© The Fellowship of Postgraduate Medicine, 1992
Postgrad Med J (1992) 68, 522- 527
Epilepsy Anthony Hopkins Research Unit, Royal College of Physicians, 11 St Andrews Place, London NWJ 4LE, UK
Introduction Epilepsy is just as much with us at the end of John Walton's professional lifetime as it was at its beginning. Nevertheless, there have been striking advances in a number of fields, of which I choose to review here those which have had, in my view, the most impact upon our understanding of epilepsy and upon the quality oflife of our patients suffering with epilepsy.
Epidemiology Most older textbooks of neurology quote the prevalence of epilepsy in the population at large as 1 in 200 (0.5%). This overall figure has proven to be correct in a number of large-scale epidemiological studies since then.'`3 However, this prevalence figure, considered in isolation, proved to be remarkably uninformative about the incidence and natural history of epilepsy. The landmark study which illuminated more clearly the true natural history of epilepsy was published from a number of general practices in 1960.4 From a review of their records, this group of general practitioners was able to show that the true cumulative incidence of epilepsy was at least 3.2%. Subsequently, Hauser and his colleagues undertook a community survey of epilepsy amongst the population of Olmstead County in Rochester, Minnesota.' Figure 1 is reproduced from their study. For Hauser's and many other epidemiological studies, the definition of epilepsy has been more than one non-febrile seizure of any type. Hauser was able to identify accurately a community population with reported seizures as the great majority of those living in Olmstead County seek both their primary and secondary care at the nearby Mayo Clinic, which has excellent medical records extending over many years. Another more recent study based upon general practices in the United Kingdom, with careful record review, again showed a documented history of recurrent seizures
Correspondence: A. Hopkins, M.D., F.R.C.P.
of at least 16.7 per thousand of the population.5 It can be seen from Figure 1 that there is a high age-specific annual incidence rate in childhood, and a high age-specific incidence rate in older age, and a relatively flat plateau of about 0.4 new cases per thousand between youth and old age. The overall annual incidence of new cases of epilepsy in this and subsequent studies by Hauser based upon the same community has been shown to be of the order of 0.61 per 1,000 per year. If the broad figure of 0.6 new cases per 1,000 per year is taken as an average annual incidence rate, it immediately follows that if a subject lives for 70 years, the cumulative lifetime risk is of the order of 70 x 0.6 = 42 per 1,000. This is to say, approximately 4% of the population will, at one stage of their lives, satisfy the criteria of being 'epileptic' that is to say having had more than one non-febrile seizure. The figures are of considerable interest in underlining the fact that epileptic seizures are not some rare and unusual event that require extensive investigation, but are so common in the general 160 - (249.6) 140 o 120 o 100
x1955-64
-
1,1965-67
a 80
c,
1
/ ,o1935-44 60 3 4o >'-*1945~~.-194-54
-- s.
20
30 40 50 60 70 80 Age (years) Figure 1 Age-specific incidence rates of epilepsy first diagnosed at different intervals in Rochester, Minnesota, 1935-1967. (Reproduced with kind permission from Hauser and Kurland.')
0
10
20
EPILEPSY
population, as are febrile convulsions,6 as to be a part of the life of a significant number of the population. Hauser's and other studies have shown that the old overall prevalence figure of 0.5% is correct. The concept, prevalent when John Walton was entering neurological practice, that epilepsy affected only 1 in 200 of the population, failed to make clear that most people with epilepsy enter long-term or permanent remissions, as illustrated in Figure 2 and its legend.7 The perspective of neurologists of epilepsy was informed by hospital practice, in which are seen patients with intractable epilepsy, poorly responsive to anticonvulsant drugs. The poor prognosis of this group8 misinformed neurologists of the much more benign prognosis of epilepsy in the community. Remission
80 -
ever
In remission
-.
60-
0
Q1)
I,
.E 40-
/
I I
Without seizuress or medication
523
Another important change of perspective has been on the prognosis after a first seizure. For many years it has been taught that 'one fit does not necessarily mean epilepsy'. Those of John Walton's generation who trained me would indulge in complex circumlocutions when dealing with patients after a single undoubted tonic-clonic seizure by trying to explain what was the nature of the fit, it if was not an epileptic fit. The apparent belief in the biological difference of single and multiple epileptic seizures remains in the practice of the Department of Transport in the United Kingdom today, which, by the current Driving Licence Regulations9 is obliged to inform a patient with epilepsy that he or she is not eligible to drive until 2 years have elapsed since the date of the last seizure. However, the Department of Transport commonly allows a patient who has had a single seizure to drive after 12 months. In fact a number of studies (for example, refs 10 and 11) have shown such a high recurrence rate after a first seizure, most commonly within the first few weeks, that there is clearly no real biological difference between one seizure and more than one seizure (Figure 3). We are now in a better position to advise patients ofthe likely risk of recurrence after a first seizure. Important variables are the interval since the first seizure (Figure 3), the time of day at which the seizure occurred, the age of the subject, and whether or not there is a family history of febrile convulsions or epilepsy.'0 In childhood, the presence of spike and wave discharges is also an important prognostic feature.'2 The value of the electroencephalogram (EEG) in
20 1.0
0.8 I
5
6
10
I 15
alIl
20
Years after diagnosis Figure 2 Remission of seizures. *-: probability of completing a period of 5 consecutive years without seizures. For example, 6 years after diagnosis 42% of subjects have been seizure-free for 5 years. ---: the probability of being in remission for at least the past 5 years. The difference between the top and middle curves is due to relapse after achievement of a 5-year remission. For example, at 20 years after diagnosis 70% are currently free from seizures, and have been for 5 years, and a further 6% have had at least one seizure-free period of at least 5 years' duration, but have subsequently relapsed. : the probability of being free of seizures for at least 5 years whilst not taking anticonvulsant drugs. In summary, 20 years after diagnosis 50% have been free from seizures without anticonvulsants for at least 5 years. A further 20% continue to take anticonvulsant medication and have also been free of seizures for at least 5 years. Seizures continue, in spite of medication, in 30%. (Reproduced with kind permission from Annegers et al.7)
.~0.6 0
° 0.4 0.2
male n = 182 0 female n = 124
0 3/12 6/12 1 yr 1/2yrs 2yrs
3yrs
4yrs
Interval since first seizure (years) Figure 3 Probability of remaining free from further epileptic seizures after an initial seizure in 306 men and women. * = male, n = 182; 0 = female, n = 124. The probabilities and 95% confidence intervals at the following times: 3, 12 and 24 months. Men 0.83 (0.77-0.88); 0.64 (0.57-0.71); 0.55 (0.47-0.63). Women 0.79 (0.71-0.86); 0.62 (0.53-0.71); 0.54 (0.45-0.64). No significant difference. (Reproduced with kind permission from Annegers et al.7)
524
A. HOPKINS
predicting seizure recurrence after a first seizure in adult life is controversial.'0"3 Epidemiological and clinic-based studies have also informed neurologists better about the prognosis of those with established seizures.7'8 It is now recognized that those with a combination of partial and tonic-clonic seizures, those whose seizures occur in clusters, those with associated cognitive or neurological impairment, and those who have already had seizures for a considerable time have a particularly poor chance of achieving a worthwhile remission of long duration. Serum anticonvulsant levels As just mentioned, the duration of epilepsy to date is a powerful predictor of the probability of achieving a future long-term remission from seizures. For those patients presenting to a neurologist who have already had recurrent seizures for a number of years, the outlook is, unfortunately, therefore unfavourable. An exception is if prior treatment has been inadequate. It is still not uncommon to see patients on clearly insufficient dosage of anticonvulsant medication. The choice of an appropriate anticonvulsant drug and of an appropriate oral dosage of that drug was revolutionized about 20 years ago by the introduction of the measurement of serum levels. The ability to measure serum levels of phenytoin, initially by gas-liquid chromatography and subsequently by immunoassay techniques, uncovered not only considerable complexities in the pharmaco-dynamics of phenytoin metabolism (for review, see ref. 14), but also established an upper level beyond which toxic effects such as nystagmus, slurred speech and ataxia were particularly likely to occur. Measurements also suggested a lower level below which, although some anticonvulsant activity may be present, an increase was clearly appropriate. The increasing availability of the more or less routine measurement of anticonvulsant drugs has brought to light a number of problems - notably the induction of hydroxylating enzymes in the liver by phenytoin, and the effects of other drugs upon the serum phenytoin level. It rapidly became clear that the anticonvulsant action of one drug widely marketed in the early 1970s - sulthiame - was largely due to its ability to elevate the serum level of phenytoin coincidentally prescribed. The possibility of measuring accurately and increasing appropriately the oral dose of one drug soon made it clear that in many cases seizure control could be achieved by adequate dosage with one drug (monotherapy). The advantages of monotherapy include increased compliance by the patient, avoidance of potential interactions with multiple drugs, and less impairment of cognitive
function. There are, however, theoretical reasons suggesting that therapy with more than one drug may sometimes be appropriate.'5 Of the principal anticonvulsant drugs in common use, measurement of the serum level is appropriate for phenytoin and carbamazepine. These have both been shown to be effective drugs in the treatment of partial and tonic-clonic seizures.'6 Phenobarbitone, and primidone, which is partly metabolized to phenobarbitone, can readily be measured in the serum, but receptor tolerance means that the serum levels do not necessarily reflect biological activity. The serum level of sodium valproate, being bound to tissue organelles, may not reflect the day to day anticonvulsant activity of the drug. Although a number of new drugs have recently been introduced, or are at an advanced stage of trial in the management of epilepsy, there has been no major 'breakthrough' of the type for which people with epilepsy and their neurologists are constantly hoping. Vigabatrin'7 has recently been introduced as it has been shown to be efficacious in add-on trials of patients with resistant epilepsy. Vigabatrin is an inhibitor of gamma-aminobutyric acid (GABA) transaminase. In one trial'8 more than 50% of those with resistant epilepsy showed a reduction in seizure frequency of more than 50%. Gabapentin is a GABA-related amino acid. Although possibly rather less effective than vigabatrin, a significant benefit to about 25% of patients has been shown in one trial.'9 Oxcarbazepine is the 10-keto analogue of carbamazepine. Its 10-hydroxy metabolite is responsible for the anticonvulsant effect. Placebo-controlled trials show a similar efficacy to carbamazepine, but with a lower allergic propensity, and with fewer side effects on the nervous system.20 Lamotrigine probably exerts its anti-epileptic effect by inhibiting the release of excitatory amino-acid neurotransmitters. It is also an effective supplementary drug in some poorly controlled epileptic patients. The investigation of seizures
This is a third area in which there have been considerable advances during John Walton's professional lifetime. When I trained as a registrar at Queen Square in the 1960s, it was not uncommon for patients who reported a single tonic-clonic seizure to be investigated either by air encephalography or by arteriography, or sometimes by both. These activities now seem extraordinary in the light of more sophisticated epidemiological evidence referred to above, but the perspective of neurologists then was very much that of neurosurgeons. About 54% of all patients with brain tumours have seizures at the time of assessment and diagnosis.22
EPILEPSY
However, this figure is meaningless when translated to the population of people with first seizures who have a much smaller risk of a brain tumour about 4% in one study.'0 Electroencephalography (EEG) has also come under increasingly critical review. A brief 25 or 30 minute paper recording will be normal in approximately 50% of patients with epilepsy, although repeated recordings increase the proportion of patients with abnormal records.23 The principal, indeed almost the only plank in the diagnostic platform of epilepsy, is the historical account given by the patient and bystanders. EEG certainly has a role in those with multiple seizures, as, if an abnormality is found, it may be sufficiently clear to diagnose accurately a seizure type, thereby influencing the choice of anticonvulsant medication. In adult practice, this means distinguishing between primary generalized epilepsy with typical 3 cycle per second spike and wave activity, and those who have an inter-ictal spike suggesting the probability that the seizure has a partial onset. It is obviously extremely helpful to obtain the record of a seizure discharge during an overt clinical seizure, but unless seizures are very frequent, this is seldom practical. However, EEG monitoring accompanied by simultaneous video monitoring has had a major impact in evaluating the origin of the seizure discharges in those with very frequent seizures who are candidates for surgery.24 Video and EEG monitoring has also revolutionized the diagnosis of simulated seizures (pseudoseizures), which account for a significant proportion of those with 'epilepsy' apparently intractable to all types of anticonvulsant medication.25 When John Walton began his neurological career, imaging studies available for epilepsy were insensitive and invasive. A skull X-ray might have shown intracranial calcification, for example, in the case of a slowly growing oligodendroglioma, or the classical tram-lines appearance of vascular calcification in the rare Sturge-Weber syndrome. As noted above, over-investigation of people with epilepsy for suspected tumours led to large numbers of air encephalograms and carotid arteriograms. The introduction of isotope brain scanning in the early 1960s did reduce these invasive tests considerably, but it was not until Hounsfield's invention of cranial computerized tomographic (CT) scanning that a sensitive and non-invasive study for the detection of abnormalities of cerebral structure became available. Since then, magnetic resonance imaging (MRI) has proved more sensitive than CT scanning in the detection ofcerebral lesions related to the genesis of seizures. MRI will detect small temporal lobe haematomas and tumours responsible for complex seizures and increasing technical advances are more frequently demonstrating mesial temporal sclerosis, the most common pathological finding on those who have
525
temporal lobectomies for temporal lobe epilepsy. MRI is only available in a few centres in the United Kingdom, and CT head scanners, although now widely available geographically, are always busy. Furthermore, the technician time and capital cost of the machine means that some sort of policy should be developed for deciding whom ofall those patients with new onset epilepsy should have a CT scan. Scheuer and Pedley26 take the view that some sort of brain imaging studies should be performed in all patients over the age of 18 years, and in all children with partial seizures (except those with benign focal epilepsy of childhood), those with abnormal findings on the neurological examination, or focal slow-wave abnormalities on the electroencephalogram. However, in the Royal College of Physicians' prospective study of first seizuresl' my colleagues Charles Clarke, Andrea Garman and I found that of 370 patients with first seizures, only 13 turned out to have brain tumours. Furthermore, scans are not always reliable - subsequent follow-up of our patients found that two further patients whose initial scans had been normal subsequently developed tumours. Two further procedures are now available for research purposes in epilepsy - positron-emission tomography, and single photon-emission CT (SPECT). These techniques allow the imaging of areas of cerebral hypometabolism interictally in the temporal lobes of those who have partial seizures, and may aid a decision about the appropriateness of temporal lobe surgery. In conclusion, although neurophysiological and imaging studies are available for people with epilepsy, few tumours are detected, and some are missed. With increasing demands for the rational use of resources, it is likely that the use of these procedures in the management of epilepsy will decline further in future years. The surgical treatment of epilepsy
This is the fourth of my topics to illustrate the advances made in the management of some cases of epilepsy in John Walton's professional lifetime. It has of course been recognized for nearly 100 years that operations on brain tumours may in themselves relieve epilepsy associated with the tumours. It was then recognized that smaller lesions, particularly in the temporal lobe, could be responsible for epilepsy, and that whereas these lesions did not in themselves warrant removal on account of progressive pathology, removal might have a remarkable effect on the frequency of the seizures, without leaving a major neurological deficit. Colleagues at the Maudsley Hospital headed first by Murray Falconer, and more recently by Charles Polkey have been amongst the international leaders in the
526
A. HOPKINS
field. The development of the technique of block dissection of the anterior part of the temporal lobe developed by Falconer and his colleagues proved to be instrumental in the development of our understanding of the pathological nature of the underlying changes responsible for epilepsy. There has been an enormous investment in clinical and neurophysiological time in attempting to predict those patients who will benefit from surgery. First of all, patients must of course be proven to be refractory to a number of anticonvulsant drugs, and combinations of drugs in dosages sufficient to achieve high therapeutic levels. Interictal scalp EEG recording is of limited value, and some sort of video monitoring with simultaneous EEG recording of the seizures is usually necessary to determine the site of origin of the seizures. Wieser and his colleagues have shown that an electrode inserted through the frame foramen ovale reliably records seizure discharges some of which do not reach scalp electrodes.27 American neurosurgeons have also written of the value of subdural and indwelling depth electrodes. As noted above, MRI and SPECT both have their value in locating the seizure focus to one or other temporal lobe. Histologically, the removed temporal lobe may show mesial temporal sclerosis, or small slowgrowing gliomas, or dysembryoplastic neuro-epithelial tumours.28
With careful case selection, more than half of all patients with intractable temporal lobe epilepsy after operation have no further seizures, and about a further third have a significant reduction in seizure frequency. There are, however, psychological effects that should not be underestimated, particularly in terms of verbal memory in left-sided operations, which may prove a significant handicap to those whose life and careers demand verbal skills. I have chosen to concentrate upon four aspects of epilepsy, but others are worthy of brief mention. There is now much better understanding of the biology of the epileptic focus,29 much better understanding ofthe relationship between febrile convulsions and epilepsy,30 much better understanding of the toxic effects of drugs, much better understanding of the frequency of pseudoseizures, much better understanding of the risks of post-traumatic epilepsy, and of the difficulty in preventing it,31 and finally, with the recent publication ofthe MRC trial on withdrawing anticonvulsants, we can reasonably hope to have better understanding of the risks and benefits of withdrawing anticonvulsants from those who have been free from seizures for 2 years or more. There remains, however, a substantial shortfall of funds for and people interested in research in epilepsy.33
References 1. Hauser, W.A. & Kurland, L.T. The epidemiology of epilepsy in Rochester, Minnesota 1935 through 1967. Epilepsia 1975, 16: 1-66. 2. Pond, D.A., Bidwell, B.H. & Stein, L. A survey of epilepsy in fourteen general practices. I. Demographic and medical data. Psychiat Neurol Neurochir 1960, 63: 217-236. 3. Gudmundsson, G. Epilepsy in Iceland: a clinical and epidemiological investigation. Acta Neurol Scand 1966, 43 (Suppl 25): 1-24. 4. Crombie, D.L., Cross, K.W., Fry, J. et al. A survey of the epilepsies in general practice: a report by the Research Committee of the College of General Practitioners. Br Med J 1960, 2: 416-422. 5. Goodridge, D.M.G. & Shorvon, S. Epileptic seizures in a population of 6,000. I. Demography, diagnosis and classification, and role of the hospital services. Br MedJ 1983, 287: 641-644. 6. Verity, C.M., Butler, N.R. & Golding, J. Febrile convulsions in a national cohort followed up from birth. I. Prevalence and recurrence in the first five years of life. Br Med J 1985, 290: 1307- 1310. 7. Annegers, J.F., Hauser, W.A. & Elveback, C.R. Remission of seizures and relapse in patients with epilepsy. Epilepsia 1979, 20: 729-737. 8. Rodin, E.A. The Prognosis of Patients with Epilepsy. Thomas, Springfield, Illinois, 1968. 9. Department of Transport (UK). Driving Licence Regulations 1989. 10. Hopkins, A., Garman, A. & Clarke, C. The first seizure in adult life. The value of clinical features, electroencephalography and computerised tomographic scanning in prediction of seizure recurrence. Lancet 1988, 1: 721-726.
11. Hauser, W.A., Rich, S.S., Annegers, J.F. & Anderson, V.E. Seizure recurrence after a first unprovoked seizure: an extended follow up. Neurology 1990, 40: 1163-1170. 12. Hauser, W.A., Anderson, V.E., Loewenson, R.B. & McRoberts, S.M. Seizure recurrence after a first unprovoked seizure. N Engl J Med 1982, 307: 522- 528. 13. van Donselaar, C.A., Geerts, A.T. & Schimsheiner, R.-J. Idiopathic first seizure in adult life: who should be treated. Br Med J 1991, 302: 620-623. 14. Richens, A. Drug Treatment ofEpilepsy. Kimpton, London, 1976. 15. Hopkins, A., Davies, P. & Dobson, C. Mathematical modelling of patterns of intervals between seizures. Arch Neurol 1985, 42: 463-467. 16. Mattson, R.H., Cramer, J.A., Collins, J.F. et al. Comparison of carbamazepine, phenobarbital, phenytoin and primidone in partial and secondarily generalised tonic-clonic seizures. N Engl J Med 1985, 313: 145- 151. 17. Reynolds, E.H. Vigabatrin: rational treatment for chronic epilepsy. Br Med J 1990, 300: 277-278. 18. Treiman, D.M. Gamma vinyl GABA: current role in the management of drug resistant epilepsy. Epilepsia 1989, 30 (Suppl 3): S31-35. 19. UK Gabapentin Study Group. Gabapentin in partial epilepsy. Lancet 1990, 335: 1114-1117. 20. Editorial. Oxcarbazepine. Lancet 1989, 2: 196-198. 21. Jawad, S., Richens, A., Goodwin, G. & Yuen, W.C. Controlled trial of lamotrigine (Lamictal) for refractory partial seizures. Epilepsia 1989, 30: 356-363. 22. Thomas, D.G.T. & Graham, D.I. Brain Tumours: Scientific Basis, Clinical Investigation and Current Therapy. Butterworths, London, 1980.
EPILEPSY 23. Salinsky, M., Kanter, R. & Dashieff, R.M. Effectiveness of multiple EEGs in supporting the diagnosis of epilepsy: an operational curve. Epilepsia 1987, 28: 331-334. 24. Gotman, J., Ives, J.R. & Gloor, P. (eds). Longterm monitoring in epilepsy. Electroencephalogr Clin Neurophysiol 1985, Suppl 37: 327-340. 25. Hopkins, A. Pseudoseizures. Q JMed 1989, NS 71:473-475. 26. Scheuer, M.L. & Pedley, T.A. The evaluation and treatment of seizures. N Engl J Med 1990, 323: 1468-1474. 27. Wieser, H.G., Elger, C.E. & Stodieck, S.R.G. The foramen ovale electrode: a new recording method for the pre-operative evaluation of patients suffering from medio-basal temporal lobe epilepsy. Electroencephalogr Clin Neurophysiol 1985, 61: 314-322.
527
28. Editorial. Lessons in the surgical treatment of epilepsy. Lancet 1989, 2: 366-367. 29. Meldrum, B.S. Anatomy, physiology and pathology of epilepsy. Lancet 1990, 336: 231-234. 30. Annegers, J.F., Hauser, W.A., Shirts, S.B. & Kurland, L.T. Factors prognostic of unprovoked seizures after febrile convulsions. N Engi J Med 1987, 316: 493-498. 31. Hauser, W.A. Prevention of post-traumatic epilepsy. N Engl J Med 1990, 323: 540-541. 32. MRC Antiepileptic Drug Withdrawal Study Group. Randomised study of antiepileptic drug withdrawal. Lancet 1991, 337: 1175-1180. 33. Shorvon, S.D. The lack of funds for clinical epilepsy research in the UK. J R Coll Phys Lond 1991, 25: 31-32.
Postgrad Med J (1992) 68, 522- 527
Epilepsy Anthony Hopkins Research Unit, Royal College of Physicians, 11 St Andrews Place, London NWJ 4LE, UK
Introduction Epilepsy is just as much with us at the end of John Walton's professional lifetime as it was at its beginning. Nevertheless, there have been striking advances in a number of fields, of which I choose to review here those which have had, in my view, the most impact upon our understanding of epilepsy and upon the quality oflife of our patients suffering with epilepsy.
Epidemiology Most older textbooks of neurology quote the prevalence of epilepsy in the population at large as 1 in 200 (0.5%). This overall figure has proven to be correct in a number of large-scale epidemiological studies since then.'`3 However, this prevalence figure, considered in isolation, proved to be remarkably uninformative about the incidence and natural history of epilepsy. The landmark study which illuminated more clearly the true natural history of epilepsy was published from a number of general practices in 1960.4 From a review of their records, this group of general practitioners was able to show that the true cumulative incidence of epilepsy was at least 3.2%. Subsequently, Hauser and his colleagues undertook a community survey of epilepsy amongst the population of Olmstead County in Rochester, Minnesota.' Figure 1 is reproduced from their study. For Hauser's and many other epidemiological studies, the definition of epilepsy has been more than one non-febrile seizure of any type. Hauser was able to identify accurately a community population with reported seizures as the great majority of those living in Olmstead County seek both their primary and secondary care at the nearby Mayo Clinic, which has excellent medical records extending over many years. Another more recent study based upon general practices in the United Kingdom, with careful record review, again showed a documented history of recurrent seizures
Correspondence: A. Hopkins, M.D., F.R.C.P.
of at least 16.7 per thousand of the population.5 It can be seen from Figure 1 that there is a high age-specific annual incidence rate in childhood, and a high age-specific incidence rate in older age, and a relatively flat plateau of about 0.4 new cases per thousand between youth and old age. The overall annual incidence of new cases of epilepsy in this and subsequent studies by Hauser based upon the same community has been shown to be of the order of 0.61 per 1,000 per year. If the broad figure of 0.6 new cases per 1,000 per year is taken as an average annual incidence rate, it immediately follows that if a subject lives for 70 years, the cumulative lifetime risk is of the order of 70 x 0.6 = 42 per 1,000. This is to say, approximately 4% of the population will, at one stage of their lives, satisfy the criteria of being 'epileptic' that is to say having had more than one non-febrile seizure. The figures are of considerable interest in underlining the fact that epileptic seizures are not some rare and unusual event that require extensive investigation, but are so common in the general 160 - (249.6) 140 o 120 o 100
x1955-64
-
1,1965-67
a 80
c,
1
/ ,o1935-44 60 3 4o >'-*1945~~.-194-54
-- s.
20
30 40 50 60 70 80 Age (years) Figure 1 Age-specific incidence rates of epilepsy first diagnosed at different intervals in Rochester, Minnesota, 1935-1967. (Reproduced with kind permission from Hauser and Kurland.')
0
10
20
EPILEPSY
population, as are febrile convulsions,6 as to be a part of the life of a significant number of the population. Hauser's and other studies have shown that the old overall prevalence figure of 0.5% is correct. The concept, prevalent when John Walton was entering neurological practice, that epilepsy affected only 1 in 200 of the population, failed to make clear that most people with epilepsy enter long-term or permanent remissions, as illustrated in Figure 2 and its legend.7 The perspective of neurologists of epilepsy was informed by hospital practice, in which are seen patients with intractable epilepsy, poorly responsive to anticonvulsant drugs. The poor prognosis of this group8 misinformed neurologists of the much more benign prognosis of epilepsy in the community. Remission
80 -
ever
In remission
-.
60-
0
Q1)
I,
.E 40-
/
I I
Without seizuress or medication
523
Another important change of perspective has been on the prognosis after a first seizure. For many years it has been taught that 'one fit does not necessarily mean epilepsy'. Those of John Walton's generation who trained me would indulge in complex circumlocutions when dealing with patients after a single undoubted tonic-clonic seizure by trying to explain what was the nature of the fit, it if was not an epileptic fit. The apparent belief in the biological difference of single and multiple epileptic seizures remains in the practice of the Department of Transport in the United Kingdom today, which, by the current Driving Licence Regulations9 is obliged to inform a patient with epilepsy that he or she is not eligible to drive until 2 years have elapsed since the date of the last seizure. However, the Department of Transport commonly allows a patient who has had a single seizure to drive after 12 months. In fact a number of studies (for example, refs 10 and 11) have shown such a high recurrence rate after a first seizure, most commonly within the first few weeks, that there is clearly no real biological difference between one seizure and more than one seizure (Figure 3). We are now in a better position to advise patients ofthe likely risk of recurrence after a first seizure. Important variables are the interval since the first seizure (Figure 3), the time of day at which the seizure occurred, the age of the subject, and whether or not there is a family history of febrile convulsions or epilepsy.'0 In childhood, the presence of spike and wave discharges is also an important prognostic feature.'2 The value of the electroencephalogram (EEG) in
20 1.0
0.8 I
5
6
10
I 15
alIl
20
Years after diagnosis Figure 2 Remission of seizures. *-: probability of completing a period of 5 consecutive years without seizures. For example, 6 years after diagnosis 42% of subjects have been seizure-free for 5 years. ---: the probability of being in remission for at least the past 5 years. The difference between the top and middle curves is due to relapse after achievement of a 5-year remission. For example, at 20 years after diagnosis 70% are currently free from seizures, and have been for 5 years, and a further 6% have had at least one seizure-free period of at least 5 years' duration, but have subsequently relapsed. : the probability of being free of seizures for at least 5 years whilst not taking anticonvulsant drugs. In summary, 20 years after diagnosis 50% have been free from seizures without anticonvulsants for at least 5 years. A further 20% continue to take anticonvulsant medication and have also been free of seizures for at least 5 years. Seizures continue, in spite of medication, in 30%. (Reproduced with kind permission from Annegers et al.7)
.~0.6 0
° 0.4 0.2
male n = 182 0 female n = 124
0 3/12 6/12 1 yr 1/2yrs 2yrs
3yrs
4yrs
Interval since first seizure (years) Figure 3 Probability of remaining free from further epileptic seizures after an initial seizure in 306 men and women. * = male, n = 182; 0 = female, n = 124. The probabilities and 95% confidence intervals at the following times: 3, 12 and 24 months. Men 0.83 (0.77-0.88); 0.64 (0.57-0.71); 0.55 (0.47-0.63). Women 0.79 (0.71-0.86); 0.62 (0.53-0.71); 0.54 (0.45-0.64). No significant difference. (Reproduced with kind permission from Annegers et al.7)
524
A. HOPKINS
predicting seizure recurrence after a first seizure in adult life is controversial.'0"3 Epidemiological and clinic-based studies have also informed neurologists better about the prognosis of those with established seizures.7'8 It is now recognized that those with a combination of partial and tonic-clonic seizures, those whose seizures occur in clusters, those with associated cognitive or neurological impairment, and those who have already had seizures for a considerable time have a particularly poor chance of achieving a worthwhile remission of long duration. Serum anticonvulsant levels As just mentioned, the duration of epilepsy to date is a powerful predictor of the probability of achieving a future long-term remission from seizures. For those patients presenting to a neurologist who have already had recurrent seizures for a number of years, the outlook is, unfortunately, therefore unfavourable. An exception is if prior treatment has been inadequate. It is still not uncommon to see patients on clearly insufficient dosage of anticonvulsant medication. The choice of an appropriate anticonvulsant drug and of an appropriate oral dosage of that drug was revolutionized about 20 years ago by the introduction of the measurement of serum levels. The ability to measure serum levels of phenytoin, initially by gas-liquid chromatography and subsequently by immunoassay techniques, uncovered not only considerable complexities in the pharmaco-dynamics of phenytoin metabolism (for review, see ref. 14), but also established an upper level beyond which toxic effects such as nystagmus, slurred speech and ataxia were particularly likely to occur. Measurements also suggested a lower level below which, although some anticonvulsant activity may be present, an increase was clearly appropriate. The increasing availability of the more or less routine measurement of anticonvulsant drugs has brought to light a number of problems - notably the induction of hydroxylating enzymes in the liver by phenytoin, and the effects of other drugs upon the serum phenytoin level. It rapidly became clear that the anticonvulsant action of one drug widely marketed in the early 1970s - sulthiame - was largely due to its ability to elevate the serum level of phenytoin coincidentally prescribed. The possibility of measuring accurately and increasing appropriately the oral dose of one drug soon made it clear that in many cases seizure control could be achieved by adequate dosage with one drug (monotherapy). The advantages of monotherapy include increased compliance by the patient, avoidance of potential interactions with multiple drugs, and less impairment of cognitive
function. There are, however, theoretical reasons suggesting that therapy with more than one drug may sometimes be appropriate.'5 Of the principal anticonvulsant drugs in common use, measurement of the serum level is appropriate for phenytoin and carbamazepine. These have both been shown to be effective drugs in the treatment of partial and tonic-clonic seizures.'6 Phenobarbitone, and primidone, which is partly metabolized to phenobarbitone, can readily be measured in the serum, but receptor tolerance means that the serum levels do not necessarily reflect biological activity. The serum level of sodium valproate, being bound to tissue organelles, may not reflect the day to day anticonvulsant activity of the drug. Although a number of new drugs have recently been introduced, or are at an advanced stage of trial in the management of epilepsy, there has been no major 'breakthrough' of the type for which people with epilepsy and their neurologists are constantly hoping. Vigabatrin'7 has recently been introduced as it has been shown to be efficacious in add-on trials of patients with resistant epilepsy. Vigabatrin is an inhibitor of gamma-aminobutyric acid (GABA) transaminase. In one trial'8 more than 50% of those with resistant epilepsy showed a reduction in seizure frequency of more than 50%. Gabapentin is a GABA-related amino acid. Although possibly rather less effective than vigabatrin, a significant benefit to about 25% of patients has been shown in one trial.'9 Oxcarbazepine is the 10-keto analogue of carbamazepine. Its 10-hydroxy metabolite is responsible for the anticonvulsant effect. Placebo-controlled trials show a similar efficacy to carbamazepine, but with a lower allergic propensity, and with fewer side effects on the nervous system.20 Lamotrigine probably exerts its anti-epileptic effect by inhibiting the release of excitatory amino-acid neurotransmitters. It is also an effective supplementary drug in some poorly controlled epileptic patients. The investigation of seizures
This is a third area in which there have been considerable advances during John Walton's professional lifetime. When I trained as a registrar at Queen Square in the 1960s, it was not uncommon for patients who reported a single tonic-clonic seizure to be investigated either by air encephalography or by arteriography, or sometimes by both. These activities now seem extraordinary in the light of more sophisticated epidemiological evidence referred to above, but the perspective of neurologists then was very much that of neurosurgeons. About 54% of all patients with brain tumours have seizures at the time of assessment and diagnosis.22
EPILEPSY
However, this figure is meaningless when translated to the population of people with first seizures who have a much smaller risk of a brain tumour about 4% in one study.'0 Electroencephalography (EEG) has also come under increasingly critical review. A brief 25 or 30 minute paper recording will be normal in approximately 50% of patients with epilepsy, although repeated recordings increase the proportion of patients with abnormal records.23 The principal, indeed almost the only plank in the diagnostic platform of epilepsy, is the historical account given by the patient and bystanders. EEG certainly has a role in those with multiple seizures, as, if an abnormality is found, it may be sufficiently clear to diagnose accurately a seizure type, thereby influencing the choice of anticonvulsant medication. In adult practice, this means distinguishing between primary generalized epilepsy with typical 3 cycle per second spike and wave activity, and those who have an inter-ictal spike suggesting the probability that the seizure has a partial onset. It is obviously extremely helpful to obtain the record of a seizure discharge during an overt clinical seizure, but unless seizures are very frequent, this is seldom practical. However, EEG monitoring accompanied by simultaneous video monitoring has had a major impact in evaluating the origin of the seizure discharges in those with very frequent seizures who are candidates for surgery.24 Video and EEG monitoring has also revolutionized the diagnosis of simulated seizures (pseudoseizures), which account for a significant proportion of those with 'epilepsy' apparently intractable to all types of anticonvulsant medication.25 When John Walton began his neurological career, imaging studies available for epilepsy were insensitive and invasive. A skull X-ray might have shown intracranial calcification, for example, in the case of a slowly growing oligodendroglioma, or the classical tram-lines appearance of vascular calcification in the rare Sturge-Weber syndrome. As noted above, over-investigation of people with epilepsy for suspected tumours led to large numbers of air encephalograms and carotid arteriograms. The introduction of isotope brain scanning in the early 1960s did reduce these invasive tests considerably, but it was not until Hounsfield's invention of cranial computerized tomographic (CT) scanning that a sensitive and non-invasive study for the detection of abnormalities of cerebral structure became available. Since then, magnetic resonance imaging (MRI) has proved more sensitive than CT scanning in the detection ofcerebral lesions related to the genesis of seizures. MRI will detect small temporal lobe haematomas and tumours responsible for complex seizures and increasing technical advances are more frequently demonstrating mesial temporal sclerosis, the most common pathological finding on those who have
525
temporal lobectomies for temporal lobe epilepsy. MRI is only available in a few centres in the United Kingdom, and CT head scanners, although now widely available geographically, are always busy. Furthermore, the technician time and capital cost of the machine means that some sort of policy should be developed for deciding whom ofall those patients with new onset epilepsy should have a CT scan. Scheuer and Pedley26 take the view that some sort of brain imaging studies should be performed in all patients over the age of 18 years, and in all children with partial seizures (except those with benign focal epilepsy of childhood), those with abnormal findings on the neurological examination, or focal slow-wave abnormalities on the electroencephalogram. However, in the Royal College of Physicians' prospective study of first seizuresl' my colleagues Charles Clarke, Andrea Garman and I found that of 370 patients with first seizures, only 13 turned out to have brain tumours. Furthermore, scans are not always reliable - subsequent follow-up of our patients found that two further patients whose initial scans had been normal subsequently developed tumours. Two further procedures are now available for research purposes in epilepsy - positron-emission tomography, and single photon-emission CT (SPECT). These techniques allow the imaging of areas of cerebral hypometabolism interictally in the temporal lobes of those who have partial seizures, and may aid a decision about the appropriateness of temporal lobe surgery. In conclusion, although neurophysiological and imaging studies are available for people with epilepsy, few tumours are detected, and some are missed. With increasing demands for the rational use of resources, it is likely that the use of these procedures in the management of epilepsy will decline further in future years. The surgical treatment of epilepsy
This is the fourth of my topics to illustrate the advances made in the management of some cases of epilepsy in John Walton's professional lifetime. It has of course been recognized for nearly 100 years that operations on brain tumours may in themselves relieve epilepsy associated with the tumours. It was then recognized that smaller lesions, particularly in the temporal lobe, could be responsible for epilepsy, and that whereas these lesions did not in themselves warrant removal on account of progressive pathology, removal might have a remarkable effect on the frequency of the seizures, without leaving a major neurological deficit. Colleagues at the Maudsley Hospital headed first by Murray Falconer, and more recently by Charles Polkey have been amongst the international leaders in the
526
A. HOPKINS
field. The development of the technique of block dissection of the anterior part of the temporal lobe developed by Falconer and his colleagues proved to be instrumental in the development of our understanding of the pathological nature of the underlying changes responsible for epilepsy. There has been an enormous investment in clinical and neurophysiological time in attempting to predict those patients who will benefit from surgery. First of all, patients must of course be proven to be refractory to a number of anticonvulsant drugs, and combinations of drugs in dosages sufficient to achieve high therapeutic levels. Interictal scalp EEG recording is of limited value, and some sort of video monitoring with simultaneous EEG recording of the seizures is usually necessary to determine the site of origin of the seizures. Wieser and his colleagues have shown that an electrode inserted through the frame foramen ovale reliably records seizure discharges some of which do not reach scalp electrodes.27 American neurosurgeons have also written of the value of subdural and indwelling depth electrodes. As noted above, MRI and SPECT both have their value in locating the seizure focus to one or other temporal lobe. Histologically, the removed temporal lobe may show mesial temporal sclerosis, or small slowgrowing gliomas, or dysembryoplastic neuro-epithelial tumours.28
With careful case selection, more than half of all patients with intractable temporal lobe epilepsy after operation have no further seizures, and about a further third have a significant reduction in seizure frequency. There are, however, psychological effects that should not be underestimated, particularly in terms of verbal memory in left-sided operations, which may prove a significant handicap to those whose life and careers demand verbal skills. I have chosen to concentrate upon four aspects of epilepsy, but others are worthy of brief mention. There is now much better understanding of the biology of the epileptic focus,29 much better understanding ofthe relationship between febrile convulsions and epilepsy,30 much better understanding of the toxic effects of drugs, much better understanding of the frequency of pseudoseizures, much better understanding of the risks of post-traumatic epilepsy, and of the difficulty in preventing it,31 and finally, with the recent publication ofthe MRC trial on withdrawing anticonvulsants, we can reasonably hope to have better understanding of the risks and benefits of withdrawing anticonvulsants from those who have been free from seizures for 2 years or more. There remains, however, a substantial shortfall of funds for and people interested in research in epilepsy.33
References 1. Hauser, W.A. & Kurland, L.T. The epidemiology of epilepsy in Rochester, Minnesota 1935 through 1967. Epilepsia 1975, 16: 1-66. 2. Pond, D.A., Bidwell, B.H. & Stein, L. A survey of epilepsy in fourteen general practices. I. Demographic and medical data. Psychiat Neurol Neurochir 1960, 63: 217-236. 3. Gudmundsson, G. Epilepsy in Iceland: a clinical and epidemiological investigation. Acta Neurol Scand 1966, 43 (Suppl 25): 1-24. 4. Crombie, D.L., Cross, K.W., Fry, J. et al. A survey of the epilepsies in general practice: a report by the Research Committee of the College of General Practitioners. Br Med J 1960, 2: 416-422. 5. Goodridge, D.M.G. & Shorvon, S. Epileptic seizures in a population of 6,000. I. Demography, diagnosis and classification, and role of the hospital services. Br MedJ 1983, 287: 641-644. 6. Verity, C.M., Butler, N.R. & Golding, J. Febrile convulsions in a national cohort followed up from birth. I. Prevalence and recurrence in the first five years of life. Br Med J 1985, 290: 1307- 1310. 7. Annegers, J.F., Hauser, W.A. & Elveback, C.R. Remission of seizures and relapse in patients with epilepsy. Epilepsia 1979, 20: 729-737. 8. Rodin, E.A. The Prognosis of Patients with Epilepsy. Thomas, Springfield, Illinois, 1968. 9. Department of Transport (UK). Driving Licence Regulations 1989. 10. Hopkins, A., Garman, A. & Clarke, C. The first seizure in adult life. The value of clinical features, electroencephalography and computerised tomographic scanning in prediction of seizure recurrence. Lancet 1988, 1: 721-726.
11. Hauser, W.A., Rich, S.S., Annegers, J.F. & Anderson, V.E. Seizure recurrence after a first unprovoked seizure: an extended follow up. Neurology 1990, 40: 1163-1170. 12. Hauser, W.A., Anderson, V.E., Loewenson, R.B. & McRoberts, S.M. Seizure recurrence after a first unprovoked seizure. N Engl J Med 1982, 307: 522- 528. 13. van Donselaar, C.A., Geerts, A.T. & Schimsheiner, R.-J. Idiopathic first seizure in adult life: who should be treated. Br Med J 1991, 302: 620-623. 14. Richens, A. Drug Treatment ofEpilepsy. Kimpton, London, 1976. 15. Hopkins, A., Davies, P. & Dobson, C. Mathematical modelling of patterns of intervals between seizures. Arch Neurol 1985, 42: 463-467. 16. Mattson, R.H., Cramer, J.A., Collins, J.F. et al. Comparison of carbamazepine, phenobarbital, phenytoin and primidone in partial and secondarily generalised tonic-clonic seizures. N Engl J Med 1985, 313: 145- 151. 17. Reynolds, E.H. Vigabatrin: rational treatment for chronic epilepsy. Br Med J 1990, 300: 277-278. 18. Treiman, D.M. Gamma vinyl GABA: current role in the management of drug resistant epilepsy. Epilepsia 1989, 30 (Suppl 3): S31-35. 19. UK Gabapentin Study Group. Gabapentin in partial epilepsy. Lancet 1990, 335: 1114-1117. 20. Editorial. Oxcarbazepine. Lancet 1989, 2: 196-198. 21. Jawad, S., Richens, A., Goodwin, G. & Yuen, W.C. Controlled trial of lamotrigine (Lamictal) for refractory partial seizures. Epilepsia 1989, 30: 356-363. 22. Thomas, D.G.T. & Graham, D.I. Brain Tumours: Scientific Basis, Clinical Investigation and Current Therapy. Butterworths, London, 1980.
EPILEPSY 23. Salinsky, M., Kanter, R. & Dashieff, R.M. Effectiveness of multiple EEGs in supporting the diagnosis of epilepsy: an operational curve. Epilepsia 1987, 28: 331-334. 24. Gotman, J., Ives, J.R. & Gloor, P. (eds). Longterm monitoring in epilepsy. Electroencephalogr Clin Neurophysiol 1985, Suppl 37: 327-340. 25. Hopkins, A. Pseudoseizures. Q JMed 1989, NS 71:473-475. 26. Scheuer, M.L. & Pedley, T.A. The evaluation and treatment of seizures. N Engl J Med 1990, 323: 1468-1474. 27. Wieser, H.G., Elger, C.E. & Stodieck, S.R.G. The foramen ovale electrode: a new recording method for the pre-operative evaluation of patients suffering from medio-basal temporal lobe epilepsy. Electroencephalogr Clin Neurophysiol 1985, 61: 314-322.
527
28. Editorial. Lessons in the surgical treatment of epilepsy. Lancet 1989, 2: 366-367. 29. Meldrum, B.S. Anatomy, physiology and pathology of epilepsy. Lancet 1990, 336: 231-234. 30. Annegers, J.F., Hauser, W.A., Shirts, S.B. & Kurland, L.T. Factors prognostic of unprovoked seizures after febrile convulsions. N Engi J Med 1987, 316: 493-498. 31. Hauser, W.A. Prevention of post-traumatic epilepsy. N Engl J Med 1990, 323: 540-541. 32. MRC Antiepileptic Drug Withdrawal Study Group. Randomised study of antiepileptic drug withdrawal. Lancet 1991, 337: 1175-1180. 33. Shorvon, S.D. The lack of funds for clinical epilepsy research in the UK. J R Coll Phys Lond 1991, 25: 31-32.
