, Rowe CC, McKay WJ, Bladin PF.SPECT in the localisation of extratemporal and temporal seizure foci.J Neurol Neurosurg Psychiatry 1995;59:26-30.
25. Kuzniecky RI, Jackson GD. Magnetic resonance in epilepsy. New York:Raven Press, 1995.
26. Barkovich AJ, Rowley HA, Andermann F. MR in partial epilepsy: value ofhigh-resolution volumetric techniques. AJNR Am J Neuroradiol 1995;16:339-43.
27. Chugani HT, Shewmon DA, Shields WD, et al. Surgery for intractable in-fantile spasms: neuroimaging perspectives. Epilepsia 1993;34:764-71.
28. Sankar R, Curran JG, Kevill JW, Rintahaka PJ, Shewmon DA, Vinters HV.Microscopic cortical dysplasia in infantile spasms: evolution of white matterabnormalities. AJNR Am J Neuroradiol 1995;16:1265-72.
29. Morris GL III, Galezowska J, Leroy R, North R. The results of computer-assisted ambulatory 16-channel EEG. Electroencephalogr Clin Neurophys-iol 1994;91:229-31.
30. Ebersole JS, Squires KC, Eliashiv SD, Smith JR. Applications of magneticsource imaging in evaluation of candidates for epilepsy surgery. Neuroim-aging Clin North Am 1995;5:267-88.
31. Wieser H-G, Engel J JR, Williamson PD, Babb TL, Gloor P. Surgically re-mediable temporal lobe syndromes. In: Engel J Jr, ed. Surgical treatment ofthe epilepsies. 2nd ed. New York: Raven Press, 1993:49-63.
32. Engel J Jr. Update on surgical treatment of the epilepsies: summary of theSecond International Palm Desert Conference on the Surgical Treatment ofthe Epilepsies. Neurology 1993;43:1612-7.
33. Spencer SS. The relative contributions of MRI, SPECT, and PET imagingin epilepsy. Epilepsia 1994;35:Suppl 6:S72-S89.
34. Shields WD, Duchowny MS, Holmes GL. Surgically remediable syndromesof infancy and early childhood. In: Engel J Jr, ed. Surgical treatment of theepilepsies. 2nd ed. New York: Raven Press, 1993:35-48.
35. Spencer SS, Spencer DD, Sass K, Westerveld M, Katz A, Mattson R. An-terior, total, and two-stage corpus callosum section: differential and incre-mental seizure responses. Epilepsia 1993;34:561-7.
36. Deonna TW. Acquired epileptiform aphasia in children (Landau-Kleffnersyndrome). J Clin Neurophysiol 1991;8:288-98.
37. Engel J Jr, Van Ness PC, Rasmussen TB, Ojemann LM. Outcome with re-spect to epileptic seizures. In: Engel J Jr, ed. Surgical treatment of the epi-lepsies. 2nd ed. New York: Raven Press, 1993:609-21.
38. Engel J JR. Epilepsy surgery. Curr Opin Neurol 1994;7:140-7.39. Devinsky O, Cramer JA, eds. Assessing quality of life in epilepsy: develop-
ment of a new inventory. Epilepsia 1993;34:Suppl 4:S1-S44.40. Vickrey BG, Hays RD, Rausch R, Sutherling WW, Engel J Jr, Brook RH.
Quality of life of epilepsy surgery patients as compared with outpatientswith hypertension, diabetes, heart disease, and/or depressive symptoms.Epilepsia 1994;35:597-607.
41. Vickrey BG, Hays RD, Rausch R, et al. Epilepsy surgery outcomes: sei-zures, medication use, employment, and self-reported quality of life. Lancet1995;346:1445-9.
42. Pilcher WH, Rusyniak WG. Complications of epilepsy surgery. NeurosurgClin North Am 1993;4:311-25.
43. Villemure J-G, Adams CBT, Hoffman HJ, Peacock WJ. Hemispherectomy.In: Engel J Jr, ed. Surgical treatment of the epilepsies. 2nd ed. New York:Raven Press, 1993:511-8.
44. Roberts DW, Rayport M, Maxwell RE, Olivier A, Marino R Jr. Corpus cal-losotomy. In: Engel J Jr, ed. Surgical treatment of the epilepsies. 2nd ed.New York: Raven Press, 1993:519-26.
45. The Vagus Nerve Stimulation Study Group. A randomized controlled trialof chronic vagus nerve stimulation for treatment of medically intractableseizures. Neurology 1995;45:224-30.
46. Velasco F, Velasco M, Velasco A, Jimenez F, Marquez I, Rise M. Electricalstimulation of the centromedian thalamic nucleus in control of seizures:long-term studies. Epilepsia 1995;36:63-71.
47. Kale R. Health information for the developing world. BMJ 1994;309:939-42.
