Special considerations in treating the elderly patient with epilepsy

Etiology.

Seizures can be caused by any disorder affecting the brain. Common risk factors too often go unnoticed in patients over 60 years of age. Approximately one-third of patients with seizures who are 60 years or older have seizures of unknown etiology, although whether or not this represents the true proportion of idiopathic seizures is debatable. In the VACS #428 study, 24.6% of patients were classified as having seizures of unknown etiology (figure 1), and the majority of patients displayed risk factors for cerebrovascular disease such as cerebral infarction (34.1%), arteriosclerosis (14.9%), head trauma (6.9%), and hemorrhage (1.7%).4 Seizures in 18.8% of patients were caused by other conditions. Data from a large population-based study revealed that cerebral infarction elevates the risk for initial late seizures in the first year to 23-fold that of the general population.5

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Etiology of newly diagnosed epilepsy in patients ≈60 years old.4 *“Other” includes known etiologies such as arteriovenous malformation and venous angioma.

In addition to cerebrovascular disease, hypertension can also be linked to seizures. Data from one case-control study suggest that hypertension alone, minus stroke, was a significant risk factor for unprovoked seizures [odds ratio (OR) = 2.02]. In patients with a history of hypertension, a previously occurring stroke was a seizure risk factor (adjusted OR = 4.81).6 However, many individuals with hypertension alone have had subclinical strokes.

Co-morbidity and polypharmacy.

The management of seizures in patients over 60 years of age is challenging because of the common presence of co-morbid conditions that require various medications. Because the elderly population possesses multiple risk factors, those who are at risk and experience a single seizure have an 80% chance of experiencing a second or third seizure. Most of the patients in VACS #428 had overt infarctions or arterial involvement. Sixty-four percent of patients presented clinically with hypertension. Previously reported evidence suggests that the chance of developing seizures increases fivefold in hypertensive patients.7 Approximately 50% of the patients in VACS #428 had experienced a stroke; half of these presented with symptomatic cardiac disease.1 Less than one-third of patients were diabetic and less than one-fourth had a history of cancer.2 Prompt treatment of one seizure in an older person will minimize recurrent episodes.1

Assessment of patients’ cognitive function is paramount before treatment for seizures is initiated. Physicians should not assume, on the basis of casual history taking, that patients possess normal cognitive function.2 Cognitive function was evaluated in a subset of patients (N = 377) in VACS #428. Patients who presented with Alzheimer’s disease were excluded from the study because of the impact of its rate of progression on the ability to control seizures. Thirty-six percent of individuals were identified as cognitively impaired. Other indications that affected cognitive ability in these patients included multiple infarctions (10.9%), alcohol-related conditions (3.4%), Parkinson’s disease (PD; 1.1%), and normal-pressure hydrocephalus (0.002%). A moderate proportion of patients had either other (9.5%) or unknown (14.1%) conditions that affected their cognitive ability.2

Because elderly persons often experience many co-morbid conditions, they are often exposed to multiple medications. The mean number of co-medications reported in this population is 6.7 (range 0–15).2 In fact, approximately 25% of the patients in VACS #428 reported taking as many as 15 prescription medications.2

Imaging findings.

In a subanalysis of 123 elderly patients enrolled in VACS #428, CT and MRI revealed that 18% of patients were considered normal via CT scans, 42.6% of patients manifested a CVA, and 40.9% displayed small-vessel disease. It is debatable whether those patients categorized as having normal CT findings would have had abnormal findings on MRI.2 Diffuse atrophy was identified in 35% of patients. A few individuals had encephalomalacia (9.1%), benign tumors (1.5%), and normal-pressure hydrocephalus (0.75%).1

Patients in VACS #428 underwent EEG study. Few patients (31.3%) exhibited normal EEG findings, and many exhibited nonspecific abnormalities. The presence of focal abnormalities did not aid in the diagnosis. In a subset analysis of 23 patients who had either a normal routine EEG (N = 9) or abnormal slowing noted in their routine EEGs (N = 14) and for whom extended EEG recordings were obtained, 12 patients were found to have epileptiform activity on the extended recordings.1,8⇓ Therefore, extended/ambulatory EEGs in elderly patients with a questionable diagnosis of epilepsy increase the yield of recorded diagnostic epileptiform activity by 50%.

Serology.

In a subset of VACS #428 patients (N = 144), less than 20% of the patients had normal lipid profiles. The low-density lipoprotein (LDL) level was normal (less than 80 mg/dL) in only 17% of patients. This low percentage of patients with normal lipid serology may be linked to the presence of cardiac or cerebrovascular diseases or to abnormal MRI findings owing to overt strokes or cerebrovascular abnormalities.1,2⇓

Clinical manifestations.

The classic features of epilepsy vary with respect to age. Elderly patients with epilepsy may not present with the typical features of complex partial seizures (CPS) but may be shown to possess an altered mental status, memory lapses, episodes of confusion, or syncope. Data reported from the VACS #428 suggest that approximately one-third (38.3%) of patients experienced CPS.1 Although the majority of CPS are of temporal lobe origin, seizures in older persons are more often extratemporal, usually confined to the frontal lobe. The variation in the clinical presentation of seizures in elderly patients is linked to the anterior frontal cortical areas of the brain that are affected in stroke patients.2

Primary generalized epilepsies that are initiated in the early years may re-emerge in older individuals and require different treatment from that prescribed for patients with partial-onset seizures. Five groups of general and/or partial seizure types were identified in VACS #428. They were classified as shown in figure 2. 1

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Frequency of seizure types in patients 60+ years old.1 CPS = complex partial seizure; GTC = generalized tonic–clonic seizure; SPS = simple partial seizure.

The clinical presentation of epilepsy in the elderly differs from that characteristic of younger patients. Auras are less common and, when they occur, are often nonspecific (e.g., dizziness); automatisms are less frequent; postictal confusion may be prolonged; and the patient’s presenting complaint may be altered mental status, confusion, or memory disturbance.4

The absence of specific symptoms may undermine an epilepsy diagnosis. For at least half of all patients who present with a symptom that ultimately was classified as a seizure, epilepsy was not a strong consideration at initial diagnosis.2 Data from VACS #428 reveal that epilepsy was a diagnostic consideration after initial evaluation by a primary care physician or internist for only 73.3% of patients who ultimately were diagnosed with epilepsy. Initial diagnoses included altered mental status (41.8%), confusion (37.5%), blackout spells (29.3%), memory disturbance (17.2%), syncope (16.8%), dizziness (10.3%), and dementia (6.9%). (A patient could have more than one initial diagnosis.) Because of the lack of specific clinical signs in a significant number of older patients with new-onset epilepsy, the disease is not identified initially.4 Consequently, patients who are treated with dementia drugs on initial diagnosis fail to improve because the seizures may become more prominent if left untreated by antiepileptic drugs (AEDs).2

Case study.

This case study is a prime example of how cognitive assessment may aid in proper diagnosis and treatment of seizures. A 72-year-old woman who lived alone was admitted for dementia. She reported experiencing periodic episodes of being “fuzzy” and had no significant past medical history. On close inspection, her EEG revealed a dynamic change from a buildup of polymorphic delta activity followed by an abrupt change to a normal recording. Although this EEG recording appears to be depicting epilepsy, it does not have the morphology of it.2 Other than the atrophy seen on the CT scan, her clinical evaluation was uneventful. However, after a video evaluation of the patient was viewed, what appeared initially to be dementia was later considered to represent seizures. The point is that physicians should consider seizures in their diagnoses in order to eliminate future episodes.2

As a result of experience with patients such as this whose diagnosis of seizures may have been delayed, VACS #428 analyzed the time to delayed of diagnosis in a subset (N = 128) of the patient population. Patients (66.7%) who presented with a tonic–clonic seizure were almost always accurately diagnosed. On the other hand, only a small number of patients (25.4%) experiencing simple partial seizures received a correct diagnosis.2,9⇓ In patients who experienced a transient ischemic attack (TIA) or stroke, the diagnosis of seizures was usually delayed.2 The time to diagnosis measured in this cohort was 1.7 years.9

Differential diagnoses.

A co-morbid disease may often be mistakenly identified as the cause of symptoms in an elderly patient with new-onset epilepsy. Co-morbid diseases include hypertension, dementias, TIA, and cardiac disease. Moreover, the uncommon occurrence of tonic–clonic seizures renders a diagnosis more challenging. Language disturbances or confusion may cause difficulty in obtaining an accurate patient history. Therefore, although a patient may exhibit symptoms of a TIA, seizures would be a more accurate diagnosis in some cases.

Although neurologic symptoms are associated with both TIA and seizures, their impact differs. Vascular syndromes, such as Wernicke’s aphasia, Korsakoff’s psychosis, pure word deafness, and nondominant parietal syndrome, are identified in some patients who have experienced a TIA. Although a TIA can cause focal neurologic symptoms, it does not cause global confusion. Seizures can cause global confusion in an ictal or postictal manner. Moreover, migraines, drugs, sleep disorders, some metabolic disturbances (such as porphyria), and transient global amnesia all cause confusion. The list of neurologic disorders that lead to confusion is relatively specific, and TIA is not included in this list.2

Results.

Once elderly patients are diagnosed with epilepsy and can tolerate a drug, the outcome is reportedly favorable. Data from VACS #428 suggest that at 3, 6, and 12 months, 60% of patients remain seizure-free.1 Twenty percent of patients were intolerant of the first drug over the first 4, 5, or 6 weeks.2 A primary limitation of these results is patient withdrawal from the study because of side effects. At 52 weeks, only 47.8% of patients enrolled in VACS #428 had completed the study. Reasons for early termination included side effects (19.4%), voluntary withdrawal (10%), death (7.2%), uncontrolled seizures (3.3%), loss to follow-up (2.8%), noncompliance (1.7%), and other (3.9%).1

Data on the plasma levels of each drug position lamotrigine as superior to gabapentin and carbamazepine for controlling epilepsy in elderly patients. Mean plasma levels at 3 months reveal that lamotrigine (3.27 μg/mL) produced complete control of seizures in patients who were tolerant of the drug. Gabapentin plasma levels (8.97 μg/mL) were acceptable. Plasma levels in patients receiving carbamazepine (6.97 μg/mL) were as expected.1,2⇓

Adverse effects of AEDs are cause for concern in the elderly population. Dose-dependent adverse effects (e.g., dizziness, lethargy, unsteady gait, and visual disturbance) and drug-specific adverse effects (e.g., hyponatremia, tremor, cardiotoxicity, ataxia, encephalopathies, and neuropathies) (table 1) are common and can occur at lower blood levels than in younger patients. Osteopenia and osteoporosis are common in patients who have used older drugs such as carbamazepine and valproate. Therefore, bone health is a major issue to be dealt with when a drug regimen is designed. Older antiepileptic drugs include phenytoin, carbamazepine, valproate, phenobarbital, and primidone. Gait impairment, sedation, and tremor are age-related adverse effects, particularly of the older AEDs.10

Drug-specific adverse effects are also associated with new-generation AEDs. In general, newer established agents may cause weight gain and pedal edema. Lamotrigine has been reported to cause rash in a small number of patients. However, a slow titration schedule minimizes the risk for serious rash. Word-finding difficulties were expected in patients treated with topiramate. However, in an ongoing study with a patient subpopulation of individuals 60 years or older, only one patient experienced this and it resolved with continued treatment.11 Single-agent topiramate has caused renal stones and paresthesias (tables 1 and 2⇓). 12–14⇓⇓

Emerging newer AEDs (e.g., tiagabine, levetiracetam, oxcarbazepine, and zonisamide) also cause drug-specific side effects (tables 1 and 2⇑). Encephalopathy and knee-buckling are associated with tiagabine. Levetiracetam causes behavioral changes in some patients. Oxcarbazepine is linked to hyponatremia. Although patients are reportedly able to tolerate fairly high doses of zonisamide (average dose up to 600 mg), renal stones, weight loss, and paresthesias can be adverse effects associated with its use.2,12⇓ In any case, older patients may not exhibit a spectrum of side effects identical to that seen in younger patients.

Selection of antiepileptic drugs.

Benefits (table 3) of using older established AEDs, such as carbamazepine, phenytoin, and valproate, include their efficacy in partial-onset seizures, extensive worldwide experience, and multiple dosage forms (e.g., intravenous, intramuscular). On the other hand, the older drugs are associated with a higher rate of drug-specific adverse effects, bidirectional drug interactions, and nonlinear kinetics that make them difficult to use.2

On the basis of a growing body of data from several clinical trials, newer established AEDs should be considered as first-line therapy for elderly patients who experience seizures.4,11,15⇓⇓ Advantages (table 2) to the use of new-generation AEDs such as gabapentin, lamotrigine, or topiramate are efficacy in partial-onset seizures, expanding dosage forms, linear or predictable kinetics, minimal and unidirectional drug interactions, and efficacy in conditions other than epilepsy. However, although these drugs possess an incompletely defined side-effect profile, they are known to cause drug-specific adverse effects.2

In a multicenter, double-blind, parallel-group randomized study including 150 elderly patients (mean age 77 years) with newly diagnosed epilepsy,15 lamotrigine was superior to carbamazepine on the basis of patient completion, ability to minimize seizures, fewer side effects, and better patient outcome. At 24 weeks, 71% of patients taking lamotrigine completed the study compared with approximately 42% of those taking carbamazepine. About 39% of patients in the lamotrigine arm remained seizure-free compared with 21% of those receiving carbamazepine. Adverse effects reported in the lamotrigine (N = 102) and carbamazepine (N = 48) arms of the study included somnolence, rash, poor coordination, dizziness, and headache. Fewer patients taking lamotrigine withdrew because of adverse events.15 In short, newer-generation AEDs are effective at relatively low doses and are well tolerated in elderly patients.

An ongoing study is being conducted with topiramate. Patients are included in the study if they are 60 years or older, have had at least two seizures in the past year, with the second seizure occurring within the previous 3 months, are considered candidates for alternative antiepileptic treatment by their primary physician, and have not had previous topiramate treatment. Patients are excluded from the study if they possess an unstable medical condition, a significant psychiatric problem, or have a history of noncompliance.11 The incidence of adverse effects associated with topiramate in this patient population has been minimal; the most common is somnolence (13%). Therefore, topiramate appears to be well tolerated in elderly patients.11

Newer emerging AEDs (e.g., levetiracetam, oxcarbazepine, tiagabine, and zonisamide) may eventually prove superior to newly established agents. As with older and newly established AEDs, they are effective against partial-onset seizures. They also possess linear kinetics and minimal or no drug interactions, like their established new-generation counterparts. Conversely, they have had limited use in the clinic, are available in limited dosage forms, and have drug-specific adverse effects with a still undefined side-effect profile. According to the International Center for Epilepsy, at present 85.5% of patients from ages <40 to >60 years are continuing on levetiracetam therapy.2

Conclusions.

The incidence of epilepsy is highest in persons over 60 years old. Because of the variation in seizure symptomatology from younger to older patients, novel diagnostic paradigms must be introduced into regular practice to establish an accurate epilepsy diagnosis. Tolerance is the therapy-limiting factor for which newer drugs may be superior to older-generation AEDs. The key to minimizing side effects is to start at a low dose and titrate up to a higher dose, to achieve efficacy while maintaining safety. Anticipation of drug interactions and adverse effects will limit their occurrence. A further advantage of newer-generation drugs is that they exhibit few drug interactions, unlike their older-generation counterparts. In light of the data reviewed in this article, the use of newer established drugs for elderly patients with epilepsy is recommended over the first-line, older-generation drugs. The data presented are for the purpose of educating patients, physicians, and health-care providers about the diagnostic challenges present in elderly patients with seizures and the benefits of newer established AEDs, in order to optimize therapy and ultimately minimize the occurrence of seizures.

Discussion

Question: A question for Gene about seizures in the elderly. My clinical experience doesn’t agree with the incidence of partial complex seizures in this age range. It seems that it would be a lot less, and we see a lot more tonic–clonic and generalized seizures, and not so many partial complex. Do you think we’re just dense, and missing them?

RER: I think they’re going to the internist and the internist is missing them. I’m taking the onus off of you. But I think that the internists are seeing the patients and that they’re misdiagnosing.

Question: How do you screen your patients into the VA cooperative study? How do you acquire those patients?

RER: We had a full-time coordinator. Initially, we were waiting for the patients to come to us. Later on, I told them to go to the patient lists that were being admitted—go to the Medicine ward. Start educating the PAs on the Medicine service that we are interested in some of these patients. The first woman that I showed you had been in the ward for 4 days. She was in complex partial status and the internists had no clue that she was having seizures. The internists just did not know. We got a lot of referrals because the nurses were concerned, not because the physicians had any clue. And I think that only by good education are we going to get some of these patients referred to us.