Ilo E. Leppik MD*
James C. Cloyd PharmD**
* Director of Research, MINCEP Epilepsy Care, Minneapolis, Minnesota
**: Professor and Director, Epilepsy Research and Education Program, College of Pharmacy, University of Minnesota, Minneapolis, Minnesota
The elderly comprise the most rapidly growing segment of the population, and onset of epilepsy is higher in this age group than in any other. The incidence of a first seizure is 52 to 59 per 100,000 in persons 40 to 59 years of age, but it rises to 127 per 100,000 in those ≥60 years old (1). Among persons ≥65 years, the active epilepsy prevalence rate is approximately 1.5%, about twice the rate of younger adults. In 1995, approximately 2.3 million residents of the United States had been diagnosed with epilepsy: 1.4 million were adults aged 15 to 64 years, 300,000 were children aged ≥14 years, and 550,000 were persons >65 years of age. Approximately 181,000 persons developed epilepsy in 1995, and approximately 68,000 of these were >65 years old (2). As the elderly population continues to grow steadily, increasing numbers of older persons are likely to require accurate diagnosis and effective treatment. The foregoing data are based mostly on a community-based population. The prevalence of epilepsy and antiepileptic drug (AED) use is much higher in nursing homes. In a review of 45,405 people ≥65 years old who were living throughout the United States in long-term care facilities serviced by Pharmacy Corporation of America, at least one AED was taken by 4,573 (10.1%) of the residents (3). Later studies have confirmed that the prevalence of AED use in the nursing home population varies between 10% and 11% (4, 5, 6). Approximately 1.5 million elderly people reside in nursing homes; thus, as many as 150,000 elderly nursing home patients may be taking AEDs.
In the elderly, the most common identifiable cause of epilepsy is stroke, which accounts for 30% to 40% of all cases (7). Brain tumor, head injury, and Alzheimer's disease are other major causes. However, in many patients, the precise cause cannot be identified. Assessment of AED treatment efficacy and toxicity in elderly patients is challenging because seizures are sometimes difficult to observe, signs and symptoms of toxicity can be attributed to other causes (e.g., Alzheimer's disease, stroke) or to comedications, and the patient may not be able to self-report problems accurately.
ANTIEPILEPTIC DRUG USE IN THE ELDERLY
In addition to their use in epilepsy, AEDs are prescribed for various other disorders, including neuralgias, aggressive behavior disorders, essential tremor, and restless legs syndrome, conditions prevalent in the elderly. Treatment of older patients with AEDs, as with many other medications, is complicated by increased sensitivity to drug effects, narrow therapeutic ranges, complex pharmacokinetics, and the increased likelihood of drug interactions because of multiple drug therapy (8, 9, 10, 11). Elderly persons also have a high probability of concomitant disorders. Use of antipsychotic therapy may also increase seizures (12,13). As a cause of adverse reactions in the elderly, AEDs rank fifth among all drug categories (14). Nonetheless, the frequent use of AEDs by nursing home residents and the growing number of elderly persons in the general population suggest that hundreds of thousands of older persons are being treated with these medications.
Phenytoin (PHT) is the most commonly used AED in nursing homes. In a 1995 study of 21,551 nursing home residents in 24 states on one day in the spring of 1995, 10.5% had an AED order (4). Of these patients, 9.2% had a seizure or epilepsy indication recorded. Of the AEDs, 6.2% of these patients were taking PHT, 1.8% used carbamazepine, 0.9% used valproic acid, 1.7% used phenobarbital, and all other AEDs combined, 1.2% (Table 13.1). The diagnosis of epilepsy is generally made only after a person has had two or more seizures. However, many physicians may begin treatment with AEDs after a single seizure in elderly patients because the risks of a second seizure are perceived to be high. Treatment in the elderly carries more risks
than in younger persons because elderly persons may experience more side effects, they have a greater risk of drug interactions, and they may be less able to afford the costs of medications. Thus, both the benefits and risks of treatment may be greater in the elderly.
TABLE 13.1. MOST COMMONLY USED ANTIEPILEPTIC DRUGS IN A NURSING HOME POPULATION (N = 21,551)
At the present time, data are limited regarding the clinical use of AEDs in the elderly. The paucity of information makes it difficult to recommend specific AEDs with any confidence that the outcomes will be optimal. Nevertheless, decisions need to be made, and indeed they are being made. Many of the recommendations will be modified as new knowledge is obtained, and at the present time, the “comfort level” with some drugs may play a larger role than actual experience or data.
Elderly persons differ in many respects from younger adults, and simply using information applicable to the younger age group will not lead to the best outcome (Table 13.2). Elderly persons represent a much more heterogeneous population, but they may, for purposes of simplification, be considered to consist of the elderly healthy (EH), except for epilepsy, and the elderly with multiple medical problems (EMMP). A drug choice optimal for one group may not be appropriate for the others. Even the EH will have a decline in functioning of the various organ systems, leading to lower hepatic and renal clearance of AEDs and a possible increase in sensitivity of the central nervous system (CNS) to side effects. In addition, the cost of AEDs is an important factor for many patients. These issues are greatly complicated for the EMMP population. In addition, the state of health many change rapidly, so a choice appropriate for one period may not be acceptable later. Each available drug is discussed in terms of the benefits and risks for the EH and EMMP populations (Table 13.3).
TABLE 13.2. PHARMACOKINETICS OF ANTIEPILEPTIC DRUGS IN ELDERLY AND YOUNGER ADULTS
TABLE 13.3. CHOOSING ANTIEPILEPTIC DRUGS FOR THE ELDERLY HEALTH AND THE ELDERLY WITH MULTIPLE MEDICAL PROBLEMS
Effect of Advancing Age on Pharmacodynamics and Pharmacokinetics
Age-related changes in physiology result in clinically significant alterations in both drug response and drug disposition. Changes in receptor number and sensitivity, alterations in cellular biochemistry, and the nature of the disorder can affect both pharmacologic and toxicologic effects (15, 16, 17). Elderly patients taking carbamazepine or valproic acid for epilepsy realize optimal seizure control, but they are also more likely to experience adverse events, at lower plasma concentrations than younger patients (18). Castleden et al. (19) have shown that elderly persons exhibit greater cognitive impairment after a dose of nitrazepam than do younger patients, although plasma concentrations are similar.
Age-related changes in physiology can alter all aspects of pharmacokinetics: absorption, distribution, metabolism, and elimination. Advanced age is associated with increased gastric pH, diminished gastrointestinal fluids, and slower
intestinal transit, and reduced absorptive area. Each of these changes can affect either or both the rate and extent of absorption. Age-related reduction in intestinal and hepatic blood flow, intestinal drug transport and metabolism, and hepatic metabolism can also affect the systemic bioavailability of some drugs. Gastric pH and intestinal transit time may exhibit intrapatient day-to-day variability, whereas other processes tend to decline slowly. Age-related alterations in absorption are most likely to affect slowly absorbed AEDs, particularly those administered as solid dosage forms, extended release formulations, or drugs absorbed by active transport (gabapentin). Depending on the process or processes affected, alterations in gastrointestinal physiology can either increase or decrease bioavailability, resulting in loss of seizure control or onset of side effects.
Total serum drug concentration reflects both drug bound to serum proteins and unbound drug. For most drugs, unbound concentration in serum is in direct equilibrium with the concentration at the site of action. Because drug concentration at the site of action determines the magnitude of both desired and toxic responses, unbound drug in serum provides the best correlation with drug response (20). Total serum drug concentration is useful for monitoring therapy when the drug is not highly protein bound (less than 75%) or when the ratio of unbound to total drug concentration remains relatively stable. This is not the case for highly bound AEDs such as carbamazepine, PHT, tiagabine, and valproic acid; these drugs frequently undergo age-related alterations in protein binding.
Older persons experience a gradual reduction in serum albumin and increased α1-acid glycoprotein (AAG) concentrations (20, 21, 22). By age 65 years, many persons have low normal albumin concentrations or are frankly hypoalbuminemic (23). Albumin concentration may be further reduced by conditions such as malnutrition, renal insufficiency, and rheumatoid arthritis. As serum albumin levels decline, the likelihood increases that drug binding will decrease. This has the effect of lowering the total serum drug concentration while unbound serum drug concentration remains unchanged. The concentration of AAG, a reactant serum protein, increases with age; further elevations occur during pathophysiologic stress such as stroke, heart failure, trauma, infection, myocardial infarction, surgery, and chronic obstructive pulmonary disease (22). Administration of enzyme-inducing AEDs also increases AAG (24). When the concentration of AAG rises, the binding of weakly alkaline and neutral drugs such as carbamazepine (and its epoxide metabolite) to AAG can increase, thereby causing higher total serum drug and metabolite concentrations.
Thus, measurements of total concentrations of highly bound AEDs in the elderly are misleading. For example, an elderly person with a low serum albumin level could have a PHT free fraction of 20% on the basis of reduced binding, rather than the usual 10% seen in younger adults. In such a patient, a total PHT level of 20 mg/L with a 20% free fraction corresponds to an unbound concentration of 4.0 mg/L; the same 4.0 mg/L of unbound concentration in a younger person with normal binding and clearance would be associated with a total drug level of 40 mg/L. Not recognizing this situation may lead to inappropriate clinical decisions.
The age-related physiologic changes having the greatest effect on pharmacokinetics are reduction in liver mass resulting in decreased drug metabolizing capacity and diminished kidney function (25, 26, 27, 28). Hepatic drug metabolism declines approximately 10% per decade beginning at age 40 years (27,28). A decrease in drug-metabolizing capacity results in a slower unbound clearance that, in turn, will cause an increase in unbound and, if there is no change in protein binding, total drug concentrations. Studies to date have included an insufficient number of people >85 years of age to know whether the decrease in drug metabolism continues in the oldest old. The effect of advancing age on enzyme induction has not been well characterized.
Renal function, as measured by glomerular filtration rate, also decreases by approximately 10% with each decade of age, beginning at age 40 years (26). Creatinine clearance is a reliable marker of glomerular filtration and correlates well with unbound renal clearance of drugs eliminated by the kidneys. Elderly persons eliminate renally cleared drugs and active metabolites more slowly than do younger adults.
The total clearance of the major AEDs, which are predominately eliminated by the liver, is influenced by both the extent of protein binding and the intrinsic metabolizing capacity (intrinsic clearance) of unbound drug. Because total clearance determines steady-state total drug concentration, age-related alterations in protein binding or intrinsic clearance can affect serum drug concentrations. Age-related reductions in intrinsic clearance cause a rise in both unbound and, if there is no simultaneous decrease in protein binding, serum drug concentrations.
If an elderly patient has a decrease in both hepatic drug metabolizing capacity and protein binding, the effect on total and unbound drug will depend on the relative magnitude of change for each parameter. If the decrease in protein binding is greater than the reduction in unbound clearance, total drug concentration can remain unchanged or can even decline in the presence of rising unbound drug concentration. Measurement of unbound concentration in elderly patients is essential when altered protein binding is suspected or when response (either therapeutic or toxic) does not correlate with total drug concentration.
Despite the effects of age-related physiologic changes on pharmacodynamics and drug disposition and the widespread use of AEDs in the elderly, few studies on AED pharmacokinetics, drug interaction, efficacy, or safety in the elderly have been published (28, 29, 30, 31). The available reports generally involve single-dose evaluations in small samples of
the young old, that is, persons 65 to 74 years old. The absence of data on AED pharmacokinetics in the oldest old increases the possibility of therapeutic failure and adverse reactions in this population (32). (A summary of available pharmacokinetic information in the elderly is presented in Table 13.4.)
TABLE 13.4. ROUTES OF ANTIEPILEPTIC DRUG ELIMINATION IN THE ELDERLY
Older patients appear to be more sensitive to the CNS and systemic adverse effects of AEDs, especially cognition (33,34). In one study, PHT was the only drug among several factors that was associated with a significant increase in nonvertebral fractures among community-dwelling elderly women (34). This study also found that both elderly men and women taking moderate doses of benzodiazepines, including clonazepam, had a greater likelihood of hip fractures (34). These reports indicate that elderly persons, as a group, are more sensitive to both the pharmacodynamic effects and the toxic effects of AEDs, although there is substantial intrapatient variability.
PHT is poorly soluble and slowly absorbed, reaching peak concentrations after a single dose ranging from 6 to 18 hours. Age-related alterations in gastrointestinal function could have an impact on PHT absorption. Preliminary results from our studies indicate that oral PHT bioavailability in elderly patients ranges from 40% to 100%. Other investigators have observed fluctuations ranging from 50% to 150% in PHT concentrations among 15 frail nursing home patients while receiving maintenance therapy with the same daily dose and formulations. These reports suggest that the elderly may be susceptible to intrapatient alterations in PHT bioavailability resulting in clinical significant changes in drug concentration.
Studies in elderly patients have shown decreases in PHT binding to albumin and increases in the free fraction (35,36). The binding of PHT to serum proteins correlates with the albumin concentration, which is typically low normal to subnormal in the elderly. As the drug concentration rises and the albumin concentration falls, PHT binding is likely to decrease.
PHT elimination in the elderly is reduced. One study compared the pharmacokinetics of PHT at steady state after oral administration in 34 elderly (60 to 79 years), 32 middle-aged (40 to 59 years), and 26 younger adults (20 to 39 years) with epilepsy. All subjects had normal albumin concentrations and liver function and received no other medications, including other AEDs known to alter hepatic metabolism. The maximum rate of metabolism (Vmax) declined gradually with age; the elderly group had a mean Vmax that was 20% smaller than in the younger adults (37). Other smaller studies have also shown that PHT metabolism is reduced in the elderly (25,38, 39, 40, 41). The smaller Vmax means that PHT metabolism becomes saturated at lower concentrations than in younger patients. Thus, smaller maintenance doses of PHT are needed to attain desired
unbound serum concentrations, and relatively small changes in dose (≤10%) are recommended when making dosing adjustments. Thus, in the elderly, a daily dose of 3 mg/kg appears to be appropriate, rather than the 5 mg/kg per day used in younger adults (42). This 3 mg/kg dose is only 160 mg/day for a 52-kg woman or 200 mg/day for a 66-kg man. The reduced elimination also results in extended half-lives. Most elderly patients with total PHT concentrations ≥10 mg/L have PHT half-lives of >40 hours. Personal observation (Cloyd and Leppik) would indicate that elderly patients can be safely given PHT doses once a day with minimal fluctuation in drug concentrations and improved compliance.
One nursing home survey revealed that residents were taking PHT doses similar to those used in younger adults (3). Thus, there is a great potential for inadvertent overdose in the nursing home population. Because of protein binding, the total levels in these persons may appear to be normal, but measurement of the unbound level may be necessary to detect overdoses (43,44). In patients with both reduced metabolism and binding to serum albumin, unbound PHT concentration increases while the total drug concentration decreases. In such cases, the total drug concentration does not correlate with response. Patients may achieve seizure control with what is thought to be subtherapeutic concentrations, or they may experience toxicity when total serum concentrations are in the therapeutic range. Measurement of unbound PHT concentrations is necessary for elderly patients who have the following: (a) decreased serum albumin concentration or total PHT concentrations that are near the upper boundary of the therapeutic range; (b) total concentrations that decline over time; (c) a low total concentration relative to the daily dose; or (d) total concentrations that do not correlate with clinical response. A range of 5 to 15 mg/L may be more appropriate as a therapeutic range for the elderly (42).
PHT is effective for localization-related epilepsies, and thus it has an efficacy profile appropriate for the elderly. However, no studies regarding the effectiveness of PHT in the elderly have been published. Some evidence from the Veterans Administration cooperative study would suggest it is equally effective as carbamazepine, phenobarbital, and primidone, but the number of subjects in that study was small. PHT has drug-drug interactions, and it needs to be used cautiously in EMMP patients receiving other medications.
PHT does have some effects on cognitive functioning, especially at higher levels (33). It is not known whether the elderly will be more sensitive to this problem. In addition, PHT may cause imbalance and ataxia. It is likely that EMMP patients, especially those with CNS disorders, may be more sensitive to these effects. One study investigated the risk factors for nonvertebral fractures in elderly, community-dwelling elderly women. Among the various lifestyle, demographic, and health factors that contributed to an increase risk, PHT, despite relatively low usage, was the only drug with a significant effect (34). PHT also is known to be a mild blocker of cardiac conduction, and it should be used cautiously in persons with cardiac conduction defects, especially heart blocks. Treatment with antineoplastic drugs may decrease PHT concentrations (45). Drug interactions with PHT are a major problem in treating the elderly (8,9).
Young adults typically require 10 to 20 mg/kg/day taken in three or four divided doses to attain serum carbamazepine concentrations within the usual therapeutic range (46). Carbamazepine doses were much lower in one nursing home study, whereas trough serum carbamazepine concentrations remained within the usual therapeutic range (5). One study evaluated carbamazepine clearance in seven elderly (mean age, 82.3 years) using daily dose and trough serum concentrations. Clearance was 40% lower than in a group of younger patients 41.0±19.6 versus 71.4±35.8 mL/hr/kg, respectively (3). This decrease is the same magnitude as seen with PHT and valproic acid in elderly patients. The smaller clearance results in a prolonged elimination half-life. These changes in carbamazepine pharmacokinetics require lower doses and less frequent dosing in elderly patients.
Carbamazepine is effective for localization-related epilepsies, and thus it has an efficacy profile appropriate for the elderly. However, no studies regarding the ability of carbamazepine in the elderly have been published. Some evidence from the VAH cooperative study would suggest that it is equally effective as PHT, phenobarbital, and primidone, but the number of subjects in the study was small. There seems to be a reduction of the drug's clearance with advancing age, so doses will need to be lower in both EH and EMMP patients. What is not known is whether its metabolism will be autoinducible to the same extent that it is in younger populations. Thus, one will need to monitor drug levels after initiation of treatment and adjust doses accordingly.
Carbamazepine has some significant drug-drug and drug-food interactions with medications that inhibit the cytochrome P450 enzyme, CYP3A4, responsible for carbamazepine metabolism. Among the inhibitors are erythromycin, fluoxetine, ketoconazole, propoxyphene (Darvon), and grapefruit juice. EH patients will need to be cautioned about the use of these medications. Many other drug interactions occur, so carbamazepine is one AED that must be used cautiously in EMMP patients receiving other medications.
Carbamazepine at higher levels does have some effects on cognitive functioning. It is not known whether the elderly will be more sensitive to this problem. In addition, carbamazepine may cause imbalance and ataxia. It is likely
that EMMP patients, especially those with CNS disorders, may be more sensitive to these effects.
One of the major concerns with carbamazepine is its effect on sodium levels. Hyponatremia is a well-known phenomenon with carbamazepine use, and it may cause significant problems in younger adults, especially in the presence of polydipsia. Although little is known about carbamazepine and sodium balance in the elderly, this could well be a major issue, especially if patients are receiving diuretics or are on salt-restricted diets. This issue needs to be investigated further. Carbamazepine also is known to affect cardiac rhythms, and it should be used cautiously, if at all, in persons with rhythm disturbances.
One of the pharmacokinetic problems of carbamazepine is its short half-life, associated with the possible need to take the drug multiple times a day. In the elderly, however, the drug's half-life may be longer. In any case, the new slow-release formulations (Carbatrol and Tegretol XR) have overcome these limitations. Carbamazepine is a moderately priced drug, and it should not present a significant cost issue, especially if lower doses are needed.
Phenobarbital is effective for localization-related epilepsies, and thus it has an efficacy profile appropriate for the elderly (47). However, the VAH cooperative study demonstrated that phenobarbital and primidone have effects on cognitive functioning, most prominent at higher levels (48). The elderly will be more sensitive to this problem. Thus, although phenobarbital is the least expensive of all the AEDs, its effects on cognition and mood make this an undesirable drug for the elderly, both the EH, who are trying to maintain independent living conditions, and EMMP patients, who may have underlying intellectual deficits.
Only a few small studies have compared the pharmacokinetics of valproic acid in young and old patients (30,49,50). In a study of steady-state valproate pharmacokinetics in six young adult and six elderly volunteers (66 to 72 years), the average unbound fraction of valproate was 10.7% in the elderly compared with 6.4% in younger subjects. In elderly subjects, mean unbound concentration was 57% higher and unbound clearance was 65% lower than in younger adults (49). In another study comparing single-dose intravenous valproate pharmacokinetics in seven young adult volunteers and in six residents of long-term care units (75 to 87 years), total clearance was similar in the two groups. Serum elimination half-life was twice as long in the elderly as in the younger subjects, 14.9 versus 7.2 hours (50). Valproic acid, like PHT, is associated with reduced protein binding and unbound clearance in the elderly. As a result, the desired clinical response may be achieved with a lower dose than usual. Because the serum elimination half-life is prolonged, the dosing interval can be extended. If the albumin concentration has fallen or if the patient's clinical response does not correlate with total drug concentration, measurement of unbound drug should be considered.
Felbamate is effective for localization-related epilepsies, and it appears to have a broader spectrum of effectiveness than some of the other AEDs. Thus, it has an efficacy profile that may be very appropriate for the elderly. Felbamate is primarily metabolized by the liver and is known to have certain drug-drug interactions, both inhibitory and inductive (51). Thus, it does not appear to be a drug that will be easy to use in EMMP patients. Because of its association with aplastic anemia and hepatic failure, felbamate has a higher risk profile than the other AEDs and may have limited use in the elderly.
Gabapentin is effective for localization-related epilepsies, and thus it has an efficacy profile appropriate for the elderly. Gabapentin is not metabolized by the liver, but rather it is renally excreted (52). Creatinine clearance is reduced with advancing age, so doses will need to be lower in both EH and EMMP patients. Thus, one will need to monitor drug levels after initiation of treatment and adjust doses accordingly. Because gabapentin has no drug-drug interactions, it may be especially useful in EMMP patients. Gabapentin may have some cognitive side effects, especially at higher levels, and the elderly may be more sensitive to this problem.
One of the problems with gabapentin is its short half-life, associated with the possible need to take the drug multiple times a day. In the elderly, however, the drug's half-life may be longer. Gabapentin is a high-priced drug, and it may present a significant cost issue, especially if higher doses are needed. A study comparing carbamazepine with gabapentin and lamotrigine is in progress, and more information regarding the various safety and efficacy issues will be available when the VAH study is completed.
Because lamotrigine is effective for localization-related epilepsies, it has an efficacy profile appropriate for the elderly. One study comparing it with carbamazepine was favorable (53). Lamotrigine is primarily metabolized by the liver by glucuronidation (54). It is not known whether there is a reduction of its glucuronidation with advancing age. Thus, one will need to monitor drug levels after initiation of treatment and adjust doses accordingly.
Lamotrigine elimination is reduced by drugs such as valproic acid, which block glucuronidation. Thus, some caution may need to be observed in EMMP patients who are taking other drugs.
Tiagabine is effective for localization-related epilepsies and has an efficacy profile appropriate for the elderly. It is primarily metabolized by the liver. There may be a reduction of its clearance with advancing age, but this may be overshadowed by the susceptibility to increased metabolizing by inducing drugs. Experience with this drug in elderly is limited.
Topiramate has an efficacy profile appropriate for the elderly. Topiramate is both metabolized by the liver and excreted unchanged in the urine. There may be a reduction of its clearance with advancing age, so levels will need to be monitored. Topiramate does have some effects on cognitive functioning, especially word finding. The elderly may be more sensitive to this problem.
Benzodiazepines used for the treatment of epilepsy include diazepam, lorazepam, clorazepate, and clonazepam. Diazepam and lorazepam are administered intravenously for the acute treatment of status epilepticus, and clorazepate and clonazepam are given orally as maintenance therapy.
Diazepam is highly protein bound (>99%) and undergoes oxidative metabolism to form an active metabolite, desmethyldiazepam. Protein binding declines with age, resulting in an increased free fraction and a greater distribution volume of diazepam and desmethyldiazepam. Unbound clearance is reduced, thus prolonging the serum elimination half-life of the drug and its metabolite. Lorazepam is less highly bound (90%) and is metabolized by conjugation to lorazepam glucuronide. The free fraction of lorazepam rises with age, and the volume of distribution is increased, but less than with diazepam. The elimination half-life of lorazepam is similar in the young and the elderly. Direct comparisons of the pharmacokinetics of clonazepam and clorazepate in the young and the elderly have not been published.
Elderly persons tend to be more sensitive to drugs that act on the CNS. Among such drugs, the benzodiazepines have undergone the most extensive pharmacodynamic investigation. In a study of diazepam sedation, this side effect was increased in the elderly, although unbound drug concentrations did not differ from those in younger subjects (55). The increased sensitivity of the elderly to such drugs is apparently independent of drug concentration, either in the serum or at the site of action.
Drug interactions involving AEDs pose a major problem because epilepsy in the elderly is often accompanied by other disorders. In one survey, half the patients in long-term care facilities were taking five or more other maintenance medications in addition to an AED (3) (Table 13.5). Many of these comedications have clinically significant pharmacodynamic or pharmacokinetic interactions with AEDs. For example, CNS depressants such as psychotropics can exaggerate the side effects of several AEDs. Pharmacokinetic interactions between AEDs and comedications can also occur by altering the absorption, hepatic metabolism, or protein binding of either drug. Pharmacokinetic drug interactions can be multi-dimensional: valproic acid competes with PHT for plasma protein binding site, whereas PHT induces valproate metabolism and valproate inhibits PHT metabolism. Additionally, some drug interactions alter the concentrations of active AED metabolites. For example, valproate inhibits carbamazepine epoxide conversion to its inactive metabolite and results in an increase in carbamazepine epoxide, whereas carbamazepine itself may remain unchanged.
PHT absorption may be significantly reduced by certain antacids. The calcium cations often used in antacids may interact with PHT and may form insoluble complexes (56, 57). Aluminum or magnesium hydroxide may decrease gabapentin absorption. There do not appear to be clinically significant interactions affecting the bioavailability of other AEDs.
Protein binding interactions that displace AEDs generally do not alter unbound drug concentrations. Hence, dosage adjustments may not be needed, but this should be verified by measuring unbound levels.
The most clinically significant drug interactions involving AEDs are those that affect metabolism (58). Medications that inhibit hepatic enzymes responsible for drug metabolism decrease the clearance of affected drugs and result in a rise in serum concentrations. Drugs that induce hepatic enzymes increase the clearance of affected drugs and produce a decrease in serum concentration. Discontinuation of an inhibitor or inducer produces the opposite effect.
An important consideration when managing therapy in the elderly is the effect of frequent changes in drugs and doses. A drug interaction reference should be checked to determine whether a specific interaction occurs (57). A list of the clinically important interactions listed by AED and the affected isoenzyme is presented in Table 13.6.
TABLE 13.5. FREQUENCY OF USE OF COMEDICATIONS WITH POTENTIAL PHARMACOKINETIC OR PHARMACODYNAMIC INTERACTIONS WITH THE ANTIEPILEPTIC DRUGS IN 4,291 RESIDENTS OF NURSING HOMES
TABLE 13.6. ISOENZYME-MEDIATED ANTIEPILEPTIC DRUG METABOLISM AND ISOENZYME-SPECIFIC DRUG INTERACTIONS
Alternate Routes of Administration
Some elderly patients may be unable or unwilling to take oral medications because of oroesophageal conditions, other physical disabilities, or altered mental status. Elderly patients also have an increased risk of seizure emergencies such as status epilepticus that require acute management. These circumstances necessitate the use of alternative routes of administration. The most commonly used method when providing maintenance therapy is a feeding gastrostomy or nasogastric tube. Carbamazepine, felbamate, PHT, phenobarbital, primidone, and valproic acid have liquid formulations that may be given through feeding tubes. Solutions such as valproic acid syrup should be well absorbed, although this has not been substantiated, and some patients experience nausea or vomiting (59). Absorption of suspensions, the contents of capsules, and crushed tablets may be more problematic. The bioavailability of PHT suspension through a gastrostomy or nasogastric tube may be reduced when the drug is coadministered with enteral feedings (60). Therefore, administration of suspension should be separated from enteral feedings by several hours, and the tube should be thoroughly flushed after a dose. The bioavailability of carbamazepine suspension through nasogastric or gastric tubes is not known. In some patients, divalproex sodium and carbamazepine sprinkle particles adhere to the end of tubes and can cause gastric fluid to leak externally around gastric tubes (21).
Rectal administration is a viable option for some, but not all, AEDs. For short-term use (days), preparation of solutions or suspension using available commercial products are preferred, and there is some information about bioavailability and safety. When rectal administration extends over weeks or longer, suppositories should be considered because of ease of administration and patient comfort. Valproic acid syrup, carbamazepine suspension, and suspensions of crushed lamotrigine or topiramate tablets can be administered rectally to adults on a temporary basis. Bioavailability varies from 50% for lamotrigine to 80% to 100% for the other drugs, but the rate of absorption is slow for carbamazepine (61). Diazepam is absorbed very well rectally. The elderly are likely to absorb rectally administered AEDs to the same extent as younger adults, although such studies have not been done. PHT, regardless of formulation, is poorly absorbed rectally (62). Intramuscular administration of PHT sodium and valproate sodium should be avoided because both are associated with tissue injury (63).
A water-soluble PHT prodrug, fosphenytoin, may be administered intravenously or intramuscularly either as
maintenance therapy when the oral route is not available or to treat seizure emergencies. The bioavailability of intramuscular fosphenytoin approaches 100% in younger adults, and preliminary results from our group indicate that it is also rapidly and completely absorbed in elderly patients. This finding suggests that intramuscular fosphenytoin may be useful in treating seizure emergencies when intravenous therapy is not possible. Injectable forms of PHT sodium and fosphenytoin may be given intravenously, but both require monitoring of blood pressure and of heart rate and rhythms, especially in the elderly. Valproate sodium injectable may be given as a rapid intravenous infusion without altering blood pressure or cardiac function (63,64).
Seizure emergencies in the elderly can be treated with rectal diazepam. The risk of respiratory depression is low, although extra caution should be exercised in elderly patients taking multiple CNS depressants or with pulmonary disease. Intranasal administration of midazolam injectable has been reported in children and younger adults and may be an option in elderly patients. The same precautions apply to intranasal midazolam as with rectal diazepam.
The use of AEDs in the elderly poses many issues not yet thoroughly studied. Consequently, many of the remarks in this chapter are subject to change as new data become available. Nevertheless, certain conclusions can be drawn. Elderly patients must have AED concentrations monitored more closely, and unbound levels are needed for some drugs. One must be particularly cautious in using dose recommendations developed for younger adults. Because of drug interactions, one must be familiar with routes of elimination of the AEDs and the influences of other drugs. PHT is the most commonly used AED in the elderly in the United States, but it has significant drawbacks. Similarly, carbamazepine may have the side effect of hyponatremia. The newer AEDs may have a more prominent role in the elderly, but more studies are needed.
Preparation of this chapter was supported in part by National Institutes of Health-National Institute of Neurological diseases and Stroke grant no. P50 NS16308.