Epilepsy is common, affecting up to 2% of the population. Thus, it is inevitable that most doctors, whether neurologists, surgeons, general practitioners or hospital physicians, will at some stage have to manage a patient with epilepsy. People with epilepsy may have an exacerbation in their seizures due to a concomitant medical condition or its treatment, a person without epilepsy may have an acute symptomatic seizure during an acute illness, or a patient may develop epilepsy associated with a medical condition.

Even though a patient’s seizures may be well controlled at baseline, an untimely seizure during a systemic disturbance may impact adversely on outcome, predispose the patient to infections and cause metabolic or cerebral dysfunction. Additionally, the situation may be further complicated by the presence of anti-epileptic drugs, as pharmacodynamics and pharmacokinetics may be altered during a systemic illness.

Unfortunately, patients with epilepsy who have comorbid diseases are often excluded from anti-epileptic drug trials, which means there are few data about this group of patients. This chapter addresses the difficult topic of managing the patient with epilepsy and a co-morbid medical condition.

Cerebrovascular disease

Stroke may cause acute symptomatic seizures in the initial post-stroke period, as well as epilepsy in the long term. In fact, stroke is the most common cause of epilepsy in older patients. Stroke treatments, such as thrombolysis and revascularization procedures, may also result in seizures. The risk of seizures is highest after intracerebral haemorrhage, but they may also occur after ischaemic stroke and in the setting of transient ischaemic attack due to focal neuronal dysfunction. Animal studies suggest that the risk of seizures is proportional to the volume of damaged tissue.

Anti-epileptic drug prophylaxis is not recommended after stroke or intracerebral haemorrhage, unless the patient has had a seizure. The risk of seizures is greatest at presentation, and in patients with lobar haemorrhage. However, seizure occurrence does not appear to affect mortality and the risk of developing epilepsy in patients who have not had a seizure at intracerebral haemorrhage onset is low, suggesting that anti-epileptic drug prophylaxis is unnecessary. In the setting of an acute symptomatic seizure following a stroke or intracerebral haemorrhage, some would advocate the use of an anti-epileptic drug for the initial few weeks, followed by gradual weaning if the patient remains seizure free. If, however, a patient has a remote unprovoked seizure in the setting of an underlying structural lesion such as a focal infarct, most epileptologists would commence long-term anti-epileptic drug treatment, as the risk of recurrent seizures is high. The choice of anti-epileptic drug is important, as many of the older anti-epileptic drugs are sedating, affect coordination and may impact on rehabilitation. Phenytoin was found to impact adversely on functional outcome post subarachnoid haemorrhage (SAH) and phenytoin, phenobarbital and benzodiazepines were all found to be associated with less independence in activities of daily living (ADL) in patients following focal brain injury. In addition, many stroke patients also have cardiac disease, therefore phenytoin may not be ideal.

Fortunately, epilepsy occurring as a result of stroke is usually relatively easily treated, most patients becoming seizure free on monotherapy and most anti-epileptic drugs being effective. Anti-epileptic drug choice depends on other factors, such as co-morbidities, medications and side-effect profiles. However, gabapentin and lamotrigine have been found to be useful because of lack of interactions, side-effect profile and efficacy. Although there is a lack of specific data on the other newer anti-epileptic drugs, levetiracetam and topiramate would also be expected to be useful in this patient population, although both may have cognitive and behavioural adverse effects.

Cardiac disease

Atrial fibrillation occurs in 9% of patients over the age of 80 years, and prevalence is increasing as the population ages. All patients with Atrial fibrillation should be anti-coagulated with warfarin for stroke prevention unless there is a clear contraindication. It is therefore particularly important in this patient group that seizures are well controlled, as a minor injury may cause devastating haemorrhage. Epilepsy, however, is not a contraindication to anti-coagulation. One study found that warfarin remained the drug of choice in elderly patients with Atrial fibrillation despite a risk of falls. Several anti-epileptic drugs may interact with warfarin, in particular the enzyme inducers (phenytoin, phenobarbital and carbamazepine), which may enhance its metabolism, requiring an increase in warfarin dose to maintain therapeutic anti-coagulation. Importantly, there is a significant risk of over-anticoagulation if the enzyme-inducing anti-epileptic drug is withdrawn. Thus, if an anti-epileptic drug is added or withdrawn, the international normalized ratio should be monitored particularly closely. Phenytoin has an unpredictable effect on the international normalized ratio as it affects both protein binding and enzyme induction.

Cardiac arrhythmias have been reported frequently during and after seizures. The most common rhythm seen periictally is a sinus tachycardia, which is a normal physiological response to the seizure. However, symptomatic sinus bradycardias, including asystole, may occur.

Table Effects of anti-epileptic drugs on the international normalized ratio

Such patients should be considered for pacemaker insertion. S-T-segment abnormalities have also been reported in association with seizures. Patients with epilepsy who have a history of ischaemic heart disease may therefore be at increased risk of cardiac ischaemic events during a seizure.

It is important to consider that not every patient with recurrent unprovoked episodes of loss of consciousness has epilepsy. It is not uncommon for a patient with syncope to have brief convulsive jerks, which may be misinterpreted as seizure activity. Cardiac conditions, such as arrhythmias, are another common cause of recurrent collapses. Confusingly, some anti-epileptic drugs may even help prevent attacks which are not seizures. For example, patients with long Q-T syndrome may have recurrent episodes of loss of consciousness due to a brief asystole. In this situation, phenytoin, which shortens the Q-T interval, may actually prevent the arrhythmia, thereby giving a false impression of ‘seizure’ control.

Cardiac medications, particularly the class I anti-arrhythmics (lidocaine, flecainide, propafenone), which block sodium channels, may lower the seizure threshold even at therapeutic doses. Some cardiac medications, in particular quinidine and flecainide, are also metabolized by the cytochrome P450 system and may interact with certain anti-epileptic drugs. Some anti-epileptic drugs have cardiovascular side-effects, particularly when administered by intravenous infusion; this may make acute seizure management difficult in patients with cardiac conditions. Phenytoin infusion causes significant hypotension in approximately 5% of patients, and maybe pro-arrhythmogenic when administered rapidly. For patients with heart disease, the rate of phenytoin infusion should be <25mg/min. Fosphenytoin, a pro-drug of phenytoin, was initially thought to be free of cardiac adverse effects; however, recent literature suggests that cardiac adverse events are not uncommon.

Carbamazepine has also been associated with adverse cardiac events. With acute overdose, acute cardiac failure and tachyarrhythmias have been reported; with chronic use and therapeutic drug levels, bradyarrhythmias and refractory hypertension have been reported. It has been suggested that sudden unexpected death in epilepsy (SUDEP) may in some cases be attributable to anti-epileptic drug-induced cardiac dysfunction ; this remains controversial.

For acute seizure treatment in a patient with cardiac disease, intravenous valproate may be preferred to phenytoin or fosphenytoin, while the second-generation anti-epileptic drugs, such as topiramate and lamotrigine, may have fewer cardiac adverse events for patients requiring long-term seizure prophylaxis.

Renal Disease

Hepatic Disease

Metabolic Disorders

The acutely unwell or periprocedural patient

People with epilepsy are often considered to be at higher risk when undergoing procedures. This is mainly due to the possibility of seizures occurring periprocedurally or due to the potential for interactions between drugs used during the procedure and the patient’s anti-epileptic drugs.

Factors which may exacerbate seizures, such as sleep deprivation and alcohol, should be avoided prior to a procedure, and patients undergoing surgery should be advised to take their usual anti-epileptic drugs on the morning of surgery, even if fasting, and should continue their usual doses as soon as it is safe to do so.

Seizures are common after neurosurgical procedures, and are also common following cardiac operations, possibly due to complex alterations in metabolism, haemodynamic changes, alteration in blood/clotting factors and cerebral perfusion which may occur during cardiopulmonary bypass.

Although most of the local and general anaesthetic agents may have pro- as well as anti-convulsant effects, the actual risk of inducing a seizure is small. Enflurane appears to be associated with the highest risk of seizure.

Benzodiazepines are often used perioperatively and may prevent acute seizures, but care must be taken when weaning off these agents to avoid withdrawal seizures. Some analgesics such as the opiates – pethidine (meperidine) in particular – are associated with seizures, and should be avoided, where possible, post-operatively.

Anti-epileptic drug levels may be altered significantly post-operatively due to changes in hydration, volume of distribution, pharmacokinetics, pharmacodynamics, altered protein binding and blood loss. It is useful to have a baseline anti-epileptic drug level at which the patient is seizure free to allow comparison and dose alterations post-operatively.

Patients with epilepsy in the intensive care unit are at high risk of seizures. Seizures are often precipitated by this environment, even in patients without any history of epilepsy. Organ failure, metabolic changes, electrolyte and fluid imbalance, cerebral oedema, hypoxia, hypotension and hypoglycaemia are all common occurrences, as is the presence of many drugs which may lower the seizure threshold. Most patients in the intensive care unit are also sleep deprived. Pre-existing anti-epileptic drug regimens are often disturbed due to altered absorption, metabolism, dosing schedules and because the patient is unable to take drugs orally. In addition, because of the frequent use of sedation and neuromuscular blocking drugs, it may be difficult to know if a patient is actually seizing.

Oral dosing may be limited in patients who are fasting, post-ictal, have a reduced level of consciousness or who are intubated. Most anti-epileptic drugs can be given via an alternative route if the patient cannot swallow tablets:

•   Phenytoin may be given intravenously, in suspension form via enteric tubes, or

rectally If given enterally, it should be given separately from feeds as these may reduce absorption.

•   Fosphenytoin, a pro-drug of phenytoin, may be substituted. It can be given intramuscularly which is useful when access is difficult, or intravenously.

•   Phenobarbital may be given intravenously intramuscularly or rectally It is also available in liquid form which may be given via enteric tubes.

•   Valproate may be given intravenously as a syrup which may be given via enteric tubes, or by rectal suppositories which give good bioavailability

•   Carbamazepine can be given by suppository using the same dose as orally, and is also available in suspension form which can be given via enteric tubes.

•   Benzodiazepines are usually used for acute seizure management rather than for chronic epilepsy treatment, but are available in many different forms which may be useful for patients who are acutely unwell. Lorazepam may be given intravenously or sublingually Rectal diazepam is well absorbed, and is also available intravenously. Midazolam can be given into the buccal cavity or intranasally

Most of the newer anti-epileptic drugs are still not available in intravenous form. However, most can be given via enteric tubes in the following preparations:

•   Oxcarbazepine suspension

•   Lamotrigine (tablets may be crushed)

•   Topiramate (sprinkles may be injected using water)

•   Levetiracetam liquid (intravenous levetiracetam is at an advanced phase of development and is expected to be commercially available in the near future)

•   Gabapentin liquid.

Elderly Patients

Drugs And Drug Interactions

Other Conditions Which May Benefit From Anti-epileptic Drug Treatment

Conclusion

Epilepsy frequently occurs in association with another medical condition. In such patients, introduction of a new anti-epileptic drug should be accompanied by close monitoring for potential adverse effects. Careful selection of an appropriate anti-epileptic drug should allow for seizure control in the majority of patients, bearing in mind potential drug interactions and alterations in pharmacodynamics and pharmacokinetics which may occur due to the underlying disease.

Seizures may occur in uraemic encephalopathy, dialysis disequilibrium syndrome and dialysis encephalopathy. In addition, renal insufficiency and dialysis may both have effects on anti-epileptic drug pharmacokinetics. Renal impairment can alter the fraction of anti-epileptic drug absorbed, volume of distribution, protein binding and renal drug clearance.

Renal impairment may alter the gastric pH, cause small intestinal bacterial overgrowth, gastrointestimal tract oedema and impaired gastrointestinal motility. These factors may cause reduced ionization of some drugs and reduce drug absorption.

The volume of distribution of drugs may be increased in patients with end-stage renal failure, resulting in lower total plasma levels. However, protein binding of acidic drugs may be significantly reduced in renal impairment; total plasma levels of anti-epileptic drugs may therefore be misleading. A total drug level may appear within the therapeutic range, despite a toxic-free level. For anti-epileptic drugs metabolized by the liver, changes in protein binding will affect the steady-state plasma concentration, but the free concentration will remain unchanged. For these reasons, it is often more useful to measure free concentrations of anti-epileptic drugs which are highly protein bound, such as phenytoin and valproate.

Renal impairment may affect drug clearance depending on the extent of renal clearance, drug filtration and tubular secretion and resorption. Gabapentin, topiramate, pregabalin and levetiracetam are almost exclusively cleared by the kidneys. These drugs have reduced elimination and prolonged half-lives in patients with renal impairment and dosages should be adjusted accordingly. For other anti-epileptic drugs such as phenytoin, carbamazepine and tiagabine, renal excretion is minimal, and unless creatinine clearance is below 25ml/min there is no need to adjust the dose.

Table Effect of renal disease on specific anti-epileptic drugs

Several methods have been proposed to calculate anti-epileptic drug dose based on creatinine clearance, volume of distribution and other variables. However, these are complex and the actual drug concentration may vary considerably from that calculated. In patients with renal impairment, anti-epileptic drugs should be slowly initiated at low doses and clinical assessment and drug level monitoring are necessary.

Haemodialysis may clear anti-epileptic drugs from the circulation depending on their molecular size, water solubility, protein binding, volume of distribution and dialysis conditions. Drugs such as ethosuximide, gabapentin, levetiracetam, phenobarbital, pregabalin and topiramate which are highly water soluble, not protein bound and with a small volume of distribution are readily removed by Haemodialysis. Carbamazepine, clonazepam, phenytoin, tiagibine and valproate have a low risk of removal, although even these will have some removal during Haemodialysis.

If an anti-epileptic drug has been partially removed by Haemodialysis, the following equation can be used to calculate the loading dose required to restore the desired plasma concentration:

LD = С _ь X VD change

where LD = loading dose; C_Aan = desired change in plasma concentration; VD = volume of distribution.

The kidney is the most commonly transplanted organ, usually as a result of hypertensive disease, diabetes or glomerulonephritis, all of which cause uraemia and may be associated with seizures. The immunosuppressant agents most often used in renal transplantation are cyclosporine, tacrolimus, azathioprine and mycophenolate; most patients are also on steroids. Cyclosporine and tacrolimus are well known to be associated with posterior leukoencephalopathy and seizures. Cyclosporine has also been shown to reduce seizure threshold. These immunosuppressants are mainly metabolized by the liver. Enzyme-inducing anti-epileptic drugs such as phenytoin, carbamazepine and phenobarbital may reduce the plasma concentration of cyclosporine; it is therefore particularly important to monitor cyclosporine levels in such patients. It has been reported that phenytoin and phenobarbital reduce renal allograft survival, possibly due to increased metabolism of immunosuppressive agents. Owing to fewer effects on the cytochrome P450 system and fewer drug interactions, some of the second-generation anti-epileptic drugs are preferred in renal transplant patients with epilepsy; however, gabapentin has been reported in association with acute renal dysfunction in an allograft.

If transplant patients with epilepsy have breakthrough seizures, it is important to consider aetiologies other than their epilepsy as the cause. In particular, uraemic encephalopathy, metabolic derangement, opportunistic infection or leukoencephalopathy should be considered.

In summary, gabapentin and levetiracetam may accumulate in patients with renal failure. Valproate, carbamazepine, oxcarbazepine and tiagabine are less likely to cause toxicity, although their protein binding will be affected.

The liver is the principal organ of drug metabolism. Some drugs are absorbed from the gut, delivered to the liver and undergo first-pass metabolism prior to reaching the systemic circulation. Metabolism of these drugs is significantly affected by hepatic vascular supply; if hepatic blood flow is reduced, first-pass metabolism is decreased and more drug reaches the systemic circulation. Other drugs reach the systemic circulation before being delivered to the liver for metabolism; hepatocyte function is more important than the blood supply for their metabolism.

Cytochrome P450 enzymes are involved in Phase I metabolism (non-synthetic metabolism which includes oxidation, reduction and hydrolysis). There is substantial genetic polymorphism within these enzymes, and people may be slow or fast metabolizers. These enzymes are also frequently induced or inhibited by drugs, including anti-epileptic drugs. Phenytoin, phenobarbital and carbamazepine are well-known enzyme inducers.

The liver is also involved in protein production; impaired protein production reduces the amount of drug that is protein bound, increasing the free fraction. This is relevant for phenytoin, valproate, tiagabine and carbamazepine, which are highly protein bound. The true plasma concentration of phenytoin in a patient with low albumin can be calculated by Equation 2.

Table Effects of liver disease on specific anti-epileptic drugs

All anti-epileptic drugs except levetiracetam, gabapentin, pregabalin and vigabatrin have some hepatic metabolism, therefore hepatic disease may affect pharmacokinetics of most of the anti-epileptic drugs. Additionally, patients with liver disease may be encephalopathic and have altered pharmacodynamics, having a lower seizure threshold as well as being more vulnerable to the central nervous system adverse reactions of anti-epileptic drugs.

Seizures occur frequently after liver transplantation. Immunosuppressants are mainly metabolized by the liver, as are most anti-epileptic drugs. In addition, many anti-epileptic drugs may induce or inhibit hepatic metabolism. These factors make management of epilepsy difficult in patients with a liver transplant. Levetiracetam may be a good choice of anti-epileptic drug in such patients because of its predominantly renal metabolism and excretion, low protein binding, lack of enzyme induction, lack of drug interactions and broad-spectrum use in different seizure types.

A number of anti-epileptic drugs may rarely be hepatotoxic; the most common culprit is valproate, which is contraindicated in patients with hepatic failure. However, other anti-epileptic drugs, including phenytoin, may cause an acute drug-induced hepatitis. These reactions are idiosyncratic and usually occur soon after commencing the drug. Hepatic toxicity should prompt immediate discontinuation of the offending drug and substitution of an alternative anti-epileptic drug.

As gabapentin, pregabalin and levetiracetam are not significantly protein bound or metabolized by the liver, these are good choices for long-term anti-epileptic drug in patients with hepatic disease. If a patient with liver disease has an acute seizure, benzodiazepines may be used, although there is an increased risk of respiratory depression.

Table Drugs that induce or inhibit hepatic enzymes

Metabolic disturbances are a common cause of seizures, even in patients without epilepsy. In most cases, acute metabolic derangements causing seizures are potentially reversible. Seizures due to metabolic derangements are often refractory to anti-convulsant medications; correction of the underlying abnormality is required.

Hyponatraemia is the most common electrolyte disturbance encountered in clinical practice. It is a common side-effect of many drugs, including carbamazepine and oxcarbazepine. Oxcarbazepine seems to be associated with a higher incidence of hyponatraemia than carbamazepine, 30% vs. 14% in one study. Risk factors for the development of hyponatraemia are age >40, female gender, use of drugs associated with hyponatraemia, psychiatric illness and surgery. It may also be due to the syndrome of inappropriate anti-diuretic hormone secretion (SIADH) which may occur in patients with cerebral pathology. Symptoms of hyponatraemia are lethargy, confusion, muscle twitching, seizures and coma. Chronic mild hyponatraemia may be asymptomatic, and may not need specific treatment. However, acute hyponatraemia associated with neurological symptoms may need urgent treatment, in some cases requiring careful administration of hypertonic saline. It is important not to increase the sodium by more than 12mmol/l/day as there is a risk of central pontine myelinolysis with a rapid change in sodium. In epilepsy patients with pre-existing hyponatraemia, carbamazepine and oxcarbazepine should not be used if possible. If a patient with epilepsy develops hyponatraemia because of their anti-epileptic drug, it may be sufficient to reduce the dose, as the severity of hyponatraemia appears to be dose related; if persistent, however, it may be necessary to use an alternative anti-epileptic drug.

Hypernatraemia is much less common than hyponatraemia, but may occur transiently following a convulsion. Significant hypernatraemia may cause intracerebral haemorrhage, cerebral vein rupture, seizures, coma and death. Treatment is water replacement, although again it is important not to correct the sodium too quickly, as there is a risk of cerebral oedema.

Symptomatic hypoglycaemia is a common cause of seizures. Hypoglycaemia is often medication related; other causes are sepsis and inborn errors of metabolism. Serum glucose usually needs to be less than 2.2mmol/l for seizures to occur. Refractory nocturnal seizures should raise the possibility of insulinoma. Patients with diabetes who suffer from recurrent episodes of symptomatic hypoglycaemia and seizures require modification of their hypoglycaemic medication rather than anti-convulsant treatment. Phenytoin has been shown to interfere with carbohydrate metabolism, the usual effect being hyperglycaemia; however, it has also been reported in association with hypoglycaemia. An alternative anti-epileptic drug should be used in patients with diabetes.

Non-ketotic hyperglycaemia usually occurs in older patients with non-insulin-dependent diabetes. It is associated with significant morbidity and mortality and often causes seizures. It may be precipitated by infection, surgery or other physiological stress. In contrast, ketotic hyperglycaemia rarely causes seizures, as ketosis is thought to have an anti-convulsant effect. Patients with Non-ketotic hyperglycaemia are often also hyponatraemic and may develop areas of potentially reversible focal cerebral damage, both of which increase the risk of seizures. Treatment of Non-ketotic hyperglycaemia is with insulin, fluids and correction of the metabolic abnormalities.

Seizures occur in up to one-quarter of patients admitted emergently with acute hypocalcaemia. Rarely, hypocalcaemia has caused seizures in patients on long-term anti-epileptic drugs such as phenytoin and phenobarbital due to their effects on vitamin D metabolism. Acute valproate overdose has also been associated with hypocalcaemia. Treatment of symptomatic hypocalcaemia is with intravenous calcium gluconate; replacement of magnesium may also be required if the patient has co-existing hypomagnesaemia.

Porphyria is a group of disorders caused by inherited deficits of enzymes involved in haem synthesis. Seizures occur in the hepatic porphyrias: acute intermittent porphyria, hereditary co-poporphyria and variegate porphyria. Between 5% and 20% of patients with porphyria have seizures. Unfortunately, many drugs can precipitate an acute attack, including the enzyme-inducing anti-epileptic drugs. This makes treatment of epilepsy and of acute symptomatic seizures very difficult. Gabapentin and vigabatrin have both been used successfully and appear to have a low risk of precipitating a crisis. Lorazepam seems to have the lowest risk of the benzodiazepines, although all cause porphyrin accumulation. Less is known about the second-generation anti-epileptic drugs; tiagabine and topiramate appear to have the potential to cause an acute attack, while oxcarbazepine and levetiracetam have both been used safely in case reports. Pregabalin is expected to be safe, although to date there are no published data on its use in porphyria.

The incidence of epilepsy is highest in patients >75 years, with a point prevalence of 1.5%, and as the population ages doctors will treat increasing numbers of older patients with epilepsy. Treatment of epilepsy in older patients is complicated by alterations in pharmacodynamics and pharmacokinetics, metabolic derangements, presence of multiple co-morbid diseases, polypharmacy and psychosocial factors.

The aetiology of epilepsy in older patients differs from that of young people; the most frequent cause being cerebrovascular disease. Seizures are usually partial in onset and may be mistaken for recurrent strokes or TIAs, which may lead to unnecessary investigations and inappropriate treatment. In addition, morbidity and mortality are higher in elderly patients with seizures. However, the literature suggests that if the elderly patient can tolerate the anti-epileptic drug, the prognosis is usually good, with seizure freedom rates of over 60% at 1 year.

Physiological changes occur with increasing age, and these affect anti-epileptic drug absorption and metabolism. Gastric emptying, gastrointestinal blood flow and bowel motility are often reduced and gastric pH higher in elderly people. These affect drug solubility and ionization and may reduce the rate and amount of drug absorbed. In addition, the volume of distribution of a drug may be altered by muscle loss and an increase in fat which occurs with age.

Protein binding is altered due to low levels of albumin, which occurs in normal elderly people as well as those with malnutrition, renal or hepatic disease. This affects anti-epileptic drugs which are highly protein bound such as phenytoin, valproate and carbamazepine, meaning that the free drug level may be more useful in these patients.

Hepatic phase I reactions and renal clearance both decrease with advancing age; elderly patients are therefore at higher risk of anti-epileptic drug toxicity at lower doses. While there are no specific recommendations with regard to anti-epileptic drug dosing in the elderly, they may develop toxic effects at lower levels and may only tolerate lower doses.

Depression is common in the elderly, as it is in epilepsy. Johnson found that interictal psychiatric features impacted more on quality of life in people with epilepsy than did seizure frequency. Depression is frequently under-diagnosed and inadequately treated, and may impact negatively on long-term outcome. Unfortunately, most anti-depressant medications can lower the seizure threshold. Selective serotonin re-uptake inhibitors have a lower risk of seizures than the tricyclic anti-depressants. The risk of seizures associated with most anti-depressants is <0.4% overall with a dose-dependent relationship; however, the risk is probably higher in patients with epilepsy and in the elderly. If a patient at higher risk of seizures requires an anti-depressant, using a low dose and slower dose escalation may reduce the seizure risk.

Many of the anti-epileptic drugs have cognitive adverse effects, which should be considered prior to prescribing them for elderly people who may be more sensitive to these side-effects. Some of the newer anti-epileptic drugs, such as lamotrigine and zonisamide, may have less cognitive adverse effects than some of the first-generation anti-epileptic drugs.

Osteoporosis is frequent in the elderly population, and fractures are a major cause of morbidity and mortality. Patients with epilepsy on anti-epileptic drugs have double the risk of fractures, probably due to a combination of increased risk of falls due to adverse effects of anti-epileptic drugs, seizure-related trauma and reduced bone health. A number of anti-epileptic drugs, predominantly phenytoin, valproate and phenobarbital have been associated with reduced bone mineral density, although more recent studies have also implicated carbamazepine and oxcarbazepine. Patients at risk should be offered dual energy X-ray absorptiometry (DEXA) scanning and may require treatment with bisphosphonates and calcium and vitamin D supplementation.

One study found that the mean number of daily medications for elderly patients in the community was eight, with 40% prescribed more than nine medications daily. Polypharmacy increases the risk of medication non-compliance, which may in turn increase the risk of seizures. In addition, polypharmacy increases the potential for drug-drug interactions. Elderly patients are more likely to have memory problems and visual impairment and these factors make medication errors more likely. Because of the narrow therapeutic index, lower tolerability and higher adverse effect rate of anti-epileptic drugs in older patients, it is preferable to initiate anti-epileptic drugs at lower doses and titrate more slowly than is necessary in younger patients.

Adverse effects of anti-epileptic drugs are one of the most important problems in epilepsy management in elderly people. Older patients are often excluded from drug trials due to co-morbid illnesses and presence of other medications; however, the literature suggests decreased tolerability of the first-generation anti-epileptic drugs with increasing age. In addition, the recent Veterans’ Affairs Cooperative Study found that retention rates in elderly people with epilepsy were better for lamotrigine and gabapentin than for carbamazepine, without any significant difference in rates of seizure freedom. Despite this, phenytoin, carbamazepine and valproate remain the most frequently prescribed anti-epileptic drugs in nursing home residents. Elderly patients are more likely to develop adverse effects of drugs at lower doses, as they tend to tolerate a narrower therapeutic range. Adverse effects such as tremor, ataxia, visual disturbances and sedation are common and occur at lower drug levels in elderly people. These adverse effects in turn increase the risk of falls and of non-compliance.

When choosing an anti-epileptic drug for an elderly patient with epilepsy, one needs to consider co-morbid conditions and medications. Monotherapy is preferable as they are often on many other medications, and the recent literature would suggest that the second-generation anti-epileptic drugs are preferable as first-line treatment rather than the more established first-generation anti-epileptic drugs (phenytoin, carbamazepine, valproate). If a patient is on other sedating medications, it would be wise to avoid prescribing potentially sedating anti-epileptic drugs such as gabapentin. A patient with tremor or ataxia may be made worse by valproate, phenytoin and carbamazepine. However, even a carefully chosen medication may be less well tolerated and require lower dose and slower titration in this group of patients.

The goal of epilepsy treatment is ‘no seizures, no side-effects’. Monotherapy is preferred if possible, because of greater patient compliance with medications, better quality of life, more favourable adverse effect profile, lack of drug interactions and cost issues. However, many patients with epilepsy will require more than one anti-epileptic drug, and many patients will also be on other medications for co-morbid conditions. This means that a large proportion of people with epilepsy are at risk of drug interactions.

When new anti-epileptic drugs are undergoing trials, they are almost invariably tested as add-on therapy, meaning that they are initially licensed only as add-on treatment rather than as monotherapy.

Drug interactions may be pharmacodynamic or pharmacokinetic. The clinical significance of a drug interaction depends on the drugs involved as well as patient factors such as age and co-morbid conditions. However, drug interactions are not necessarily harmful. There are several combinations of anti-epileptic drugs with synergistic effects on each other, such as lamotrigine and valproate, meaning that lower doses of both are required, which may reduce cost. The combination of phenytoin and valproate also appears to have some synergistic activity, so lower drug levels may confer the same anti-convulsant effect. Using an anti-epileptic drug which prolongs the half-life or reduces the metabolism of another may mean prolonging the time between doses, which may improve patient compliance.

There are several ways in which anti-epileptic drug absorption may be altered. Activated charcoal, antacids and resins reduce drug absorption by adsorbing drugs onto their surface and preventing passage of drug across the stomach wall. As soon as this occurs, serum concentration of the drug begins to decrease, the rate of decrease depending on its half-life. Metoclopramide and other drugs that increase gastric motility may increase the absorption of drugs, while anticholinergics and other drugs that slow gastric emptying may delay absorption. An increase or decrease in the rate of absorption may not alter drug bioavailability, but may be important if a rapid effect is required. Drugs that alter gastric pH, such as histamine receptor blockers, proton pump inhibitors and antacids, may change ionization of acidic or basic drugs, thereby affecting absorption. All of these local factors may be reduced by the patient taking the interacting drugs several hours apart.

Alterations in protein binding may be another mechanism for drug interactions. This is largely important for drugs which are highly protein bound and have a narrow therapeutic index, such as phenytoin. If one drug is commenced when a patient is stable on another, the new drug will displace some of the original drug from plasma proteins, causing an increase in the unbound fraction. However, when the unbound fraction of a drug increases, more is available for metabolism and clearance, so in fact the drug concentration may decrease. These interactions are unpredictable, so it is important to monitor free anti-epileptic drug levels as total levels are misleading.

The most common cause of drug interaction, however, is induced alterations in hepatic metabolism. Phenytoin, phenobarbital, carbamazepine, topiramate, tiagabine, zonisamide and f elbamate are all metabolized by the cytochrome P450 enzyme system. Several drugs can induce or inhibit these enzymes in a dose-dependent fashion. Carbamazepine, for example, is an autoinducer as it induces the isoenzyme which is responsible for its own metabolism. Phenytoin, phenobarbital, carbamazepine and topiramate all induce cytochrome P450 enzymes, while valproate is an inhibitor. Interactions due to enzyme induction occur relatively slowly, as new proteins have to be synthesized. Those resulting from enzyme inhibition depend on the half-life of the affected drug as it takes approximately five half-lives for a drug to reach a steady-state concentration. Gabapentin, levetiracetam and vigabatrin are only minimally metabolized and so have less potential for these type of interactions.

An important interaction occurs in young women who are using the oral contraceptive pill. Phenytoin, phenobarbital, carbamazepine, oxcarbazepine, felbamate and topiramate increase the metabolism of ethinyl estradiol and progestogens, reducing the effectiveness of the oral contraceptive pill. With these anti-epileptic drugs, the dose of oestrogen should be increased to at least 50µg, or an alternative form of contraception used.

One should always remember that patients with chronic medical conditions such as epilepsy are more likely to use over-the-counter non-prescription medications. Several of these, in particular St John’s Wort and Ginkgo biloba, may have effects on enzyme activity and have the potential for altering anti-epileptic drug metabolism. In addition, some may have pharmacodynamic interactions, and may have a clinically significant effect without any effect on drug levels. Most patients using complementary medicine do not inform their doctor, so it is important to ask the patient directly, as self-medication with over-the-counter medications may be a reason for subtherapeutic or toxic drug levels.

It is not an exhaustive list of interactions, as there are many other non-anti-epileptic drugs which are also inducers or inhibitors and may cause interactions. Consult the British National Formulary (BNF) for more information.

In conclusion, although the potential for drug interactions is high with anti-epileptic drugs, the majority of patients prescribed potentially interacting drugs do not in fact develop any clinical consequences. Patient factors are probably the most important reason behind this. The physician may prevent serious consequences by being aware of the potential interaction, anticipating it, monitoring the patient clinically and using drug levels wisely. As a general rule, the second-generation anti-epileptic drugs have far fewer pharmacokinetic interactions as most do not have any appreciable effect on metabolizing enzymes. If the risk of a serious interaction is high, it may be preferable to use an alternative drug.