Some patients can suffer unrecognised seizures for many years [9], particularly children and teenagers who are often thought to be daydreaming. Relatives may have witnessed a number of episodes and not recognised them as seizures.A useful sequence of questions includes:

1. Have you ever seen the patient suddenly interrupt what they were doing and stare into space, where their eyes were open but they did not respond to you? If the answer is yes, this is in keeping with absence or complex–partial seizures.

2. What was the patient doing at the time the episode commenced?

3. What was the first thing that you noticed and how long did it last?

4. What was the next thing that you noticed and how long did it last?

5. What was the next thing that you noticed and how long did it last?

Keep asking this question until the whole episode has been described in detail. More specific questions include:

• During the episode was the patient able to hear what you were saying or were they out to it? If the answer is no, this indicates a loss of awareness suggesting a generalised seizure or complex–partial seizure.

• Did you see any excessive blinking or abnormal chewing movements of the mouth? These occur with absence and complex–partial seizures, respectively.

It is often useful to question more than one eyewitness or relative as one individual may not have observed all the details. When seeing a patient with suspected epilepsy or a seizure, more is to be gained by picking up the telephone and ringing a family member or an eyewitness than undertaking investigations.

As a general rule, the more complex and varied the behaviour during an episode and the longer the duration of the abnormal behaviour, the less likely one is dealing with a seizure disorder.

Epilepsy in the elderly

It is a common misconception that epilepsy is a disease of childhood. While this is largely true of absence (petit-mal) seizures, which are very rare in adults and often present when they do occur with absence status epilepsy, seizures do occur in the elderly and the incidence increases with increasing age, as shown in Figure 8.1. Elderly patients with epilepsy most often present with tonic–clonic or complex–partial seizures that have a higher recurrence rate than in the younger population. The complex–partial seizures are often difficult to diagnose since they present with atypical symptoms, particularly prolonged post-ictal symptoms including memory lapses, confusion, altered mental status and inattention [10].

FIGURE 8.1 Schematic diagram to show that both maturational factors in seizure susceptibility derived from onset of age-related epileptic disorders and febrile convulsions (a) and environmental factors derived from age-related risk of cerebral injury (b) combined to yield the incidence of epilepsy by age (c) Reproduced from ‘Seizures and Epilepsy’, by J Engel, Jr. In: Contemporary Neurology Series, edited by F Plum, Vol 31, 1989, FA Davis Company, Figure 5.3, p 115 [12]

Although there is a continuing incidence of seizures throughout life, they are more common in the first 5 years. There is also a higher incidence in the 70- to 80-year-old age group [11].

Seizures can occur in any age group and either sex.

CHARACTERISATION OF THE TYPE OF SEIZURE

The clinical manifestations of the more common types of seizures is discussed here; more detail can be found in textbooks [12–14]. The commonest seizures in clinical practice are:

• benign rolandic seizures.

Although the traditional approach is to classify seizures as focal versus generalised or primary (no obvious cause) versus secondary, an alternate way to characterise the various types of seizures is based on what can be witnessed:

1. seizures that cause the patient to fall with or without loss of consciousness (see Figure 8.2)

FIGURE 8.2 Seizures that cause a fall with or without loss of consciousness (LOC) ^∗A fall will result if the patient is standing and the myoclonus affects the legs. Patients with brief focal myoclonic seizures may not fall if they are seated.

2. seizures that are not associated with a fall with or without loss of consciousness (awareness; see Figure 8.3).

FIGURE 8.3 Seizures that do not cause a fall with or without loss of consciousness (LOC) ^∗Patients with myoclonic or focal–motor seizures may fall if they are standing but there is no LOC.

Drop attacks due to epilepsy mainly occur in childhood; drop attacks that occur in elderly adults are thought not to be epileptic in origin and are discussed in Chapter 7, ‘Episodic disturbances of function’. The Lennox–Gastaut syndrome is seen in childhood and consists of multiple seizure types including atypical absence seizures, myoclonic, tonic–clonic and atonic seizures with often a degree of intellectual disability [5].

Myoclonic seizures

Myoclonus is a shock-like contraction of a muscle or a group of muscles. Myoclonus occurs in people without epilepsy, often when they are just falling asleep but also occasionally any time of the day. The frequency would usually be less than once a month. There is also an entity called segmental myoclonus that is not a seizure disorder, where a particular muscle or group of muscles supplied by a particular nerve or nerve root will have myoclonic jerks.

In patients with myoclonic epilepsy, these myoclonic jerks are more frequent, often occur during sleep but characteristically occur first thing in the morning on awakening. They affect the whole body or a single limb and may be single or repetitive jerks. In myoclonic seizures the patient is fully aware of what is happening. There is no aura, loss of awareness or post-ictal confusion, i.e. the patient is normal immediately before and after the event. Myoclonus induced by movement is a feature of post-hypoxic myoclonus [15].

JUVENILE MYOCLONIC EPILEPSY

Juvenile myoclonic epilepsy (JME) is myoclonus, absence seizures and occasional tonic–clonic seizures. This usually presents between ages 6 and 22, but there is a lifelong predisposition to recurrent seizures without treatment [16, 17].

Complex–partial seizures

Complex–partial seizures most commonly arise from the temporal lobe in association with mesial temporal sclerosis. Complex–partial epilepsy (CPE) was originally referred to as temporal lobe epilepsy. However, the recognition that minor seizures that impair consciousness without causing the patient to fall may arise from the occipital, frontal or parietal lobes has led to the use of the term CPE.

• Immediately before ictus: The duration of the aura is usually measured in seconds; the symptoms of aura have been described above.

• During ictus: The ictus usually lasts seconds to a few minutes. The patient is not aware of what is happening, nor are they able to respond to any verbal or painful stimuli. In the words of two relatives witnessing minor seizures: ‘the lights were on but no one was at home’ or ‘there were no sheep in the top paddock’. There may be involuntary movements of the mouth or limbs, depending on which area of the brain is the focus for the seizure.

• After ictus: There is a period of post-ictal confusion lasting several minutes during which the patient can usually respond to outside stimuli but is clearly disoriented and confused [18]. The patient is unable to relate what happened during the event.

Absence (petit-mal) seizures

Absence epilepsy generally occurs in childhood and is very rare as a primary presentation in adults. In retrospect, however, when one takes a careful history from a patient with their first tonic–clonic seizure, many patients have had unrecognised absence seizures [9] in childhood and have simply been regarded as either dull or a daydreamer.

• Immediately before ictus: The absence seizure is characterised by no warning unless it is an atypical absence.

• During ictus: The period of impaired consciousness is brief. In one study the average seizure duration was 9.4 seconds (range 1–44 seconds, SD 7 seconds), 26% of seizures were shorter than 4 seconds and 8% were longer than 20 seconds [19]. The patient simply stares into space and may blink, but they are clearly unresponsive to either verbal or painful stimuli. They have no recollection of what happened or was said to them during the episode.

• After ictus: Immediately following the ictus, the patient is able to resume a conversation and the activity that they were undertaking without any post-ictal confusion.

It is surprising how many absence seizures patients can experience without people actually noticing them because they are so brief. Absence seizures can readily be induced by hyperventilation (HV). In one study of 47 children with childhood absence seizures, HV induced seizures in 83% (39/47) of children. Of the eight children who did not have seizures induced by HV, four were too young to perform HV. In the other four children, HV was performed but may not have been performed well [19].

The important clue that helps differentiate absence seizures from daydreaming is that with absence seizures the patient will interrupt their behaviour. For example, they will momentarily pause unexpectedly during a conversation or a game they have been playing, and then appear to resume it as if nothing has happened. Many children daydream or are concentrating on television or some other activity and, although it can be difficult to attract their attention, they are not unresponsive. One common presentation of absence epilepsy is the child who has previously done well at school and who begins to fail. Unfortunately, many children in the past were assumed to be daydreamers who were not working hard enough when in fact they were missing large parts of the lesson because of minor absence seizures. In the author’s experience, many teachers fail to detect children with minor epilepsy. It is when these children who fail have further teaching that the seizures are actually discovered by the tutor who spends more time with them on a one-to-one basis.

The EEG in absence seizures shows a characteristic 3 per second spike and wave (see Figure 8.4).

FIGURE 8.4 EEG in absence epilepsy showing 3 per second spike and wave

Simple partial seizures

The essential characteristic of simple partial seizures is that they can occur at any age in either sex and consist of a brief stereotyped sensation without loss of awareness, i.e. the patient is fully conscious of what is happening and is able to describe the whole episode from start to finish. There is no warning or aura, nor is there any post-ictal drowsiness or confusion. Examples of this type of epilepsy are patients who experience a tingling sensation that may commence in their feet and rise up to the top of the head and sometimes go back down again to their feet, over a matter of seconds.

Focal motor seizures

These episodes are very brief, usually lasting less than a minute or two, although they can go on repetitively in patients, in which case the term focal motor status epilepsy is used. The patient is fully aware of what is going on around them, there is usually no warning and certainly there is no post-ictal drowsiness or confusion. The seizures consist of repetitive contractions down one side of the body.

Benign focal seizure of childhood

There are a number of benign focal seizures of childhood [20, 21]. There are currently three identifiable electroclinical syndromes recognised by the International League against Epilepsy (ILAE) [22]: rolandic epilepsy, Panayiotopoulos syndrome (PS) or common autonomic epilepsy and the idiopathic childhood occipital epilepsy of Gastaut (ICOE-G). They produce terrifying manifestations. Details are given in Appendix B, ‘Benign focal seizures of childhood’, because they are rare and atypical with the duration of symptoms being longer than for most other seizure types.

Febrile convulsions

Febrile convulsions affect 2% to 5% of all children and usually appear between 3 months and 5 years of age. Febrile convulsions may be provoked by any febrile bacterial or viral illness and no specific level of fever is required to diagnose febrile seizures. It is essential to exclude underlying meningitis either clinically or, if any doubt remains, by lumbar puncture. The risk of epilepsy following a febrile seizure is 1% to 6% [23].

Reflex epilepsies

Reflex epilepsies are very rare, but are discussed briefly so that the reader is familiar with this entity. The clue to the clinical diagnosis is repeated precipitation of a brief stereotyped episode by a particular stimulus. One of the commonest is the tonic–clonic seizure that is precipitated by the stroboscope at a dance.

The reflex epilepsies are syndromes in which all epileptic seizures are precipitated by sensory stimuli. Generalised reflex seizures are precipitated by visual light stimulation, thinking and decision making. Numerous triggers can induce focal reflex seizures, including reading, writing, other language functions, startle, somatosensory stimulation, proprioception, auditory stimuli, immersion in hot water, eating and vestibular stimulation [24].

Consider possible multiple types of seizures

Having established that the patient has suffered a seizure, the next step is to determine whether they suffer from any other seizure types and whether they have suffered from undiagnosed seizures in the past. Although most patients who present with their first tonic–clonic seizure have not had other manifestations of epilepsy, with detailed questioning it is not uncommon for patients to have had unrecognised myoclonic seizures and/or minor absence or complex–partial seizures in the past [9].

Most seizures conform to the common seizure types listed above and the diagnosis rarely presents any difficulty. On the other hand, there are rare cases where the events during the ictus are atypical and can lead to an erroneous diagnosis of a non-organic event. Consider the following actual case history (Case 8.1) where the principles discussed earlier in this chapter are highlighted.

CASE 8.1

Dear Doctor,

Thank you for seeing this young man with very strange behaviour. He has had recurrent episodes lasting less than a minute or two. He is witnessed to walk across the room and shakes whoever is in the room a few times. After this he is confused for a few more minutes and states that he does not recall what happened.

Yours sincerely,

This patient did exhibit very strange behaviour. However, the episodes were stereotyped and brief, lasting less than 90 seconds, and during the episode there were positive phenomena in that he walked, grabbed people and shook them. He lost awareness and was amnestic during the episode and it was followed by a short period of post-ictal confusion. Subsequent video-EEG monitoring confirmed that these episodes were complex–partial seizures.

Pseudoseizures

Pseudoseizures are paroxysmal changes in behaviour that resemble epileptic seizures, but which are without organic cause and the expected EEG changes. Longer ictal duration and less stereotypy (variation of the manifestations from one episode to the next) are characteristic. Other characteristic features include asynchronous extremity movements, atypical vocalisation, alternating head movements, talking or screaming, opisthotonic posturing (arching of the back) and pelvic thrusting [25].

An increasing frequency of ‘seizures’ with escalating doses of anticonvulsants should alert the clinician to the possible diagnosis [26]. The diagnosis can be very difficult as up to 30% of patients with pseudoseizures have epileptic seizures as well and pelvic thrusting can be a very rare manifestation of temporal lobe or frontal lobe epilepsy [27]. The diagnosis of psychogenic pseudoseizures has improved with the availability of video-EEG monitoring [28]. If pseudoseizures are suspected, one way to confirm the diagnosis is to instruct family members, nursing staff on the ward or anyone who may witness an episode to try and interrupt what is happening either by talking to the patient or by employing a painful stimulus, e.g. pinching the skin over the medial aspect of the elbow. Organic seizures, as opposed to pseudoseizures, cannot be interrupted.

ASSESSING THE FREQUENCY OF SEIZURES

The patient with their ‘first seizure’

As already stated it is not uncommon to see patients with their first apparent tonic–clonic seizure who have had a previous seizure or a long history of unrecognised minor seizures or myoclonus [9]. Ask the patient whether they have frequent involuntary brief jerking of one limb or all of the body, similar to the sensation that they experience when they are just falling asleep. These involuntary jerks (termed myoclonus) can occur infrequently, perhaps a few times per year in normal individuals, whereas in patients with epilepsy they occur more frequently and often first thing in the morning upon wakening, particularly in patients with juvenile myoclonic epilepsy syndrome. Although the patient is very aware of these myoclonic jerks, they interfere so little with their life that often they do not seek medical attention. Similarly, minor seizures may go undetected or undiagnosed for many, many years. Often detailed questioning of relatives or friends using the questions described above will elicit a prior history of previously unrecognised minor seizures. Ascertain how frequently the various types of seizures occur (i.e. many times per day, daily, weekly etc).

Status epilepsy

Status epilepsy is a seizure that ‘persists for a sufficient length of time or is repeated frequently enough that recovery between attacks does not occur’. Both convulsive and non-convulsive status epilepsy are recognised.

CONVULSIVE STATUS EPILEPSY

Status tonic–clonic (convulsive) epilepsy consists of continuous or rapidly repeating seizures and is a medical emergency. The treating doctor should not leave the bedside until seizures cease. There are published guidelines for the management of status epilepsy [29–31]. Guidelines are often updated and every institution should have access to them. Patients who do not respond to initial therapy should be treated in an intensive care unit, as artificial ventilation and haemodynamic support are required.

NON-CONVULSIVE STATUS EPILEPSY

Non-convulsive status epilepsy is an epileptic state in which there is some impairment of consciousness associated with ongoing seizure activity on the EEG [32]. It is often a sequel to convulsive status epilepsy [32]. Patients may present in a coma without any overt signs of seizure activity [33].

IDENTIFYING ANY PRECIPITATING CAUSES

Identifying a reversible precipitating cause is important as avoidance in the future can reduce the risk of subsequent seizures. Most seizures are due to an abnormal electrical discharge within the brain, often referred to as primary epilepsy. Seizures can also be secondary to the hypotension that occurs for example with syncope, a Stokes–Adams attack, alcohol abuse or withdrawal of sedative or hypnotic drugs.

• In patients with primary epilepsy, sleep deprivation, photic stimulation, alcohol abuse [34], menstruation and an intercurrent infective illness may potentially predispose to recurrence [35].

• Vomiting and/or diarrhoea alter drug levels, predisposing to recurrent seizures, and many drugs, including psychotropics, local and general anaesthetics, narcotics, antiarrhythmics and antibiotics, are recognised precipitating factors for seizures [36].

• Poor compliance is one, if not the commonest, cause of recurrent seizures in patients on anticonvulsants [34].

ESTABLISHING AN AETIOLOGY

Most patients with epilepsy do not have any identifiable underlying pathology. It is possible that with future advancements in the understanding of epilepsy more causes will be identified. Focal-motor seizures, complex–partial seizures and, particularly, seizures associated with a focal neurological deficit are more likely to have identifiable pathology on imaging [37].

Seizures may occur with alcohol or drug withdrawal, infective processes within the central nervous system, drugs, benign and malignant tumours, just to mention a few possible aetiologies. Seizures can be the presenting symptom of disturbances of cardiac rhythm [38] or, rarely, secondary to severe pain, for example in patients with trigeminal neuralgia [39]. Detailed lists can be found in textbooks on epilepsy [5–7, 14, 40, 41]. This means that it is not possible to be dogmatic about what investigation(s) should be performed in an individual patient; suffice to say that every effort should be made to find a cause.

A word of warning: the presence of a family history of epilepsy or a past history of a head injury only suggests that the seizure may be familial or post-traumatic, and other pathological processes need to be considered.

Investigations

• Some would argue that it is justifiable to perform medical imaging of all adult patients who present with tonic–clonic, complex–partial and focal motor seizures [42, 43]. Others would argue that there is currently insufficient data to support or refute recommending any of these tests for the routine evaluation of adults presenting with an apparent first unprovoked seizure [43]. The yield of currently available imaging techniques is about 10%. Current recommendations advocate an EEG, CT or MRI brain scan in all patients presenting with a first unprovoked seizure [43]. In reality most, if not all, patients will demand imaging for reassurance that they do not have a brain tumour.

• Laboratory tests, such as blood counts, blood glucose and electrolytes, particularly sodium, lumbar puncture and toxicology screening may be helpful as determined by the specific clinical circumstances based on the history, physical and neurological examination, but there are insufficient data to support or refute recommending any of these tests for the routine evaluation of adults presenting with an apparent first unprovoked seizure [43].

• On the other hand, although metabolic disturbances causing seizures are rare they are readily correctable, particularly hypoglycaemia. When a patient presents with a seizure, probably the single most important test to perform immediately is a serum glucose to exclude hypoglycaemia. Other metabolic disturbances, such as hyponatraemia, hypocalcaemia and an elevated urea, may cause seizures and, as a general principle, these investigations are without risk and the subsequent management is easy. It would seem reasonable to exclude metabolic causes in all patients.

DECIDING WHETHER TO TREAT OR NOT

In many patients with simple partial seizures, reassurance that nothing sinister is the matter suffices and they do not wish to take medication to stop such trivial symptoms. In patients with an isolated idiopathic (no particular cause) tonic–clonic seizure the subsequent risk of further seizures varies from study to study. In one study of children it was 54% [44], while in another study of adults there was a 27% risk of recurrence at 36 months [45]. A first seizure provoked by an acute brain disturbance is unlikely to recur (3–10%), whereas a first unprovoked seizure has a recurrence risk of 30–50% over the next 2 years [46].

The number of seizures of all types at presentation, the presence of a neurological disorder and an abnormal EEG are significant risk factors for recurrent seizures. An abnormal EEG is defined as specific focal or generalised epileptiform or slow wave abnormalities. Individuals with two or three seizures plus a neurological disorder and/or an abnormal EEG have been identified as a high-risk group with a 73% incidence of recurrent seizures at 5 years [47]. In the same study the recurrence rate at 5 years in patients with a single seizure, a normal EEG and no neurological abnormality was 39% [47].

In a study of immediate versus delayed treatment with currently available anticonvulsants, immediate treatment did not reduce the long-term recurrence rate. At 5-year follow-up, 76% of patients in the immediate treatment group and 77% of those in the deferred treatment group were seizure free (difference –0.2%, 95% CI –5.8% to 5.5%) [48]. When confronted with such a low risk of recurrence many patients elect not to take antiepileptic drug (AED) treatment after a first seizure [46]. It would therefore seem reasonable not to recommend therapy in patients with an isolated tonic–clonic seizure in the absence of a positive family history, an abnormal EEG and pre-existing cerebral pathology, particularly if there was an easily reversible precipitating cause such as sleep deprivation. The patient in this situation often decides. The question to put to the patient is: ‘What effect would another seizure have on your life?’ The commonest reason patients give for wanting to go onto medication is that a further seizure would terrify them, although more often it is the unacceptable social implications and the effect on their ability to drive, and thus on their employment, that influences their decision whether to take medication or not. Being employed is a major reason why patients continue to drive against medical advice [49].

Occasional patients can have two seizures many years apart. If the patient had been placed on medication after the first seizure, the clinician would have regarded the treatment as excellent when in fact the natural history was such that the patient was not going to experience a second seizure for many years. In general, AEDs should be offered after a first tonic–clonic seizure if:

• the patient has had previous myoclonic, absence or partial seizures

• the EEG shows unequivocal epileptic discharges

• the patient has a congenital neurological deficit

• the patient considers the risk of recurrence unacceptable.

CHOOSING THE APPROPRIATE DRUG, DOSE AND ONGOING MONITORING OF THE RESPONSE TO THERAPY

Choosing a drug

Treatment is targeted primarily to:

• assist the patient in adjusting psychologically to the diagnosis and in maintaining as normal a lifestyle as possible

• reduce or eliminate seizure occurrence

• avoid or minimise the side effects of long-term drug treatment.

Most patients with more than one seizure require prophylactic AED therapy. Some seizure types or epilepsy syndromes may respond to certain drugs while other drugs may exacerbate the seizures [50, 51].

The general principle is to choose a drug with the greatest efficacy in terms of preventing seizures, provided there is no contraindication to its use. If there are several possible drugs, the ‘correct drug’ is the drug the patient deems to have the fewest undesirable potential side effects from their perspective, e.g. the risks of weight gain or interference with oral contraceptives are common reasons why young females refuse a particular drug. Patients can only make such a decision when they receive detailed information regarding the proposed treatment options.

Side effects are a significant reason for discontinuing an AED, particularly in the elderly [52].

When monotherapy does not control seizures, review the diagnosis of epilepsy and check for compliance with medication. Failure of the first AED due to lack of efficacy (and not due to incorrect choice of drug, wrong dose, incorrect dosing schedule or poor compliance) implies refractoriness and trying multiple AEDs one after another is unlikely to be successful [53]. Combination therapy should be considered when treatment with two first-line AEDs has failed or when the first well-tolerated drug substantially improves seizure control but fails to produce freedom from seizure at maximal dosage. In this instance, an AED with different and perhaps multiple mechanisms of action should be added. Strategies for combining drugs should involve individual assessment of the patient’s seizure type, together with an understanding of the pharmacology, side effects and interaction profiles of the AEDs [54]. The choice of drugs to use in combination should be limited to two or at most three AEDs.

An important principle when altering therapeutic regimens is to avoid changing two things at a time. In that situation it is difficult to know what has done the good or harm.

Current treatment recommendations and the potential side effects of specific drugs are given in Appendix C, ‘Currently recommended drugs for epilepsy’, and links to websites that should provide more up-to-date information are given in Chapter 15, ‘Further reading, keeping up-to-date and retrieving information’.

Choosing the dose

Other than with status epilepsy, there is no need to reach a therapeutic dose rapidly. Therefore, treatment should commence with the lowest possible dose and increase very slowly over weeks as this may reduce the incidence of side effects [55, 56]. The number of doses per day is dependent on the half-life of the drug, but medications prescribed more than twice a day are associated with an increasing incidence of poor compliance [57].

Monitoring the response

In many patients complete freedom from seizures is not possible and sensible decisions regarding the efficacy of medication can be made only with careful assessment of seizure frequency. Seizure frequency has a significant impact on quality of life [58]. Although it is recommended that patients keep a diary to record the frequency of their seizures, very few do. It is important for the treating doctor to keep detailed notes on behalf of the patient and refer to these notes when deciding if a particular AED has been effective in reducing the number of seizures. Monitoring the response to treatment at times can be extraordinarily difficult, particularly in young patients who through embarrassment will often deny forgetting their medication, drinking too much alcohol or staying out all night.

Intractable epilepsy

One of the major difficulties in patients with intractable epilepsy is that many do not know what drugs have been used, what the doses were and whether they were stopped due to an apparent lack of efficacy or because of side effects. The lack of efficacy may relate to an incorrect choice of drug or drug dosage and scheduling. It is sometimes useful to show patients pictures of the specific drugs or, better still, have all the common AEDs stuck onto cardboard so that patients can identify the drug they have taken. Unfortunately, even when they can identify the drug, they are usually unaware of the dose tried, the frequency of dosing and the reason for stopping the drug. It is useful to advise patients to keep a diary of the date and time of any seizures that occur for two reasons:

1. To see whether there is a decrease in the frequency of seizures when therapy is modified.

2. Perhaps more importantly, to determine what time of the day the seizures occur in relation to the timing of medication.

In patients on a twice daily dosage regimen, a number will experience seizures in the hour or two before or within the first half hour after the dose. Simply increasing the individual dose in this setting results in symptoms of toxicity as the peak serum level increases but then falls below the threshold for the patient’s seizures just prior to the next dose. The appropriate course of action is to increase the frequency of the medication if the drug has a half-life of 8–12 hours (although compliance is better when drugs are prescribed less frequently) or, alternatively, change the patient to a drug that has a longer half-life.

Unfortunately, with many drugs, there is a ‘honeymoon period’ during the first 6 months of treatment when the number of seizures decreases, but subsequently seizures may become just as frequent.

Measuring serum levels

Serum AED measurements are useful when the level directly reflects efficacy; this is not the case for all drugs. Other clinical situations where AED monitoring is useful include checking for compliance or toxicity, as a guide to adjusting the dose when another drug is added or during pregnancy [59].

ADVICE REGARDING LIFESTYLE

Driving and epilepsy

Each country has different regulations with regard to driving and epilepsy, and it is important for all clinicians who care for patients with epilepsy to have a copy of their respective country’s (or state’s) guidelines. In recent years restrictions have been less stringent, reflecting the low contribution of seizures to the overall road toll [60]. In some countries the restriction on driving is very severe, whereas in others a more lenient approach is taken.

Essentially, a sufficient period of time needs to elapse to be certain that a recurrent seizure is unlikely to occur. In one study patients who had seizure-free intervals of 12 months or longer had 93% reduced odds of motor vehicle accidents compared to patients with shorter seizure-free intervals. The majority (54%) of patients who crashed were driving illegally, with seizure-free intervals shorter than legally permitted [61], and 20% had missed an AED dose just prior to the crash. Patients should be informed that, although the risk of a seizure is small, the potential consequences could be disastrous. It is also a useful tactic to show patients the guidelines, explaining that the medical practitioner does not make the rules and that these rules are not there to punish patients with epilepsy but to protect the patient from hurting themselves or innocent people.

The risk of a seizure occurring while driving is also a reflection of the time spent driving. If the patient drives for only half an hour per day the risk of a seizure while driving is much lower than if they drive for 12 hours per day. For this reason in many countries the requirements are more stringent for commercial and heavy goods vehicle licenses than for private motor vehicle licenses [62].

Pregnancy and epilepsy

Although the increased risk of major congenital malformations in patients with epilepsy taking AED(s) during pregnancy is possibly 2–3 times that of the normal population [63], most pregnancies will result in a normal child. There is some debate about the exact role of AED exposure in pregnancy and any increased risk [64]. Some AEDs and the use of multiple AEDs (polypharmacy) may be associated with a greater risk [65–67]. Long-term cognitive problems in the child may also occur [65].

In clinical practice many patients attend their doctor to discuss withdrawal of medication when they are 2–3 months pregnant. Any teratogenic effects will have already occurred, and it is probably unwise and essentially too late to withdraw medication at this stage of the pregnancy. The risk of uncontrolled seizures during pregnancy needs to be weighed against the risk of AED exposure. An increased risk to the mother and infant is often quoted but it is difficult to find good evidence to justify this statement. In the European Registry of Antiepileptic Drugs and Pregnancy (EURAP) study, of 1956 pregnancies 58.3% were seizure-free throughout pregnancy. Seizures occurred during delivery in 60 pregnancies (3.5%), more commonly in women with seizures during pregnancy (odds ratio (OR): 4.8; 2.3–10.0). There were 36 cases of status epilepsy (12 convulsive), which resulted in stillbirth in only 1 case, but no cases of miscarriage or maternal mortality [68]. A Cochrane review has concluded that, based on the best current available evidence, it would seem advisable for women to continue medication during pregnancy using monotherapy at the lowest dose required to achieve seizure control [65].

Some patients seek advice about ceasing their AEDs prior to pregnancy and here the decision is more difficult. Even if the epilepsy is easily controlled, the patient has to be willing to run the risk of a recurrent seizure and the subsequent consequences regarding the alteration in their lifestyle, in particular driving. In general, if the epilepsy has been difficult to control, an argument can be made for the patient to remain on medication throughout the pregnancy. Accumulating evidence from drug registries suggests that the lowest possible dose and avoidance of certain drugs may be appropriate [68–70]. Patients willing to consider a mid-trimester termination can have testing for major malformations.

Some patients may experience an increased seizure frequency while others have fewer seizures [68]. Occasionally, patients not known to suffer from epilepsy have a tonic–clonic seizure during labour and, although other serious disorders need to be considered, often a detailed history will elicit a long history of infrequent minor and previously unrecognised seizures.¹ Most patients with epilepsy will maintain control during pregnancy [68].

Potentially hazardous activities

Patients should be warned that it may be dangerous to go swimming or fishing, have a bath or walk near water if they are alone as a seizure could result in drowning [71]. Patients should also be advised not to scale heights, walk near the edge of cliffs, sky dive or scuba dive.

CONSIDERATION OF SURGERY IN PATIENTS WHO FAIL TO RESPOND TO DRUG THERAPY

Approximately one-third of patients with epilepsy are refractory to AED therapy and many of these patients are candidates for surgical treatment [72]. Among patients who do not respond to the first drug, the percentage who subsequently become seizure-free with additional AEDs is smaller (11%) [53]. Patients with refractory epilepsy should be referred to an appropriate centre earlier rather than later for potential surgery.

Investigation and treatment

The use of magnetic resonance imaging (MRI) has been pivotal in the evaluation of patients with partial seizures [73]. Patients with MRI-negative partial epilepsy may be candidates for additional neuroimaging techniques including positron emission tomography, MR spectroscopy and single photon emission tomography. Peri-ictal imaging may allow identification of the epileptogenic zone in patients with normal MRI scans.

In patients with refractory epilepsy, vagal nerve [74], hippocampal [75] and bilateral cerebellar [76] electrical stimulation have been advocated. These techniques have resulted in a modest reduction in seizure frequency but none has eliminated seizures altogether. On the other hand surgery, in particular temporal lobectomy for mesial temporal sclerosis, has been shown to be highly effective in a randomised controlled trial, with more than 50% of patients seizure-free as opposed to 8% in the non-surgical group [77].

Macroscopic and radiological evidence of total lesional excision with isolated structural lesions such as dysembryoplastic tumours, low-grade astrocytomas or focal vascular abnormalities is associated with excellent seizure-free outcome [78].

Corpus callosotomy is recommended for patients with atonic seizures or drop attacks [79, 80].

Hemispherectomy is now a widely accepted procedure for medically refractory, catastrophic hemispheric epilepsy. The classic anatomical hemispherectomy procedure has been abandoned in favour of functional or modified hemispherectomy [81].

WHETHER AND WHEN TO WITHDRAW THERAPY IN ‘SEIZURE-FREE’ PATIENTS

A patient who has been free of seizures for some time often asks whether AEDs can be stopped. The juvenile myoclonic epilepsy syndrome consisting of absence and myoclonic seizures and infrequent tonic–clonic seizures is associated with a lifelong predisposition to seizures and AEDs should not be ceased.

In other forms of epilepsy where remission can occur and a trial off AEDs is not unreasonable, the initial question to the patient must be: ‘What effect would a recurrent seizure have on your life?’ For example, the patient would be unable to drive for a prescribed period after the seizure and this could impact on their life and work. A recurrent seizure may occur in a situation where it would cause significant embarrassment or, worse still, possible injury. A minimum of 2 years free of seizures is recommended before considering withdrawal of medication [82]. Even when the risk of recurrence is low, it must be stated that there is no guarantee that there will not be a recurrent seizure. Recurrent seizures may occur after as many as 8 years off medication.²

In a review of 28 studies accounting for 4571 patients (2758 children, 1020 adults and a combined group of 793), most with at least 2 years of seizure remission, the proportion of patients with relapses during or after AED withdrawal ranged from 12% to 66% [83].

A higher-than-average risk of seizure relapse included:

• adolescent-onset epilepsy

• partial seizures

• the presence of an underlying neurological condition

• abnormal EEG findings at the time of AED withdrawal in children [83].

Most relapses occur during or within the first 6 months after withdrawal [84]. At the time of writing there was no evidence to determine the rate of withdrawal [85], but it seems reasonable to reduce the dose slowly over several weeks to months.