SURGERY FOR EPILEPSY
From: NIH Consensus Development Conference on
Surgery for Epilepsy
National Institutes of Health Consensus
Development Conference Statement
March 19-21, 1990 This statement
was originally published as: Surgery for Epilepsy . NIH
Consensus Statement 1990 Mar 19- 21; 8(2):1-20.
For making bibliographic reference
to the statement in the electronic form displayed here, it is recommended
that the following format be used: Surgery for Epilepsy.
NIH Consens Statement Online 1990 Mar 19-21[cited year month
The National Institutes of Health Consensus Development Conference
on Surgery for Epilepsy brought together neurologists, neurosurgeons,
psychologists, other health care providers, and the public to address
issues regarding epilepsy surgery including patient selection and
management, localization of seizure site onset, appropriate diagnostic
techniques, and postoperative outcome assessment.
The panel concluded that brain surgery
is an alternative treatment when medication fails. Seizure frequency,
severity type, possible brain damage or injury from frequent seizures,
and effect on quality of life all must be considered in deciding
to evaluate for surgery. An appropriate medication trial must have
been conducted, using the correct drugs for the patient's seizure
type at adequate doses and blood levels. Nonepileptic attacks must
be ruled out, and diagnostic tests to detect any underlying cause
should be performed.
If surgery is considered, patients
should be evaluated by a team including neurologists, neurosurgeons,
neuropsychologists, social workers, and, if needed, psychiatrists.
Assessment of outcome should include standardized methods of information
collection. Measures assessing quality of life and overall health
status can compare epilepsy to other chronic conditions. Assessment
of economic and social impact on the patient's family should be
The full text of the panel's statement
Epilepsy is common. About 10 percent
of all Americans will have at least one seizure at some time. Many
people have one or a few attacks and then never have another one.
For those with recurrent seizures, about 70 percent are satisfactorily
controlled with antiepileptic drugs. Of the 150,000 people who develop
epilepsy each year, 10 to 20 percent prove to have "medically
intractable epilepsy." Many of these patients and their families
have to deal with a chronic disorder that impairs the quality of
life for all concerned.
Brain surgery may be an alternative
treatment if antiepileptic drugs fail, and it is being used more
often. Several centers have reported success, and increasing numbers
of patients are being referred for surgery, including many children.
Improved technology has made it possible to identify more accurately
where seizures originate in the brain (epileptic regions), and advances
in surgery have made operative management safer. As a result, investigators
have estimated that 2,000 to 5,000 new patients in the United States
might be suitable for operations each year, compared with the present
annual rate of about 500.
Nevertheless, controlled trials have
not been done, and there is disagreement among investigators about
the choices and application of methods to evaluate patients for
surgery, which procedures should be done, and how best to assess
outcome. For these reasons, this consensus conference was organized.
There is no precise definition of
intractable epilepsy. Among the considerations are seizure frequency,
seizure type, severity of attacks, and impact on quality of life.
Before seizures are deemed intractable it is necessary to be certain
that the correct drugs have been used in the correct amounts. Complex
partial seizures are more likely to be intractable than tonic-clonic
or other common forms of epilepsy. In uncontrolled complex partial
seizures, the frequency of seizures varies from fewer than one a
week to five or more each day. The clinical manifestations also
vary in different patients. Some are not apparent to anyone but
the patient; others disrupt daily activities and are socially embarrassing.
If a patient falls during seizures that occur only a few times a
year, repeated injuries and trips to emergency rooms can make life
miserable. Even one seizure a year may disqualify a person from
having a driver's license. Disability is also influenced by the
reaction of the patient's family, friends, teachers, or employers.
All of these factors have an impact on what is judged severe enough
to warrant consideration of surgical therapy.
There are other reasons to consider
surgical therapy. For instance, repeated seizures may have adverse
effects on the brain, leading to progressive cerebral degeneration
and more severe clinical handicaps. Chronic use of antiepileptic
drugs may cause toxic syndromes and may also have adverse effects
on learning, scholastic achievement, development, and job performance.
On the other hand, surgery has risks
and costs that have to be considered. Before a patient is accepted
for surgery, it is necessary to be certain of the diagnosis and
the adequacy of previous drug therapy. Evaluation for surgery includes
several methods: video- monitoring, brain imaging, different kinds
of electroencephalography, and neuropsychological tests. The best
combination of these studies has yet to be defined. Investigators
do not agree about the preoperative need to identify areas of the
brain that control speech. Similarly, there are differences of opinion
about the specific surgical procedures that produce the best outcomes
for different kinds of attacks. Anterior temporal lobe operations
and other cortical resections involve the removal of epileptic regions
from the temporal lobe and other areas of the brain and are done
for complex partial seizures. Corpus callosotomy involves the severing
of connections between the right and left sides of the brain and
is used for some types of generalized attacks; corpus callosotomy
and hemispherectomy can be used for seizures with childhood hemiplegia.
This consensus conference was designed
to address the following questions:
- How should patients be selected?
- What evaluation is necessary to
localize epileptic regions?
- What procedures are appropriate
for specific epilepsies?
- How should outcome be assessed?
- Directions for future research--should
a controlled trial be done? If so, for what seizure types?
To address these questions, the National
Institute of Neurological Disorders and Stroke and the Office of
Medical Applications of Research of the National Institutes of Health
convened a Consensus Development Conference on Surgery for Epilepsy
on March 19-21, 1990. After a day and a half of presentations by
experts and discussion by the audience, a consensus panel drawn
from specialists and generalists from the medical profession and
related scientific disciplines, clinical investigators, and public
representatives considered the evidence and came to the following
PATIENTS BE SELECTED?
Patients with unsatisfactory seizure
control often seek alternative care. The number or severity of the
seizures may be unacceptable to the patient, family, or treating
physician. Other reasons for referral include the results of diagnostic
tests that may show a structural focal brain lesion, unsatisfactory
psychosocial adaptation due to poor seizure control, unacceptable
sedation, or other drug side effects. Such patients, especially
those with persistent complex partial seizures and some types of
generalized seizures, may be candidates for surgical treatment.
First, however, these patients should be referred to an adult or
pediatric neurologist for further evaluation of diagnosis and treatment.
Evaluation and medical treatment of
these patients may take place under the supervision of a neurologist
or in an epilepsy center that provides comprehensive diagnostic
and treatment services designed for patients with intractable epilepsy.
By "intractable" we mean that seizures have not been brought
under acceptable control with the resources available to the primary
care physician or neurologist.
To be effective and comprehensive,
the staff of a center should include the following: neurologists
with special training and experience in epilepsy; neurosurgeons
with experience in epilepsy surgery; neuropsychologists; and personnel
trained to deal with social, psychological, and psychiatric problems
and rehabilitation for school and work. Personnel to perform ancillary
neurodiagnostic assessment must also be available, including closed-circuit
TV and electroencephalography (EEG) monitoring telemetry; modern
neurophysiological and EEG equipment; neuroimaging capabilities,
including magnetic resonance imaging (MRI); and neuropsychological
testing. Some centers also have positron emission tomography (PET),
single photon emission computed tomography (SPECT), or other methods
of evaluating cerebral blood flow and metabolism.
Before a patient is considered for
surgery, evaluation should be sufficient to ensure the following:
- Nonepileptic attacks have been excluded
and epilepsy is, in fact, present. Cardiogenic syncope, psychogenic
seizures, and other nonepileptic states can closely mimic epileptic
- The epileptic seizure type and syndrome
have been clarified. Primary and secondary epilepsies, partial
seizures, and tonic- clonic seizures respond to different antiepileptic
drugs and different surgical procedures.
- Diagnostic tests have been performed
to define a metabolic or structural cause of the epileptic attacks.
- The patient has had a reasonable
trial of the appropriate antiepileptic drugs, with adequate monitoring
of compliance and the effects of the treatment.
- The patient and family have received
detailed information about the specific seizure disorder, available
drug treatments and side effects, and alternative treatments such
as surgery. If, after this evaluation, seizures prove to be intractable
or drug treatment is unsatisfactory, appropriate patients should
be referred to an epilepsy center to be evaluated for surgery.
Referrals should be made as soon as it is clear that medical treatment
is unlikely to result in further benefit. Early referrals may
prevent the development of chronic psychosocial and physical problems
that result from uncontrolled seizures.
Coexisting disorders may affect the
decision to operate; they may include severe psychiatric disorders,
profound developmental retardation, or progressive neurodegenerative
diseases. After the initial evaluation and a full unsuccessful trial
of medical therapy, surgery may be considered. Patients with partial
seizures and secondarily generalized seizures (attacks that begin
locally and spread to both sides of the brain) are potential candidates.
Secondarily generalized seizures may take the form of atonic, tonic,
or tonic-clonic attacks. Patients with seizures and childhood hemiplegia
may also benefit from surgery. Patients with the following seizure
types are potential candidates: complex partial seizures of temporal
lobe origin or other focal seizures; generalized, atonic, akinetic,
or myoclonic seizures; and partial seizures with childhood hemiplegia.
IS NECESSARY TO LOCALIZE EPILEPTIC FOCI?
Precise clinical, electrophysiologic,
and imaging data are necessary to carry out surgical therapy. Neurological
assessment is necessary to identify and exclude other forms of neurological
disease. In all cases, EEG and MRI are used. Additional tests are
often necessary for precise localization. The following electrophysiological
techniques are used in establishing the diagnosis and focality of
an epileptic disorder:
- EEG is essential, sometimes with
sleep deprivation or other activation techniques.
- EEG monitoring with video (video-EEG)
is used widely in the evaluation of potential surgical candidates
to exclude nonepileptic seizures and to define the electroclinical
characteristics of the seizures. It is often used to establish
and localize consistency and validity of the epileptogenic region.
- For precise cerebral localization,
other more or less invasive techniques are used in some cases
to establish a high degree of confidence in the electrical localization.
These methods include sphenoidal leads, subdural and epidural
electrodes and grids, and depth electrodes placed stereotaxically.
Decisions about the need for these procedures must be individualized.
Imaging techniques include x-ray computed
tomography (CT), MRI, PET, and SPECT.
CT has a limited role in the investigation
of partial epilepsy because MRI is superior to CT in demonstrating
brain tumors, vascular malformations, and focal brain atrophy. The
diagnostic value of MRI in visualizing mesial temporal sclerosis
and atrophy is under study. MRI is useful postoperatively to assess
the extent of surgical resection.
PET measures regional cerebral metabolism
and blood flow. PET imaging has been quite successful in identifying
the focus as an area of hypometabolism between attacks. This observation
may be used in selecting patients with partial and secondarily generalized
seizures for resective surgery. Because of the high costs and complexities
of PET, this technology has been confined to a limited number of
SPECT can also be used for functional
imaging of the brain because it demonstrates regional cerebral blood
flow, which is linked to cerebral metabolism and can therefore be
used to identify the epileptic focus. SPECT uses conventional and
readily available equipment and radiopharmaceuticals. These compounds
can be used to study both ictal and interictal states. In the past
decade, this relatively affordable technology has become widely
available. More work is needed to determine whether SPECT is as
sensitive as PET in localizing the epileptic regions.
Psychological tests are essential
for the evaluation of varied cerebral functions, including memory
and language. The intra-arterial amobarbital test is used to localize
language function and to assess memory preoperatively.
These diagnostic methods should be
available at specialized epileptic centers. However, the data are
insufficient to determine which particular patients require the
more invasive and detailed techniques.
Combining data from the major epilepsy
centers would allow the development of a data bank or registry that
should clarify many unanswered questions about the use of these
ARE APPROPRIATE FOR SPECIFIC EPILEPSIES?
For 60 years, there has been continuous
development of the surgical management of epilepsy. There have been
many advances in the scientific understanding of epilepsy, in new
technologies for localizing epileptic foci, and in methods for reducing
operative risk. Neither randomized controlled trials nor large community-based
clinical trials have been undertaken; nevertheless, several surgical
approaches have been reported to be successful and rational in managing
some types of intractable epilepsy. The data are not definitive
and are primarily derived from single-institution studies. Surgery
for epilepsy in patients with preoperatively detected neoplasms
or vascular malformations depends as much on the nature and site
of the lesion as on the seizure disorder.
The following discussion relates specifically
to surgical procedures performed with the primary goal of alleviating
a seizure disorder rather than removing a specific brain lesion.
Data collected to date suggest that patients in three general categories
are suitable candidates for epilepsy surgery: partial seizures of
temporal or extratemporal origin, secondarily generalized seizures,
or unilateral, multifocal epilepsy associated with infantile hemiplegia.
The largest group of surgical candidates
comprises patients with complex partial seizures of temporal lobe
origin. Preoperative evaluation identifies those patients with tumors
or vascular malformations and can determine whether the epileptic
focus is deep (in the amygdala or hippocampus) or superficial (cortical).
Appropriately directed, surgical resection
of epileptogenic tissue has resulted in success rates of 55 to 70
percent of patients, when success was defined as no seizures (some
auras may be present) for 5 years after surgery sometimes with some
patients still taking anticonvulsant medication. In some patients,
surgery also results in an improved psychosocial outcome, but this
has not been studied adequately. Combined morbidity and mortality
rates for surgery are below 5 percent. The charges for diagnostic
evaluation and surgery vary from $25,000 in uncomplicated cases
to over $100,000 in those who require extensive preoperative testing,
with a median charge of $40-60,000.
Partial seizures of frontal origin
and from other extratemporal sites may also be treated surgically
when the clinical manifestations and diagnostic studies indicate
an epileptic region in a resectable area. Appropriately directed
surgical resection of epileptogenic tissue may result in improvement
(as defined above) in 30-50 percent of cases. The mortality rates
are less than 2 percent, and the charges are slightly more than
the cost of temporal lobe resection.
With Secondarily Generalized Seizures
Some patients with generalized seizures
may be candidates for surgical management. Specifically, patients
with the Lennox-Gastaut syndrome or drop-attacks may be helped by
section of the corpus callosum, a procedure designed to prevent
rapid bilateral generalization of epileptic discharges. The procedure
is most frequently recommended in patients who are prone to violent
falls that often result in head injury. The seizure disorder usually
persists postoperatively but seizures may become less frequent,
less disabling, and less violent. Evaluation and selection of candidates
have not been standardized. There is also variability of surgical
technique, particularly how completely the corpus callosum is sectioned.
Postoperative mortality is low, but significant complications may
occur in as many as 20 percent. The charge for the surgery is often
higher than for lobectomy because the procedure as done in some
centers requires two operations, but the charge for the preoperative
evaluation is often less.
Associated With Infantile Hemiplegia
In patients with intractable unilateral,
multifocal epilepsy associated with infantile hemiplegia, hemispherectomy
or callosotomy may be beneficial. These cases are rare, accounting
for about 2 percent of all patients treated surgically for epilepsy.
Success is measured not only by improvement in seizure frequency
and type but also by improvement in behavior. Avoidance of complications
(superficial cortical hemosiderosis and hydrocephalus) is a major
consideration in the selection of surgical technique and has led
to a current consensus for a "functional" hemispherectomy
rather than one that is "anatomically complete." Success
rates of 50 to 70 percent are balanced by combined mortality and
morbidity rates that, in the past, reached 50 percent with anatomical
hemispherectomy. Initial charges are similar to those for callosotomy
but are increased by subsequent charges of dealing with late complications.
OUTCOME BE ASSESSED?
Most studies that have assessed the
outcome of either medical or surgical treatment of epilepsy have
emphasized a single measure: seizure elimination or reduction in
frequency. This measurement of seizure frequency needs to be clarified.
For example, is a 50-percent reduction in a person who has two complex
partial seizures a week equivalent to a 50-percent reduction in
a person who has two seizures a year? We recommend the use of standardized
methods to collect information about the frequency and kinds of
seizures the patients are having.
Although we recognize the importance
of seizure frequency, we recommend that future studies should use
"general measures" of outcome that would take advantage
of validated and quantitative methods to assess the quality of life
and health status of individuals. This could be achieved by a standardized
survey to assess the following: short-term surgical mortality and
morbidity or complications (e.g., death, paralysis, or infection
in the postoperative period); physical health (symptoms, functioning,
role activities, sleep-wake cycle, and mobility); mental health
(psychiatric diagnoses as well as symptoms such as anxiety or depression);
neuropsychological assessment, including cognitive functioning and
memory, both verbal and nonverbal; social health (personal interactions,
employment, sexuality, driver's license, and community interactions);
and general health (health perceptions, including fear of death
and pain, life satisfactions, and energy). Evaluation of children
should include assessment of developmental progress and school performance.
To evaluate each of the above, we
recommend using general measures of health status and quality of
life so that patients with epilepsy can be compared to patients
with other chronic conditions. Epilepsy is a unique condition and
we also recommend that, in addition to the general measures, supplemental
information should be collected about the specific aspects of the
quality of life of patients with seizures, as discussed above. This
information should be gathered from family members in addition to
the patients themselves. Because epilepsy affects the whole family,
the family's well-being should be part of the assessment.
The assessment of any treatment of
epilepsy must include analysis of the economic impact on patients,
families, and society. Economic impact includes expenditures on
medical care directly (surgical and hospitalization costs, medication
costs, costs of allied health personnel) as well as indirect costs
such as contributed care by family members and whether a patient's
income and productivity are affected positively or negatively in
Assessment must be done repeatedly
for several years. Data from treatment centers should be pooled
to achieve statistical power sufficient to test the efficacy of
treatment, as assessed by the multiple outcome measures.
We recognize that individuals will
emphasize different outcomes. Some may be more concerned with the
reduction in seizure frequency; others may be more concerned with
the effect of treatment on memory or social function. Evaluation
of any therapy for intractable epilepsy must explicitly consider
these patient preferences.
For all of these considerations, there
is a need for a standardized method of data collection so that results
from different treatment centers can be combined and compared.
FOR FUTURE RESEARCH--SHOULD A CONTROLLED TRIAL BE DONE? IF SO, FOR
WHAT SEIZURE TYPES?
- The panel is impressed that surgery
is beneficial for selected patients, but the optimal timing of
surgery is not known. Because of current referral patterns, patients
considered for temporal lobe resection tend to have had uncontrolled
(intractable) epilepsy for 10 to 20 years. We therefore recommend
a controlled trial of early versus late surgery to determine whether
early surgery or optimal medical treatment followed by later surgery
of patients with complex partial seizures will result in better
health status and quality of life and may prevent additional brain
damage or chronic social disability.
- Investigators differ in the selection
of tests for preoperative evaluation. In particular, it is not
known when more extensive diagnostic tests are needed, including
ictal surface EEG recording, invasive intracranial electrode recording,
PET, or SPECT. A program should be developed to assess the value
of these tests, and should include the development and evaluation
of algorithms. This would require standardization of definitions,
data collection, and central analysis of the data.
- We recommend development of an outcome
assessment method that combines validated measures already used
to assess general health status and function in a population of
patients with other chronic conditions, with special items that
are sensitive to the unique characteristics of people with epilepsy
and those close to them, as described in the answer to question
- We recommend that psychiatric and
behavioral functions be systematically assessed before surgery
and during followup to determine whether there are specific contraindications
to any particular surgical procedure and whether these procedures
subsequently affect behavior.
- In temporal lobe surgery for partial
seizures, standard and "tailored" resections are used
by different groups but the results are apparently similar. The
circumstances in which each technique is maximally effective should
be clarified by standardized data collection including documentation
of extent of surgical resection and multivariate analysis so that
an appropriate trial may ultimately be planned, if needed.
- Because epilepsy surgery now may
be used more often in children than in the past, we recommend
additional studies to determine the effects of uncontrolled seizures
and antiepileptic drug therapy on the developing brain. These
studies might include, but not be limited to, evaluation of sequential
neurodevelopmental assessment, anatomic and metabolic imaging
procedures, cognitive-linguistic- academic achievement in school,
and psychosocial adaptation of the child and family.
- Surgical treatment of epilepsy might
not be needed if we knew more about ways to prevent brain injury
or if we had more effective and less toxic anticonvulsant drugs.
It is therefore necessary to support fundamental research in the
basic sciences of epilepsy: developmental neurobiology, neural
science, cellular pathology, neuropharmacology, and preventive
- Most epilepsy surgery is performed
by teams of committed physicians at sophisticated medical centers.
The number of operations is increasing rapidly.
- As currently used, surgery for intractable
epilepsy is capable of stopping seizures or reducing their frequency.
Effects on overall health status and quality of life have not
been adequately studied.
- Several different diagnostic studies
and surgical techniques have been used but do not clearly differ
- Before surgery is performed, there
are three absolute requirements. First, the diagnosis of epilepsy
must be ascertained. Second, there must have been an adequate
trial of drug therapy; that is, the correct drugs used in the
correct dosage, carefully monitored for an appropriate time. Finally,
the electroclinical syndrome must be defined.
- As demand for surgery grows and
it becomes available at more hospitals, quality of care must be
maintained. Surgery should be performed at hospitals equipped
with modern technology and staffed by multidisciplinary teams
capable of preoperative diagnosis, selection of medical and surgical
treatments, comprehensive postoperative evaluation, and ambulatory
rehabilitative care. Rehabilitation should include the transition
of patients to a seizure-free or almost seizure-free lifestyle
with respect to psychological and social adjustment, education,
and vocational training.
- Physicians in any center or independent
hospital that offers surgery for epilepsy should agree to use
standardized data collection for all patients. The data should
be maintained in a central registry, with respect for confidentiality.
The data should include demographic information, diagnosis, clinical
history, results of preoperative evaluation, and outcome assessment
of quality of life and health status for at least 5 years. Outcome
information should be provided to patients considering surgery
so that expectations about the benefits and risks can be discussed
- The Panel
Members , Conference Speakers
, and Conference Organizers
National Institute of Neurological
Disorders and Stroke. Murray Goldstein, M.D. Director
Office of Medical Applications of
Research. John H. Ferguson, M.D. Director
NIH CONSENSUS DEVELOPMENT PROGRAM
NIH Consensus Development Conferences
are convened to evaluate available scientific information and resolve
safety and efficacy issues related to a biomedical technology. The
resultant NIH Consensus Statements are intended to advance understanding
of the technology or issue in question and to be useful to health
professionals and the public.
NIH Consensus Statements are prepared
by a nonadvocate, non-Federal panel of experts, based on
- presentations by investigators working
in areas relevant to the consensus question during a 2-day public
- questions and statements from conference
attendees during open discussion periods that are part of the
public session; and
- closed deliberations by the panel
during the remainder of the second day and morning of the third.
This statement is an independent report of the panel and is not
a policy statement of the NIH or the Federal Government.
Copies of this statement and bibliographies
prepared by the National Library of Medicine are available from
the Office of Medical Applications of Research, National Institutes
of Health, Federal Building, Room 618, Bethesda, MD 20892.