Strategies for Optimizing Antiepileptic Drug Therapy in Elderly People

Thomas E. Lackner, Pharm.D., FASCP


Pharmacotherapy. 2002;22(3) 

In This Article

Seizure Disorders

The frequency of seizure disorders in persons aged 75 years and older exceeds that in all other age groups. Complex partial seizures followed by generalized tonic-clonic seizures are the most common seizure types in the elderly.[3] However, in many older people, the type is unknown or is undocumented, potentially complicating assessment of AED appropriateness ( Table 1 ).[9,10,11,12,13,14,15,16,17,18] Before considering pharmacotherapy, reversible etiologies should be identified. The most common conditions are thrombotic and hemorrhagic strokes (accounting for about one-third of all cases of epilepsy), dementias, central nervous system (CNS) neoplasms, subdural hematoma such as with mild-to-severe head injury (often after falls), CNS and systemic infections, metabolic abnormalities (sometimes caused by drugs) such as uremia and hepatic failure, nonketotic hyperglycemia, hypoglycemia, hyponatremia, hypocalcemia, hypoxia, and tachyarrhythmias. Other etiologies are emotional stress, insufficient sleep, certain drugs and herbal products (rarely as direct causes), and abrupt drug discontinuation.[3,19,20,21,22,23,24,25] Seizures, generally partial, occur in 4-43% of persons after stroke, with greatest frequency after cortical infarcts and large hemorrhagic infarcts; recurrent seizures occur in 20-80% of patients experiencing a poststroke seizure.[25]

Drugs most often implicated in inducing typically generalized tonic-clonic seizures in patients with or without a history of epilepsy are phenothiazines (especially chlorpromazine), loxapine, clozapine, olanzapine, quetiapine, fluoroquinolones (especially trovafloxacin and enoxacin), penicillins, and cephalosporins. Other provocative agents are maprotiline, amoxapine, bupropion, clomipramine, fluvoxamine, tricyclic antidepressants, venlafaxine, lithium, theophylline, metoclopramide, isoniazid, ephedrine, meperidine, and insulin.[21,22,23] Several herbal products (St. John's wort, ginkgo biloba, ginseng, kava kava, ma huang) may provoke seizures, especially when taken with drugs known to lower the seizure threshold.[24]

Elderly people often have many risk factors for drug-induced seizures. These include a history of seizures, head injury (falls), recent stroke, excessively high drug dosages or serum drug concentrations (SDCs) due in some cases to diminished ability to eliminate drug, increased pharmacodynamic sensitivity, abrupt dosage increase or decrease, intrinsic epileptogenicity, and concomitant therapy with an agent that lowers the seizure threshold.[22] Seizures also can be precipitated by rapid withdrawal of many traditional AEDs, benzodiazepines, phenothiazines, serotonin reuptake inhibitors, venlafaxine, baclofen, and alcohol.[22,23]

Seizures caused by drugs are clinically indistinguishable from primary seizures.[26] Other symptoms of drug toxicity (e.g., serotonin syndrome) and a temporal relation with a newly prescribed agent, increased dosage, or discontinuation help identify the causative agent. Remedial measures involve avoiding provocative agents and prescribing the lowest effective dosage when the drug is necessary, stroke prophylaxis, and mitigating other underlying factors such as metabolic disturbances. Since alcohol abuse is prevalent in elderly people, this possibility should be elicited carefully. After a suspected drug-induced seizure, the provocative agent should be discontinued as soon as possible.

In addition to antiepileptic pharmacotherapy for repeat drug-induced seizures and status epilepticus, seizures related to theophylline toxicity may necessitate hemodialysis or hemoperfusion to decrease the theophylline concentration rapidly. Seizures related to barbiturate withdrawal are best managed by temporarily restarting the drug and withdrawing it more gradually.[22]

A detailed description and documentation of seizures (time of onset, duration, events before and after) are necessary for accurate diagnosis and optimal treatment. This is especially important in elderly people, as seizures can be confused with acute behavioral disorders, including those caused by some drugs (delirium), cardiac arrhythmias, syncope, transient ischemic attacks, transient global amnesia, vertigo, hypoglycemia, nonketotic hyperosmolar states, and drug-induced dystonia.[27,28,29]

When selecting an AED, its expected efficacy and risk of recurrent seizures and their complications must be weighed against the potential for adverse reactions as well as the patient's desire for treatment. Whereas the rate of recurrence after a single untreated seizure in the elderly is somewhat greater than that in younger adults, the difference is not significant.[30] However, elderly people are especially vulnerable to sequelae of seizures and status epilepticus, including physical injury such as compression fractures of thoracic vertebrae (especially with underlying osteoporosis) or fractures from falls with potential loss of mobility and independence, institutionalization, and death.[31,32,33] Antiepileptic drug therapy is generally beneficial in patients with recurrent seizures. In addition, because of the higher risk of seizure recurrence in a patient with a structural cortical lesion (i.e., brain tumor, encephalomalacia due to a stroke, or trauma), starting AED therapy after a single seizure is appropriate in this case.[25]

As in younger adults, epilepsy in the elderly can diminish cognitive function and self-esteem, increase social isolation and anxiety, and cause depression that may reduce quality of life.[34,35,36,37,38] The disease exacts a considerable economic toll. Although not analyzed specifically in elderly people, its economic burden may be particularly problematic in these individuals, many of whom have limited fiscal resources. Whereas indirect costs related to epilepsy (loss of earnings, household productivity) among persons aged 65 years and older are low compared with younger adults, elderly people account for a dispropor-tionately larger fraction of direct costs (diagnostic procedures, physician and hospital services, laboratory tests).[39] In one study, the lifetime economic cost of managing epilepsy in all individuals with frequent seizures was 32-fold greater than that incurred by patients experi-encing remission after a first seizure.[40]

Limited data indicate that the effectiveness of AEDs in reducing seizure frequency in elderly patients with epilepsy (complex partial, secondary generalized tonic-clonic) is similar to that in younger adults.[41,42,43] Moreover, control may be achieved with lower dosages and SDCs (carbamazepine, valproate) than required in younger adults.[43] Despite numerous shortcomings, particularly in elderly people, traditional AEDs (phenytoin, carbamazepine, valproate) are considered by many to be agents of first choice for partial and generalized tonic-clonic seizures. This is largely because of more seizure indications, greater clinical experience, and lower cost compared with newer AEDs ( Table 2 ).[44,45] Phenobarbital and primidone are as effective as other traditional AEDs, and phenobarbital is inexpensive, but both agents should be avoided in the elderly because of their propensity for untoward effects such as behavior disorders, cognitive impairment, and sedation that can impair quality of life, complicate care, and increase the risk of falls and hip fractures.[6,46,47] Despite these troublesome effects, phenobarbital is the second most commonly prescribed adjunctive AED in nursing home residents.[2] However, even preferred AEDs, such as phenytoin, the most commonly prescribed AED in nursing home residents, often are implicated in adverse reactions in elderly people ( Table 3 ).[1,2,4,5,48]

Because of less complex pharmacokinetics and a smaller risk for drug-drug and some drug-disease interactions, newer AEDs should be considered in the elderly with partial seizures refractory to traditional AEDs, individuals requiring many agents for seizure control, those intolerant to traditional AEDs, and possibly elderly people at risk for drug-drug or drug-disease interactions ( Table 4 , Table 5 , and Table 6 ).[24,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69] The role of newer agents as initial therapy in all elderly people is not established, and they are more expensive than traditional drugs. In addition, only one controlled clinical trial of efficacy of a new AED has been conducted exclusively in the elderly.[12] Whereas newer AEDs are not approved by the Food and Drug Administration (FDA) for monotherapy or primary generalized seizures, some may be effective when administered in this manner for some seizure types.[13,14,15,16,17,18]

Since the elderly are predisposed to problems associated with polypharmacy (nonadherence, drug interactions) and limited fiscal resources, AED monotherapy is preferred and often can achieve good seizure control.[41,43,70,71] For example, 75% of 67 elderly patients (mean age 73 yrs) taking primarily phenytoin or carbamazepine monotherapy were adequately controlled (< 3 seizures/yr) after an average of 33 months treatment for predominantly partial seizures.[71] In a retrospective analysis of 251 outpatients and nursing home residents at least 60 years of age, 62% of previously untreated patients with established epilepsy were seizure free with AED monotherapy, compared with 47% of those with established epilepsy after at least 12 months of observation.[41] Seizure prognosis was unrelated to age at seizure onset or degree of dementia (at least 8 patients had dementia). Since the time to steady state of some AEDs can be longer than that in younger adults (delayed time to maximum benefit), a longer than usual trial is advised before considering drug substitution. In addition, before an alternative AED is considered in patients inadequately controlled with current therapy, prescription refill records and SDCs should be assessed to evaluate adherence with therapy. Factors that can adversely affect adherence include cognitive impairment, disturbed vision, hearing impairment, dysphagia, impaired dexterity, limited fiscal resources, and lack of transportation to a pharmacy. Progression of seizure etiology and iatrogenic factors (insufficient dosage possibly related to overcompensating for advanced age) also should be considered if control declines.

Since up to 40-50% of adults of all ages refractory to an initial AED will achieve adequate control with a different AED as monotherapy, another drug known to be effective as monotherapy should be substituted and titrated to its maximum tolerated dosage.[72] If seizure control is not achieved with two attempts using different, single AEDs, adjunctive therapy with a second AED that has a different mechanism of action and side effects should be considered. Approximately 10-38% of young adults refractory to monotherapy achieve at least a 50% decrease in frequency of partial seizures after adding another traditional AED.[73,74,75] Similar results were achieved with newer AEDs as adjunctive therapy in young adults with refractory epilepsy (20-40% efficacy).[76] Considering the good response to monotherapy, it is appropriate to taper the first drug's dosage gradually over 3-6 months once acceptable control is achieved with an add-on AED.[72,77] However, this assumes antiepileptic efficacy of the add-on drug as monotherapy.[14,15,16,17,18,70]

Given the risk of adverse reactions and economic costs, gradual withdrawal of therapy should be considered in the absence of an indication and/or documented absence of seizures for at least 3-5 years, especially with lower than usual therapeutic AED total SDCs (unbound SDC of phenytoin and possibly valproate).[10,11,78,79,80,81,82,83,84] However, evidence supporting this recommendation specifically for elderly people is limited.[11,81] Often complicating this decision are incomplete seizure records and difficulty interpreting SDCs because of possibly altered pharmacokinetics and pharmacodynamics.

Whereas therapy with carbamazepine or phenytoin is recommended for a few weeks, but no longer than 6 months, to prevent seizure recurrence after an ischemic stroke, routine primary prophylaxis is not advised.[70,78] Primary prophylaxis also is generally not recommended after intracerebral hematomas. However, to reduce the risk of cerebral rebleeding after subarachnoid hemorrhage from a ruptured aneurysm, primary seizure prophylaxis generally is limited to the immediate posthemorrhagic period up to 1 week, although some believe continuing therapy up to 6 months may be appropriate. In fact, the value of this intervention is controversial because data are limited to nonrandomized trials and conflicting findings.[79,80] Long-term AED prophylaxis after a subarachnoid hemorrhage should be considered only in the presence of high-risk factors, such as preexisting seizures, infarctions, middle cerebral artery aneurysms, intraparenchymal hematoma, history of hypertension, and possibly dementia and other neurologic disorders.[79,80]

Since abrupt AED withdrawal is likely to cause seizure relapse and status epilepticus, the agent should be withdrawn gradually.[81,82] However, without controlled studies in the elderly, it is uncertain whether the time of withdrawal differs from that recommended in younger adults. In two studies of children and adults of unspecified mean age in one study[81] and mean age of 24 years in the other,[83] the age at onset of withdrawal was not prognostic of seizure relapse. The rate of seizure relapse was evaluated in a controlled study of 1013 children and adults (median age 27 yrs), including an unspecified number of individuals aged 50 years and older who were seizure free for at least 2 years before enrollment.[81] The rate of AED withdrawal in adults was decrements of every 4 weeks of no more than phenobarbital 30 mg, phenytoin 50 mg, carbamazepine 100 mg, valproate 200 mg, primidone 125 mg, and ethosuximide 250 mg, with the target withdrawal period being at least 6 months. The seizure relapse rate 2 years after randomization was 41% after AED discontinuation, compared with 22% in patients continuing therapy. The nonsignificant difference in median (95% confidence interval [CI]) relative risk of seizure relapse in patients at least 50 years of age (unspecified number) was 1.25 (0.86-1.82) compared with 1.56 (1.10-2.21) for persons aged 35 to less than 50 years.

In an open-label study, 18 of 72 elderly nursing home residents (mean age 66 yrs, 45.7% received ≥ 2 AEDs) had the drugs discontinued over an unspecified time period.[11] Withdrawal was predicated on a prolonged (undefined) absence of seizures and on total SDC less than the usual therapeutic range.[81] No patients experienced a seizure during the 6-12 months after AED discontinuation. However, because the medical indication for the AED was unknown in many patients and 56% of the patients did not have seizures documented in medical records, the rate of relapse may underpredict relapse in an elderly population with documented seizures. In a retrospective study of adverse drug events in 132 nursing home residents (mean age 70 yrs), phenytoin had to be reinstituted in 30% of those for whom it had been discontinued since admission 6 months earlier.[5] The rationale for, and rapidity of, AED withdrawal and patient characteristics were not reported. Moreover, the study design could not distinguish whether a seizure after AED discontinuation would have been prevented by continuing therapy.

For adults of all ages, conditions prognostic for seizure recurrence after AED withdrawal are neurologic abnormalities such as stroke, head injury, mental retardation, congenital defects, and focal neurologic signs. Additional risk factors are an abnormal electroencephalogram (EEG) and severe epilepsy (many seizures, difficult to control seizures, > 1 AED).[84] When they occur after stroke, recurrence is more likely after relatively late than early seizures. Therefore, unless it is urgently necessary to discontinue therapy (serious adverse AED reaction), slow tapering over 3-6 months should be considered in patients who are seizure free for 3-5 years. Additional factors associated with successful drug withdrawal are single type of partial or generalized tonic-clonic seizure, normal EEG without epileptic activity before and after treatment, normal neurologic examination and intelligence quotient (IQ), reliable observer, and absence of alcohol or other drug abuse; these patients would likely survive a recurrence.[81,84]

No specific guidelines exist for tapering AED dosages in the elderly, but data from predominantly young adults suggest that it should be no faster than every 4-6 weeks in decrements of approxi-mately 20%. During and after withdrawal, patients should be monitored carefully for any type of seizure activity.