INTRODUCTION — Seizures are a common and potentially devastating complication of both primary and metastatic brain tumors [1]. Such seizures are focal in origin and may either remain focal or secondarily generalize. The diagnosis of a seizure disorder is usually made clinically.
The etiology, epidemiology, evaluation, and treatment of seizures and the limited role of prophylactic antiseizure therapy in patients with brain tumors will be reviewed here. The clinical manifestations of brain tumors and the general management of seizures are discussed elsewhere. (See "Overview of the clinical features and diagnosis of brain tumors in adults", section on 'Clinical manifestations' and "Overview of the management of epilepsy in adults".)
CAUSES OF SEIZURES — Brain tumors and metastases can cause seizures by a variety of mechanisms; alterations in excitatory neurotransmission and extracellular ion currents are considered the most likely [2-5]. Other potential mechanisms include hypoxia, acidosis, inflammation, mechanical effects, and disruption of local homeostasis with changes in electrolytes, perfusion, and metabolism.
Seizures may also be caused by factors other than metastases in patients with tumors arising outside the central nervous system (see "Evaluation and management of the first seizure in adults", section on 'Causes of seizures'). Examples of potential seizure etiologies that are more common in patients with cancer compared with the general population include metabolic encephalopathies such as hyponatremia or hypoglycemia, opportunistic infections, or side effects of therapy. Paraneoplastic encephalitis is another potential cause of seizures and altered mental status in patients with systemic cancer [6]. (See "Paraneoplastic and autoimmune encephalitis".)
Case series emphasize that nonconvulsive status epilepticus NCSE may occur in patients with systemic cancer who do not have intracranial metastases [6,7]. One study of patients with systemic cancer reported NCSE in 6 percent of those with impaired mental status; none had brain metastasis, and the origin of the seizure activity was presumably related to systemic factors such as metabolic encephalopathy (eg, hyponatremia or hypoglycemia), infection, or side effects of therapy [7].
EPIDEMIOLOGY — Seizures are a relatively common problem in patients with brain tumors. Seizures may be the initial manifestation of a brain tumor or may occur during the course of disease. (See "Overview of the clinical features and diagnosis of brain tumors in adults", section on 'Seizures'.)
The main factors that influence the incidence of seizures are tumor type, grade, and location:
●Primary brain tumor type and grade – Among patients with primary brain tumors, seizures are more common with low-grade tumors than with high-grade tumors [8]. The prevalence rates of epilepsy in a series of 1028 patients with primary brain tumors were 85, 69, and 49 percent among patients with low-grade glioma, anaplastic glioma, and glioblastoma, respectively [9]. Other studies have found that at least one seizure occurs in up to 80 percent of patients with high-grade glioma at some time during the course of disease [10]. Other less common low-grade tumor types also appear to have a high incidence of seizures. As an example, ganglioglioma and dysembryoplastic neuroepithelial tumors (DNETs) have an incidence of seizures greater than 80 percent [11,12].
Genetic characteristics of gliomas may also influence seizure propensity. In one study, patients with gliomas harboring mutations in isocitrate dehydrogenase type 1 (IDH1) were more than twice as likely to present with seizure as those with IDH1-wildtype gliomas (59 to 74 percent versus 18 to 34 percent) [13]. Preclinical studies suggested that D-2-hydroxyglutarate, a byproduct of the mutant IDH1 enzyme, was epileptogenic. (See "Molecular pathogenesis of diffuse gliomas", section on 'Isocitrate dehydrogenase (IDH) gene'.)
●Metastatic tumors – Seizures are less common in patients with metastatic lesions compared with those with primary brain tumors [9,14]. A systematic review published in 2017 that included over 2000 patients with intracranial metastatic disease found that seizures affected approximately 15 percent; the seizure rate with metastatic melanoma tumors was higher than expected, while the seizure rate with prostate tumors was lower than expected [15]. In a retrospective series that included 470 patients with brain metastases, seizures were diagnosed in 24 percent at some point during the illness [16]. Seizures were most common in melanoma (67 percent) and least common in breast cancer (16 percent). Lung (29 percent) and unknown primary sources (25 percent) were intermediate. The high rate of seizures in patients with melanoma may reflect the tendency of melanoma to involve the cerebral cortex and to hemorrhage.
●Tumor location – Clinical situations in which the risk of seizures may be relatively higher include patients with metastases involving areas of high epileptogenicity such as the motor cortex or when metastases involve both the brain and leptomeninges.
CLINICAL MANIFESTATIONS — Seizures that arise in patients with brain tumors are either focal or secondarily generalized. The specific ictal and peri-ictal manifestations typically reflect the tumor's location within the brain and are described in detail separately. (See "Focal epilepsy: Causes and clinical features".)
Patients with brain tumors can also develop status epilepticus, which may be either convulsive or nonconvulsive. (See "Convulsive status epilepticus in adults: Classification, clinical features, and diagnosis" and "Nonconvulsive status epilepticus: Classification, clinical features, and diagnosis".)
Nonconvulsive status epilepticus (NCSE) has been reported in patients with systemic cancer. The clinical manifestations of NCSE are often nonspecific, such as an altered personality or a decreased level of alertness. These can be difficult to distinguish from, and may coexist with, other causes of altered mental status. (See "Nonconvulsive status epilepticus: Classification, clinical features, and diagnosis".)
DIAGNOSTIC EVALUATION
Neuroimaging — A new seizure should prompt a neuroimaging study with computed tomography (CT) or magnetic resonance imaging (MRI), depending on the clinical setting, regardless of a known diagnosis of systemic cancer or primary brain tumor. Brain MRI has better sensitivity than CT; nevertheless, a CT scan is suitable to exclude a major change in the size of a tumor, hemorrhage, or large stroke under emergency situations or if MRI is unavailable or contraindicated. (See "Evaluation and management of the first seizure in adults", section on 'Neuroimaging' and "Neuroimaging in the evaluation of seizures and epilepsy".)
Seizures themselves, particularly if prolonged, can produce changes on brain MRI or positron emission tomography (PET), which broadens the differential for patients with brain tumors and new imaging abnormalities [17,18]. (See "Magnetic resonance imaging changes related to acute seizure activity".)
Electroencephalography — Electroencephalography (EEG) is essential for making the diagnosis of nonconvulsive status epilepticus (NCSE). As a general rule, any fluctuating or unexplained alteration in behavior or mental status warrants consideration of NCSE and evaluation with EEG. (See "Nonconvulsive status epilepticus: Classification, clinical features, and diagnosis".)
However, EEG may not be required in a patient who has had a clinically obvious seizure and has recovered. EEG is also not routinely needed for those without clinical evidence of a seizure.
Laboratory testing — Rapid point-of-care glucose should be checked in all patients with a first seizure. Other laboratory evaluations that are appropriate for the evaluation of a first seizure include electrolytes, glucose, calcium, magnesium, complete blood count, renal function tests, liver function tests, urinalysis, and toxicology screens, although the likelihood of finding a relevant abnormality in unselected patients is low. An electrocardiogram (ECG) should be performed in all patients with loss of consciousness, as cardiogenic syncope can manifest as a secondary hypoxic seizure. (See "Evaluation and management of the first seizure in adults", section on 'Initial evaluation'.)
In rare cases, patients with systemic cancer (but no brain metastases) may develop a paraneoplastic encephalitis with manifestations that include focal seizures. The clinical features and diagnostic evaluation of paraneoplastic encephalitis is discussed in detail elsewhere. (See "Paraneoplastic and autoimmune encephalitis".)
APPROACH TO MANAGEMENT
Treating seizures — Seizures in and of themselves are an important source of morbidity and mortality in patients with primary and secondary brain tumors and require aggressive treatment [19]. Patients who experience a seizure and those who have a history suggestive of previously unreported or unrecognized seizure activity due to a brain tumor should be treated with monotherapy with a standard first-line antiseizure medication because of the high risk of seizure recurrence. (See 'Patients with seizures' below.)
Tumor-directed therapies — Treatment of the causal tumor with surgery, radiation therapy, or systemic therapy (chemotherapy, molecularly targeted agents) may ameliorate seizure activity whether or not patients can be adequately managed with drug therapy [20]. Seizure control is an important potential benefit of treatment in patients with low-grade gliomas and other tumors, independent of tumor response and overall survival [21].
●Surgery – Experts widely agree on gross total resection of a primary brain tumor with clear margins when this is feasible. A systematic review of 910 reported patients with low-grade glioneuronal tumors (gangliomas and dysembryoplastic neuroepithelial tumors [DNETs]) found that seizure remission was achieved after gross total resection in 80 percent of patients [22]. While some studies have suggested that extratemporal epilepsy may be more difficult to control with surgery compared with temporal lobe epilepsy, outcomes in this report were not different between those with temporal lobe versus extratemporal tumors. As has been suggested in other reports, extended resection of temporal lobe tumors with hippocampectomy conferred additional benefit [22-24]. Prognostic factors that have been associated with improved seizure outcomes after brain tumor resection include shorter duration of epilepsy prior to surgery, a single focus on electroencephalography (EEG), a single lesion on neuroimaging, and complete tumor resection [25]. (See "Surgical treatment of epilepsy in adults".)
●Radiation therapy – Radiation therapy can result in improved seizure control, even in the absence of an objective tumor response by magnetic resonance imaging [26-28]. In a multicenter trial that compared early versus delayed radiotherapy for low-grade gliomas, early radiation treatment was associated with a lower seizure rate at one year (25 versus 41 percent), although it did not affect overall survival [26].
●Systemic therapy – There are reports of patients with low-grade primary brain tumors and epilepsy having improved seizure control and a few achieving seizure freedom with the use of temozolomide or nitrosoureas [28-31]. The mechanistic target of rapamycin (mTOR) inhibitor, everolimus, has been shown to decrease the size of subependymal giant-cell astrocytomas in tuberous sclerosis and also resulted in improved seizure control in an open-label study of 28 patients [32]. It was unclear if seizure reduction with everolimus was a result of decreased tumor burden or perhaps an antiseizure effect of mTOR inhibition, such as regulating protein synthesis and microanatomic changes that underlie synaptic, long-term potentiation and depression.
Driving — Patients with uncontrolled seizures should not drive.
There are no general guidelines on driving restrictions for patients with seizures that are controlled on medication or during the period of antiseizure medication taper. State regulations differ as to the duration a patient must be seizure free in order to legally drive. This is discussed in detail separately. (See "Driving restrictions for patients with seizures and epilepsy".)
Patients with intracranial tumors who have not had seizures and who drive are often placed on prophylactic antiseizure medications despite the lack of evidence that this practice reduces the incidence of seizures. (See 'Patients without seizures' below.)
The implications for driving in patients who have not experienced a seizure are discussed separately. (See "Overview of the treatment of brain metastases", section on 'Driving'.)
PATIENTS WITH SEIZURES
Indications for antiseizure medication therapy — Antiseizure medication therapy is indicated for patients who experience a seizure and for those who have a history suggestive of previously unreported or unrecognized seizure activity due to a brain tumor.
Choosing initial therapy — Antiseizure medications with no or minimal hepatic enzyme-inducing or -inhibiting properties, are generally preferred for the treatment of patients with tumor-related seizures and epilepsy; options include:
These agents have a more favorable safety profile compared with older agents and often have fewer problematic drug-drug interactions (table 1) [33-37]. The choice among non-enzyme inducing agents should be individualized according to patient preferences and comorbidities, as there is no compelling evidence to support the superiority of one agent over another in this setting [38,39].
To minimize toxicity, patients should be treated with monotherapy using the lowest effective dose of the antiseizure medication. Monitoring of serum drug levels is indicated only with older agents such as phenytoin, valproate, carbamazepine, and phenobarbital. Newer agents such as levetiracetam, topiramate, lamotrigine, lacosamide, and pregabalin have wider indications and dosing ranges but no generalizable reference ranges. A similar antiseizure medication level may provide seizure-free quality of life in some patients but correlate with symptoms of toxicity in other patients. With newer agents, level monitoring can be helpful to confirm compliance and establish a patient-specific effective level, which is therefore useful only when compared with subsequent levels from the same individual. Otherwise, drug level monitoring is rarely useful with newer agents [40]. These patients should be followed clinically for neurotoxicity and seizure control. Potential interactions between the antiseizure medication and any chemotherapeutic agent should also be considered when choosing an antiseizure medication and monitoring levels (table 2). (See "Initial treatment of epilepsy in adults", section on 'Selection of an antiseizure medication'.)
Approximately 50 percent of patients with tumor-related epilepsy will respond adequately to a single antiseizure medication [10]. The initial use of multidrug regimens should be avoided whenever possible. Monotherapy increases the likelihood of compliance, provides a wider therapeutic window, and is more cost effective than combination drug treatment. There are fewer side effects and idiosyncratic reactions associated with single-drug therapy, and drug-drug interactions are avoided. (See "Overview of the management of epilepsy in adults", section on 'Subsequent drug trials'.)
As in the general epilepsy population, levetiracetam is generally well tolerated but can cause neuropsychiatric side effects, including irritability, agitation, and anxiety. Patients with frontal lobe tumors may be at highest risk for neuropsychiatric adverse effects on levetiracetam and should be monitored for new psychiatric symptoms [41]. (See 'Drug-drug interactions' below and "Initial treatment of epilepsy in adults", section on 'Selection of an antiseizure medication'.)
Of note, several observational studies, including one small prospective study, have suggested that valproate may be associated with improved survival when taken during radiation therapy in patients with glioblastoma [42-44]. However, a pooled analysis of over 1800 patients enrolled in four randomized trials for newly diagnosed glioblastoma found no difference in overall survival among patients taking valproate compared with those who were not (hazard ratio [HR] 0.96, 95% CI 0.80-1.15) [45]. Similarly, levetiracetam was not associated with improved outcomes [45], in contrast with results of a separate, smaller observational study [46]. Among patients with gliomas and epilepsy, one retrospective study found that levetiracetam treatment was associated with a lower frequency of treatment failure from uncontrolled seizures compared with valproate [47]. Other studies have found that valproate may also be associated with a higher rate of hematologic toxicity that can lead to treatment delays. Preferential use of valproate over other antiseizure medications in patients with high-grade gliomas therefore does not seem justified.
Addressing precipitating factors — Brain tumors may also precipitate seizures due to tumor enlargement, alterations in intracranial pressure, metabolic changes, and vascular complications [38,48]. Management of these precipitating factors may improve seizure control in conjunction with antiseizure medication therapy.
Recurrent or refractory seizures — If a patient has recurrent seizures after initiation of therapy, doses of the initial agent should be titrated as tolerated and adequate serum concentrations (when applicable) should be verified before switching drugs or adding a second agent. If adequate seizure control cannot be obtained, an alternative antiseizure medication should be prescribed or a second antiseizure medication added.
Lacosamide has been increasingly studied as an adjunctive agent in brain tumor patients with refractory epilepsy, is generally well tolerated, and was associated with improved seizure control in a prospective study [49-51]. Brivaracetam is an analogue of levetiracetam and is approved in Europe and the United States as an adjunct for focal onset epilepsy. A small study showed tolerability and efficacy of this agent in tumor-related epilepsy, although side effects occurred in 21 percent [52]. An observational pilot study of patients with brain tumor-related epilepsy found that adjunctive treatment with zonisamide was associated with seizure reduction at six months [53]. A variety of other agents are equally reasonable in this setting, taking into consideration mechanism of action, side effect profile, drug interactions, and concomitant medications. (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects".)
The evaluation and treatment of convulsive and nonconvulsive status epilepticus is reviewed separately. (See "Convulsive status epilepticus in adults: Classification, clinical features, and diagnosis" and "Convulsive status epilepticus in adults: Management" and "Nonconvulsive status epilepticus: Classification, clinical features, and diagnosis" and "Nonconvulsive status epilepticus: Treatment and prognosis".)
Adverse effects of antiseizure medications
●Common and serious adverse effects – The important side effects of the antiseizure medications are listed in the table (table 3A-B). In patients with brain tumors, the incidence of adverse effects from antiseizure medications ranges from 30 to 40 percent, which is higher than that observed in seizure patients without brain tumors [14,54]. Overall, 24 percent of patients with brain tumors who are receiving antiseizure medications experience side effects that are severe enough to warrant a change in or discontinuation of antiseizure medication therapy [54]. Phenobarbital is associated with a higher incidence of shoulder-hand syndrome [55].
●Drug rash – Patients with brain tumors who are undergoing radiation therapy have a higher likelihood of drug rash with antiseizure medications; a small percentage develop Stevens-Johnson syndrome [56-58]. In addition, the combination of phenytoin, cranial irradiation, and the gradual reduction of concomitant glucocorticoid therapy may be particularly associated with erythema multiforme and/or Stevens-Johnson syndrome [58]. The mechanism is unknown but may result from depletion of suppressor T cells by the radiation therapy, allowing the development of a hypersensitivity reaction to phenytoin [57]. Stevens-Johnson syndrome has also been described in glioma patients receiving carbamazepine and lamotrigine [59,60]. Cross-sensitivity rash is common, particularly when switching from carbamazepine to phenytoin or vice versa [61]. Levetiracetam and valproate rarely cause rash [61]. Patients of Asian ancestry should be tested for the HLA-B*1502 allele before prescribing carbamazepine and possibly phenytoin or other antiseizure medications [62]. (See "Stevens-Johnson syndrome and toxic epidermal necrolysis: Pathogenesis, clinical manifestations, and diagnosis" and "Initial treatment of epilepsy in adults", section on 'Side effect profiles'.)
Drug-drug interactions — Several antiseizure medications have clinically important interactions with other medications commonly used in patients with brain tumors. These drug-drug interactions can have several different mechanisms [20]:
●Enzyme induction – Induction of hepatic metabolism via the cytochrome P450 (CYP) system can increase the clearance of other agents. Antiseizure medications that induce the P450 system can significantly reduce serum levels of antitumor agents such as the nitrosoureas, cyclophosphamide, procarbazine, vincristine, and others (table 4 and table 2). As a result, a higher dose of these antineoplastic drugs may be required in patients taking enzyme-inducing antiseizure medications. Examples of antiseizure medications that are potent inducers of hepatic drug metabolism include phenobarbital, phenytoin, and carbamazepine. Of note, CYP isoenzymes play only a minor role in metabolism of temozolomide, an alkylating agent often used for patients with primary brain tumors; therefore temozolomide levels are not affected by enzyme-inducing antiseizure medications [38].
In addition, antitumor drugs may induce the CYP enzymes, thus changing the serum levels of antiseizure medications (table 2) [8].
Antiseizure medications may also interact with glucocorticoids used to control cerebral edema. As an example, phenytoin induces the hepatic metabolism of dexamethasone, significantly reducing its half-life and bioavailability [8]. Conversely, dexamethasone induction of these enzymes may reduce serum phenytoin levels, potentially compromising seizure control.
●Enzyme inhibition – Valproic acid inhibits multiple components of the CYP system. This can result in decreased metabolism of chemotherapy agents such as the nitrosoureas, and therefore increased toxicity due to bone marrow suppression [63].
Antiseizure medications that are not metabolized via the CYP enzyme system and do not inhibit these enzymes thus have important advantages over older agents. These include levetiracetam, pregabalin, lacosamide, lamotrigine, topiramate, and perampanel (table 1). (See "Antiseizure medications: Mechanism of action, pharmacology, and adverse effects".)
●Displacement of protein-bound agents – Antiseizure medications and chemotherapy agents that are highly protein bound can interact by displacing each other. This displacement can alter levels of free and protein-bound drug for both antiseizure medications and the chemotherapy drug. Levetiracetam has little protein binding.
Duration of antiseizure medication therapy — For select patients with a sustained period of seizure freedom who are deemed at low risk for recurrent seizures, tapering an antiseizure medication in order to minimize side effects is an option. Careful re-evaluation of the need for ongoing drug treatment is important, since antiseizure medication-related somnolence, dizziness, and mood changes can severely impact quality of life [64]. However, quality of life can also be affected by recurrent seizures, as has been shown in several studies [65].
Note that seizure worsening (eg, increased frequency or altered semiology) after a seizure-free interval in a patient with a known brain tumor, especially a low-grade tumor, could be an indication of a change in the tumor and should prompt repeat imaging.
●Factors associated with risk of seizure recurrence – Numerous factors may affect the risk of seizure recurrence upon withdrawing antiseizure medication therapy [64]:
•Seizure type – Generalized seizures are associated with a lower risk and partial seizures a higher risk of seizure recurrence
•Seizure frequency prior to remission – Low seizure frequency prior to remission is associated with a lower risk of recurrence
•Duration of seizure remission – One to two years of seizure freedom portends a lower risk and shorter periods a higher risk of seizure recurrence [64,66-69]
•EEG findings – Normal EEG findings are associated with a lower risk, while the presence of epileptiform abnormalities such as sharp or spike waves portends a higher risk of seizure recurrence
•History of status epilepticus – A history of status epilepticus portends a higher risk of seizure recurrence
•Tumor location – Midline and occipital lobe tumor location are associated with a lower risk, while insular, frontal, and temporal lobe locations are associated with a higher risk
•Tumor type – Anaplastic glioma, glioblastoma multiforme, metastatic disease (with the exception of melanoma), and lymphoma are associated with a lower risk of seizure recurrence; low-grade gliomas, glioneural tumors, and hemorrhagic metastases are associated with a higher risk
•Tumor treatment – Total surgical resection is associated with a lower risk, while radiotherapy, chemotherapy, and subtotal/partial resection are associated with a higher risk
•Tumor status at the time of antiseizure withdrawal – Long-term stable tumor after treatment and low short-term risk of tumor recurrence are associated with lower risk of seizure recurrence, while short-term stable disease and high risk of tumor recurrence are associated with a higher risk of seizure recurrence
•Patient-related factors – Factors that may influence patients to request antiseizure medication withdrawal include a high burden of adverse effects, undesired drug interactions, substantial financial burdens (eg, limited insurance coverage), or drug-associated teratogenic risk
Few studies have evaluated the risk of breakthrough seizures after antiseizure medication withdrawal; one small prospective observational study of patients with gliomas found that 26 percent had seizure recurrence during follow-up after antiseizure medication withdrawal, while seizure recurrence affected only 8 percent of patients who continued antiseizure medication treatment [70].
●Patient preferences – We find that antiseizure medication tapering is often driven by patient preferences. Some patients desire antiseizure medication withdrawal after a seizure-free period because of side effects or the inconvenience of taking medications; others will prefer continued antiseizure medication prophylaxis due to concern of seizure recurrence after antiseizure medication withdrawal [71].
●Timing of drug withdrawal – For most patients with brain tumors, a minimum period of one year of seizure freedom after the end of antitumor treatment seems to be appropriate before initiating drug withdrawal. However, the optimal timing of antiseizure medication withdrawal for patients with brain tumor is unknown [64]. Extrapolating from patients with nontumor-related epilepsy undergoing epilepsy surgery, a 2015 Cochrane review of five trials found that withdrawing antiseizure medications before two years of seizure freedom was associated with a higher risk of seizure recurrence compared with two or more seizure-free years [66].
●Method of drug withdrawal – Slow rates of antiseizure medication tapering over one to three months may reduce the risk of provoking seizures and also allow the treating physician to resume previous dosages if seizures occur during a drug taper. However, the optimal rate of tapering is unknown [72]. An EEG prior to and during drug withdrawal can aid management. One study found that in patients with worsening EEG findings (eg, the presence of sharp or spike waves) during antiseizure medication withdrawal, seizures recurred in 83 percent [73]. The optimal rate of drug withdrawal is also unknown.
ROLE OF PROPHYLAXIS
Patients without seizures — Prophylactic antiseizure medications are generally not recommended for patients with a primary or metastatic brain tumor and without a history of antecedent seizure [74-78]. An exception might be brain metastases from melanoma.
●Routine prophylaxis not indicated – A 2004 meta-analysis that included data from five randomized trials using phenobarbital, phenytoin, or valproic acid as prophylactic antiseizure medications concluded that there was no evidence to support their use in patients with brain tumors without a history of seizures, regardless of the type of tumor [74]. A 2008 systematic review by the Cochrane collaboration also concluded that there was no difference between placebo and treatment with phenobarbital, phenytoin, or valproic acid in preventing a first seizure [75]. Moreover, adverse events were more frequent with antiseizure medication therapy. Similarly, systematic reviews from 2010 and 2019 concluded that routine prophylactic use of antiseizure medications is not recommended, but evidence from randomized clinical trials was sparse [76,77].
●Melanoma metastases – In patients with metastatic brain lesions from melanoma, particularly those with multiple, hemorrhagic, or supratentorial brain metastases, the threshold to start antiseizure prophylaxis is lower compared with patients who have other metastatic or primary brain tumors. In a retrospective analysis of 109 patients with melanoma and brain metastases, there were 95 who did not have a seizure at diagnosis; of these, seizures developed later in 23 percent [79]. None of the 14 patients treated with prophylactic antiseizure medications developed seizures. In contrast, of the 81 patients who did not receive prophylaxis, 22 developed seizures, with a three-month seizure rate by univariate analysis of 17 percent. Risk factors included hemorrhagic and multiple supratentorial brain metastases. These risk factors were not disproportionately present in patients who did or did not receive prophylaxis.
Postoperative prophylaxis — Some experts suggest antiseizure medication prophylaxis for patients undergoing surgery for brain tumors. This recommendation places a high value on prevention of early postoperative seizures, which are uncommon but can be devastating, and a lower value on the potential adverse effects of antiseizure medications, which are of particular concern for older antiseizure medications such as phenytoin [80].
Other experts do not recommend prophylactic postoperative antiseizure medication treatment [15]. The data are drawn from observational studies and a limited number of small randomized trials. The results have been inconsistent; some but not all have reported a lower risk of early postoperative seizures in patients treated with a prophylactic antiseizure medication compared with no antiseizure medication [80-84]. A Cochrane analysis found only limited, low-certainty evidence regarding the effectiveness of prophylactic antiseizure medications after craniotomy [84].
●Choice of therapy – For prophylaxis after brain tumor surgery, levetiracetam is probably the most widely used and appears to be effective and well tolerated in the postoperative setting. Support for the use of levetiracetam is drawn from observational data and small randomized trials [82,85-89].
In one trial, 147 patients undergoing craniotomy for a supratentorial tumor were randomly assigned to receive levetiracetam (500 mg twice per day) or phenytoin (fosphenytoin 15 to 18 mg phenytoin equivalent [PE]/kg intravenous [IV] loading dose then 5 to 7.5 mg PE/kg/day maintenance until taking oral pills, then phenytoin 125 mg twice per day) starting at the induction of anesthesia and continuing for seven days [89]. Three-quarters of the patients had no history of seizures preoperatively. By the seventh postoperative day, patients who received levetiracetam had a significantly lower incidence of seizures compared with those who received phenytoin (1 versus 15 percent). Of note, the phenytoin group had blood levels in the low therapeutic range, which may have contributed to increased seizure frequency. All patients who received levetiracetam completed seven days of therapy; phenytoin was withdrawn due to adverse effects in five patients (rash, atrial fibrillation, or liver dysfunction).
A smaller trial randomly assigned 81 postcraniotomy patients (50 percent with intracranial tumors) to three days of IV levetiracetam or phenytoin followed by up to three months of oral therapy [82]. Levetiracetam was similarly tolerated compared with phenytoin, and there was no difference in discontinuation rate or number of patients reporting side effects. Seizures occurred in significantly fewer patients taking levetiracetam (zero versus six), and there was a trend towards more major side effects in patients assigned to phenytoin (mostly allergy and/or drug rash). Mood disturbance/irritability was the most common side effect in patients taking levetiracetam (19 percent).
●Duration of therapy – The optimal duration of postoperative seizure prophylaxis is not well established. Based upon a review of older studies, the American Academy of Neurology recommended that antiseizure medications should be gradually tapered beginning one to two weeks after the surgery and then discontinued in patients who remain free of seizures [54].
SEIZURES AT THE END OF LIFE
Frequency — Seizures have been reported in 35 to 50 percent of brain tumor patients in the last month of life [90-93]. Patients with a history of seizures, particularly those with late-onset epilepsy, appear to be at highest risk. In a retrospective study of 157 patients with high-grade glioma cared for at home, 37 percent of patients had at least one seizure in the month before death; 79 percent of seizures were focal, 18 percent were generalized, and 2.5 percent were status epilepticus. The risk of seizure was as high as 76 percent in patients with late-onset epilepsy and 11 percent in patients with no prior history of seizure [92]. In one study, the risk of seizure in the last month of life was higher for patients with high-grade glioma than for those with brain metastases (34 versus 10 percent) [94].
Management — It is useful for clinicians to discuss an end-of-life antiseizure medication strategy in advance with the patient, family, home care nurses, and other caregivers. Management of seizures and epilepsy in patients with brain tumors at the end of life has not been well studied. General principles include the following:
●Role of drug therapy – Patients with a history of seizures who are near the end of life should continue antiseizure medication therapy as long as they are able to take oral medications. As with earlier stages of the disease, there is no clear role for prophylactic antiseizure medications in patients who have never had a seizure.
●Dysphagia – Dysphagia is common in brain tumor patients at the end of life [93,95], and the route of administration of antiseizure medications may need to be adapted. In some cases, switching to a suspension or sprinkles (for valproate) can extend the length of time a patient is able to continue on oral therapy. When patients can no longer take medication orally, clinical judgment should be used as to whether to continue antiseizure medications in this setting; it can be appropriate to simply stop antiseizure medications, particularly if the patient's life expectancy is short.
●Nonoral administration – For patients unable to take oral medications, a variety of nonoral routes of administration may be considered, including rectal, transdermal, transmucosal, subcutaneous, and intravenous (IV) delivery. The availability of alternative routes varies by antiseizure medication and care setting, and pharmacokinetic data for rectal administration are limited for many drugs [90]. (See "Palliative care: The last hours and days of life", section on 'Non-oral routes of medication administration'.)
Phenobarbital, carbamazepine, valproate, and lamotrigine can be given rectally without the need for dose adjustments [96]. Rectal administration of levetiracetam suspension may lead to steady state trough concentrations of 50 percent less than with oral administration, suggesting that a dosing adjustment of approximately 2:1 may be necessary [97]. Some drugs can be given subcutaneously, including benzodiazepines and phenobarbital [98].
●Abortive medications for prolonged seizures – When prolonged seizures occur, benzodiazepines should be available for use as an abortive medication. Rectal diazepam (0.2 mg/kg or 10 to 20 mg) can be repeated hourly until the seizure stops. Sublingual or subcutaneous lorazepam can also be used. Buccal clonazepam and intranasal midazolam, when available, are also effective nonoral abortive options [99]. For patients who develop frequent seizures, scheduled doses of diazepam or lorazepam can be given.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Primary brain tumors".)
SUMMARY AND RECOMMENDATIONS — Seizures are a common complication of both primary and metastatic brain tumors. Management of seizures caused by brain tumors is similar to that in noncancer patients. (See "Overview of the management of epilepsy in adults".)
●Indications for antiseizure medication therapy – Patients who have experienced one or more seizures associated with a primary or metastatic brain tumor require antiseizure medication treatment. (See 'Choosing initial therapy' above and "Initial treatment of epilepsy in adults", section on 'Selection of an antiseizure medication'.)
For patients without a history of seizures who undergo brain tumor resection, we suggest that prophylactic antiseizure medications be given perioperatively (Grade 2C). Prophylactic antiseizure medications can be gradually tapered postoperatively if the patient remains seizure free. (See 'Patients without seizures' above.)
For other patients (those without a history of seizures and not undergoing a neurosurgical procedure), we recommend not using prophylactic antiseizure medication therapy (Grade 1B). (See 'Patients without seizures' above.)
●Choice of antiseizure medication – To avoid clinically significant drug-drug interactions, we suggest monotherapy with a nonenzyme-inducing antiseizure medication such as levetiracetam (Grade 2C). However, other antiseizure medications that can be titrated quickly when drug interactions are not a concern should also be considered. (See 'Choosing initial therapy' above.)
●Recurrent seizures – If patients have incomplete seizure control while on therapy, adequate serum levels should be verified before switching drugs or adding a second agent. (See 'Recurrent or refractory seizures' above.)
●End of life seizures – Seizures are common in brain tumor patients at the end of life, particularly in patients with late-onset epilepsy. It is useful to discuss an antiseizure medication strategy in advance with home care nurses and family members. For patients receiving hospice care, benzodiazepines should be available in the home for use as an abortive medication. (See 'Seizures at the end of life' above.)