Indications and Criteria for Surgical Intervention

Michael S. Duchowny

A. Simon Harvey

Michael R. Sperling

Peter D. Williamson

Introduction

The estimated lifetime cumulative incidence and prevalence of epilepsy are 3% and 0.5% respectively, with approximately 60% of patients manifesting partial seizures.⁴⁷^,⁴⁸ Although most patients with new-onset epilepsy have few seizures or are well-controlled, an estimated 5% to 10% of patients become sufficiently medically intractable to consider surgical therapy.⁴⁷^,⁸⁶ Data from a Danish registry of unselected patients with partial epilepsy suggest a cumulative incidence for intractable epilepsy of 135 per 100,000.⁶² Intractable epilepsy is a major risk for personal injury, quality of life, and in some cases, mortality. Furthermore, recurrent seizures have a significant socioeconomic cost to the individual and society.

In recent decades, there has been increasing interest in the surgical treatment of patients with intractable seizures. At the second Palm Desert conference on epilepsy surgery, the number of procedures performed at major epilepsy surgery centers between 1986 and 1990 was more than twice that reported before 1985.³⁴ This remarkable increase has been fuelled by a combination of technological advances in neuroimaging and electroencephalography, new appreciation of the devastating personal and socioeconomic effects of recurrent seizures, delineation of surgically remediable epilepsy syndromes, and recognition of the long-term benefits of early surgical intervention.

Temporal lobe resections make up approximately two-thirds of all procedures currently performed at epilepsy surgery centers, reflecting the predominance of adult patients with intractable temporal lobe epilepsy (TLE). In pediatric epilepsy surgery centers, a greater proportion of patients undergo extratemporal, multilobar, and hemispheric resections.⁵⁴^,⁸⁵^,¹²⁶

It is currently estimated that as many as 5,000 new patients annually in the United States might benefit from epilepsy surgery,⁸⁶ yet less than one-third receive treatment at the major centers.³⁴ Underutilization of epilepsy surgery is also apparent in other countries.⁷⁹^,⁹⁷ The education of primary care physicians to refer patients with surgically remediable epilepsies, instruction of health insurers and government health agencies regarding reimbursement for epilepsy surgery, and refinement of epilepsy surgical protocols should increase the number of patients with intractable epilepsy who may benefit from epilepsy surgery.

Who Is a Candidate for Epilepsy Surgery?

A simple statement can be made that epilepsy surgery may be considered by anyone whose seizures recur despite use of appropriate antiepileptic drugs (AEDs). For properly selected patients, surgery offers a relatively safe and effective means of either abolishing seizures, diminishing their severity, or reducing seizure frequency. Because most of the benefits conferred by surgery accrue when seizures are completely stopped, palliative surgery is a viable option only if it would improve quality of life or lessen the risk of injury or death. Surgery may seem a rather drastic measure for intermittent symptoms that may not prevent someone from living a happy and fulfilling life. However, uncontrolled epilepsy has numerous adverse medical and psychosocial consequences, some of which appear only after years or decades of illness. Surgery is performed precisely because it is effective, and its risks are generally less than the aggregate long-term risks of uncontrolled seizures.

What kinds of seizures and types of epilepsy provoke the consideration of epilepsy surgery? The most common types of seizures that lead people to consider surgery are those that cause alteration of consciousness or awareness or injury. These seizures have the greatest potential to cause injury and impair quality of life. Hence, complex partial and secondarily generalized tonic–clonic seizures are common indications for surgery because of their adverse psychosocial and medical repercussions. Loss of awareness prevents legal operation of a motor vehicle, forces people to rely upon others for transportation (which limits independence), reduces employment opportunities, diminishes educational choices, and imposes psychological burdens. Common activities, such as using pubic transportation, are fraught in ways that people who do not suffer from seizures cannot imagine. In addition, when seizures might cause someone to precipitously fall with resultant injury (e.g., tonic–clonic, atonic, tonic, and some simple partial seizures), activity is further limited. In addition, some seizures, although they might not cause injury or loss of awareness, might be so unpleasant or psychologically disruptive that surgery could still be an option. For example, seizures that either produce intense nausea and vomiting or cause socially unacceptable behavior might warrant surgical therapy. Lastly, some postictal behaviors, such as recurrent postictal psychosis and prolonged postictal confusion or lethargy, may cause more difficulty than the seizures themselves and prompt a patient to consider surgery. In general, the individual who suffers from the seizures is the only one who can state with certainty whether residual symptoms are inadequately controlled or not, although it is the physician’s responsibility to assess medical risks.

How often should seizures occur to consider surgery? There is no objectively determined seizure frequency required to consider surgery. Although most people who have surgery have seizures at least once every month or so, others have had fewer seizures and decided that surgery was worthwhile. Having just one or two seizures per year prohibits the legal operation of a motor vehicle and can lead to serious medical and psychosocial consequences. In our experience, some patients with as few as two or three seizures per year have had surgery and viewed it as worthwhile; careful preoperative counseling
should always be performed so that patients will have realistic expectations.

The time at which seizures occur might also influence the decision to have surgery. Seizures that occur at predict-able times pose fewer problems than seizures that appear randomly. For example, the person who has seizures occurring only while asleep (nocturnal seizures) may drive an automobile and live a relatively unrestricted life. However, these seizures might still be psychologically disturbing, disrupt sleep, and pose medical risk so that surgical consideration is legitimate.

Last, the type of epilepsy syndrome must influence the decision to consider surgery. Some types of epilepsy are known to disappear with the passage of time, so that it would be overly aggressively to perform surgery for them. Other types of epilepsy are associated with a progressive downhill course, (e.g., progressive myoclonic epilepsy, seizure associated with malignant tumors) so that surgery will certainly not afford long-term benefit and probably should not be done. Other types of drug-resistant epilepsy (e.g., medial TLE) often favorably respond to surgical treatment, and surgery might be considered early in the course of treatment.

Rationale for Surgical Intervention

In addition to the physical harm posed by epileptic seizures, accumulating evidence suggests that chronic epilepsy results in deterioration of brain function and structure. It is recognized that many patients with epilepsy, particularly patients with seizures of temporal lobe origin, suffer a progressive decline in learning and memory.⁵¹ The loss of verbal memory skills has particularly devastating consequences on quality of life and may lead to a significant decline in socioeconomic status. It is not altogether clear to what degree seizures, their underlying pathologic substrate, or the cumulative effects of drug treatment contribute to decline, but a longer duration of epilepsy appears to be a critical factor in cognitive decline. Helmstaedter et al.⁵⁰ performed a longitudinal neurocognitive study of 147 surgically treated and 102 medically treated adults with TLE. They found that a higher proportion of the medically treated group suffered a significant decline in memory skills between 2 and 10 years after baseline evaluation. Seizure-free patients had the best outcome, with superior recovery of both memory and nonmemory functions. Cross-sectional studies have further shown that a duration of TLE greater than two decades is associated with the greatest deterioration of cognitive ability and is independent of patient age and age at seizure onset.⁶¹ Additional evidence suggests that frequent interictal discharges and short nonconvulsive seizures can cumulatively impair cognition and educational achievement.²

In parallel with studies of cognition, serial neuroimaging studies in patients with refractory seizures reveal evidence of progressive degeneration at anatomic loci remote from seizure origin. Recurrent temporal lobe seizures are associated with decreased hippocampal volume 3.5 years after seizure onset if recurrent seizures are present,⁴¹ whereas progressive volume reduction occurs in the amygdala and entorhinal cortex ipsilateral to the atrophic hippocampus.¹⁴ Voxel-based morphometry of unilateral TLE reveals abnormalities in the thalamus, cerebellum, and extratemporal neocortex as well as temporal and extratemporal white matter.⁸⁰ These findings collectively reveal that chronic TLE is associated with widespread secondary changes that affect multiple brain regions and the connectivity between cerebral hemispheres.

Criteria for Surgical Candidacy

Medical Intractability

The decision to refer patients for epilepsy surgery rests on the perceptions of the patient and physician that medical treatment is failing. These perceptions are based on the belief that the physician has exercised “due diligence”—she has administered optimal medications at high serum therapeutic concentrations. However, there are few established guidelines to assist the physician’s choice of medications, dosing interval, duration of treatment, monitoring of serum concentration, and withdrawal of therapy. Thus, medical treatment of seizures is still largely individualized, with no unique point in time when intractability is established.

Certain risk factors dictate a low probability of seizure remission. The likelihood of persistent seizures is increased by additional factors such that patients with multiple risk factors have the poorest outcome.²¹ High seizure frequency in the form of daily or weekly episodes constitutes a major risk factor for medical intractability, whereas seizure clustering increases the risk still further.¹ Early seizure onset, particularly in infancy, predicts seizure persistence,⁷⁷ and infantile hemiconvulsive status epilepticus (SE) is specifically linked to the development of TLE.²⁰^,⁴²^,⁴⁵ Motor convulsions in a nonconvulsive disorder are a risk factor independent of the number of lifetime episodes.⁹⁰ Patients with organic brain damage are less likely to undergo spontaneous seizure remission.¹^,⁵⁶ Thus, abnormal neurologic status by physical examination or neuroimaging criteria is associated with both a greater risk of developing epilepsy and a reduced likelihood of remission. As a rule, the more severe the brain damage, the greater the likelihood of seizure persistence.¹¹³

Deficient medical management may, in some cases, simulate medical intractability. Patient noncompliance or intermittent compliance may masquerade as “pseudointractable” seizures. Physician failure to administer a first-line or appropriate adjunctive antiepileptic medication, or more commonly failure to attain high therapeutic serum concentrations, is not infrequently discovered in patients referred for surgical evaluation.⁴³ Nonepileptic disorders and psychogenic seizures must be exposed at the outset, but approximately 10% of epileptic patients also manifest coexistent psychogenic seizures.²³ The range of conditions resembling epilepsy in early childhood is particularly broad.¹⁰⁴ Mistaking complex partial seizures for absence epilepsy, or failing to identify precipitating factors, such as sleep deprivation, leads to suboptimal therapy. Neurodegenerative disorders and inborn errors of metabolism are a rare but important cause of seizures that do not respond to medications, and inborn errors of metabolism may go unrecognized in patients with structural brain abnormalities. Indolent gliomas occasionally manifest as refractory epilepsy but are usually detectable with serial neuroimaging.

Seizure frequency and duration of epilepsy should not be the sole determinants of medical intractability, because infrequent seizures or short duration of epilepsy may still pose significant risks for the patient and impair quality of life. For example, yearly seizures may make it impossible for a patient to drive or perform certain occupations. Similarly, infrequent but severe seizures or recurrent bouts of SE pose a significant medical risk. Occasionally, patients with a relatively brief seizure history present with extremely frequent or disabling seizures, such as in epilepsia partialis continua or certain infantile syndromes. Medical intractability must therefore be assessed in the context of the patient’s quality of life and likelihood of seizure remission.

Recent studies suggest that patients at risk for intractable epilepsy can be identified with a surprising degree of accuracy early in the course of their epilepsy. This would suggest that patients can and should be placed on a “watch list” by their treating health care providers. Epilepsy patients who have already experienced multiple seizures before seeking treatment, and who have an inadequate response to initial attempts to control seizures using AEDs are likely to have refractory epilepsy.⁷¹ In children, for example, after adjusting for epilepsy syndrome, the occurrence of high seizure frequency, focal electroencephalographic (EEG) slowing, and acute symptomatic or neonatal SE are associated with a higher risk of intractability, whereas seizure onset between 5 and 9 years was associated with a lower risk.¹⁰ In contrast, lack of early remission in patients who are not medically refractory is not associated with a poor outcome.⁹

Identification of a Surgically Remediable Syndrome

A significant proportion of epilepsy surgery candidates manifest seizures in the context of specific epilepsy syndromes. Candidates for excisional procedures usually have localization-related epilepsy syndromes, whereas candidates for commisural surgery typically have generalized epilepsy syndromes. Patients with surgically amenable epilepsy syndromes fall within the cryptogenic and symptomatic etiologic groups, because idiopathic epilepsy syndromes are genetically determined and are therefore unlikely to be influenced by surgical intervention. Many epilepsy syndromes are well characterized and have defined prognoses, simplifying the selection process; thus, idiopathic partial and generalized epilepsies are not surgically amenable, whereas surgery might be considered the treatment of choice for some specific epilepsy syndromes such as mesial TLE; neocortical epilepsy caused by discrete, easily resectable lesions; chronic epilepsy associated with Sturge-Weber syndrome; tuberous sclerosis; focal cortical dysplasia; hemimegalencephaly; and Rasmussen syndrome.

The Landau-Kleffner syndrome of acquired epileptic aphasia is a rare disorder characterized by regression of language in early childhood, prominent EEG abnormalities, and often seizures. In some children with impaired language who fail medical treatment, multiple subpial transection of perisylvian cortex on one side may be of benefit.⁵³

The syndrome of gelastic epilepsy and hypothalamic hamartoma is another rare epilepsy syndrome for which cortical resection and anterior callosotomy are ineffective in alleviating seizures.¹⁷ Recently, seizure remission has been achieved by hamartoma resection,⁸⁸^,¹¹⁶ confirming seizure origin in the hypothalamic lesion. These observations suggest that other cases of intractable epilepsy syndromes associated with subcortical lesions such as subcortical heterotopias may also be surgically amenable.

Surgical treatment is advocated for some patients with infantile spasms. Infants who fail to respond to treatment with conventional anticonvulsants and corticosteroids, and who demonstrate predominantly unilateral abnormalities on EEG, positron emission tomography (PET), and magnetic resonance imaging (MRI) may be suitable for lobar or multilobar cortical resection.²²^,¹⁰⁵ Early resective surgery is also indicated in infants with catastrophic presentations of partial seizures.²⁸

Loss of Quality of Life

Chronic epilepsy is associated with considerable comorbidity and deterioration in overall quality of life. Decline in social, behavioral, and intellectual domains is a prime motivating factor in seeking surgical intervention. This is especially true for children, who are at higher risk for behavioral and cognitive disturbances as well as the emergence of depression in adolescence.²⁹^,³⁰^,⁸⁶ Children with medically resistant seizures who are intellectually disadvantaged show an even greater compromise in their quality of life, irrespective of their intellectual ability level.¹⁰¹

The benefits of surgical intervention have recently been demonstrated for adults with TLE. Aydemir et al.⁴ found that successful surgery improved social activities and resulted in greater independence. Similarly, Lowe et al.⁷⁸ found that long-term seizure-freedom after temporal lobectomy uniformly improved quality of life.

Earlier definitive surgical intervention would be expected to set the stage for improved self-esteem, greater social opportunity, and career advancement. Two separate studies have confirmed this hypothesis. Van Empelen et al.¹¹⁸ performed a longitudinal follow-up of 21 children and found that, 2 years after surgery, children began to perceive themselves as being socially more competent and having greater self-worth. Adolescents began to improve sooner after surgery; at 2 years, they demonstrated improvement in the domains of athletic competence and romance. Sabaz et al.¹⁰² also found improved quality of life in a prospective study of families of 35 children with medically resistant epilepsy undergoing epilepsy surgery, but significant gains occurred primarily in patients who became seizure-free.

Role of Early Intervention

The evidence for neuronal injury and subsequent epileptic and cognitive deterioration in chronic epilepsy has been discussed. The detrimental effects of recurrent seizures pose special risks to the developing brain of infants and young children. Uncontrolled seizures also jeopardize the chances for an independent lifestyle, and children with intractable epilepsy are likely to be excluded from normal schools and from social and vocational opportunities.⁷⁴^,⁷⁵^,⁷⁶ Thus, uncontrolled chronic epilepsy has the potential for irreversible cognitive, behavioral, and psychosocial problems in later life. For this reason, syndrome characterization, medical intractability and likely prognosis should be established early in patients with uncontrolled seizures, allowing the consideration of early surgical intervention. Additional consideration takes into account the neural plasticity data from lesioning studies,⁶⁵^,⁶⁶^,⁶⁷ which suggest that functional recovery from cerebral resection is superior if surgery is performed earlier in postnatal life.

Goals of Surgery

The goal of any treatment for epilepsy is to permit the patient to live as normal a life as possible. Maximizing normal function and minimizing adverse effects is part of the overall goal of therapy, whether medical or surgical. Treatment must therefore restore a sense of well-being and alleviate the psychosocial disability, medical morbidity, mortality, and risk for associated seizures. Refractory-seizure patients often have cognitive, linguistic, motor, sensory, psychiatric, and social impairments,³^,³⁸^,³⁹^,⁶⁰^,⁶⁴^,¹¹¹ and treatment must also address these problems. A comprehensive rehabilitative plan is an important component in an epilepsy surgery program.

Two broad categories of surgical therapy for epilepsy, curative and palliative, define the relative success of surgical intervention. Curative surgery eradicates seizures and the need for medication, whereas palliative surgery lessens seizure severity or frequency or prevents the occurrence of some seizure types.
These outcome measures are similar to the goals of surgery in other conditions (e.g., for cancer surgery, complete tumor removal vs. debulking).

Curative surgery should eliminate the psychosocial disability associated with seizures and therefore remains the best hope for achieving a “normal” life, including improved schooling, greater personal independence, enhanced employment opportunities and attainment of a driver’s license. Studies have shown that return to a normal lifestyle after surgery rarely occurs in patients who do not achieve seizure freedom.¹¹⁰ However, curative surgery may also create unanticipated psychosocial problems due to newly acquired independence or upward vocational mobility.¹⁵

In some patients, only one of several seizure types is cured by surgery, but this may be a worthwhile outcome. For example, patients with mixed seizure disorders usually benefit from elimination or marked reduction in the frequency of tonic or atonic seizures, due to the reduction in medical risk and attendant injuries (e.g., fractures, lacerations). Similarly, patients with TLE may benefit from abolition of complex partial seizures and tolerate occasional auras without loss of consciousness. Palliation requires a clear a priori definition of the surgical objectives, so that an intelligent appraisal of results can follow surgery.

Only gold members can continue reading. Log In or Register to continue