If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The outpatient services at the Epilepsy Unit in the Western Infirmary, Glasgow, Scotland was set up in September 1982. From the outset patient data were collected prospectively. A focused approach to patients with newly diagnosed epilepsy was developed and a series of 4 analyses have been undertaken over the intervening years, with results from the latest still being written up for publication. A total of 16 published papers have described patient outcomes over the years, focusing on response to different drug schedules. A number of factors contributing to a poorer prognosis has been identified and follow up data over 30 years has confirmed the lack of overall improvement in prognosis despite the introduction of 14 new AEDs for the common epilepsies in the UK with different mechanisms of action over this time. Patterns of response have confirmed that a majority of patients will go into remission with around 25% of the population appearing to have refractory epilepsy de novo. Since all available options are antiseizure and not antiepilepsy drugs, some patients, who are initially well controlled, are seen to relapse over time and to develop refractory epilepsy. A new approach in identifying and treating epileptogenesis is necessary, if this disappointing scenario is to be reversed with the next generation of antiepileptic drugs.
The journey of a thousand miles begins with a single step.Lao Tzu 500 BC
My research journey in epilepsy began with a single idea. When I set up my seizure and epilepsy clinics in September 1982, I subsequently collected data from every patient using a “pink folder” system including a standard data collection form and investigative protocol. All subsequent information has been collected prospectively. A dedicated telephone line was made available to patients, families, and their primary care physician to facilitate optimal management. Patient data were included in a computer database and in pink folders which were stored in metal cabinets. The appropriate pink folders accompanied my team to every clinic and were available when patients, relatives and general practitioners phoned the Epilepsy Unit. Over the next 34 years, each new patient was registered in our database and a folder developed for storage. There are currently 8068 pink folders in the system.
One of the first patients referred to my service in 1982 was a 44 year old man reporting more than 10 focal and generalized tonic-clonic seizures each month despite taking high doses of 4 antiepileptic drugs (AEDs). He was markedly ataxic and was helped into the consulting room by his elderly parents. I remember thinking “how did he get into this state” and could I have done better for him? This unfortunate individual was the inspiration for the subsequent outcome work emanating from our newly diagnosed epilepsy programme over the next 35 years!
2. Database
Once the decision was taken to focus our outcome programme on adolescents and adult patients with newly diagnosed epilepsy, a letter was sent to all general practitioners in the West of Scotland offering to review suitable patients reporting a first seizure or with likely untreated epilepsy within a few weeks of referral. A direct line to the Epilepsy Unit office was set up to expedite the review of urgent cases. Around the same time an audit of patients presenting to the casualty department at the Western Infirmary with a first seizure or untreated epilepsy was made and the results published in the British Medical Journal [
]. Most of these patients were sent home with no investigations or follow up arranged. Hence, a direct referral arrangement was established with the Epilepsy Unit, which initially bypassed the patient’s general practitioner. Lastly, our epilepsy nurse specialists reviewed patients admitted to the emergency and general medical wards with untreated seizures or epilepsy. Appropriate investigations were arranged and rapid referral to our clinic was organised as appropriate. All clinical information available from these patients was included in our database.
A summary of top line outcomes in the 4 major analyses undertaken over the past 20 years is shown in Table 1. Seizure freedom was defined as having no seizures for at least the previous year. Each series of analyses will now be described in detail. Modern AEDs licensed for adjunctive use and as monotherapy in the UK for adolescents and adults during the time of all 4 analyses are listed in Table 2. Most of the analyses used the older classification of seizures and syndromes published in 1989 by the International League against Epilepsy (ILAE) [
Commission on classifications and terminology of the International League against Epilepsy: proposal for revised classification of epilepsies and epileptic syndromes.
The first set of analyses was undertaken in 1999 focusing on 470 previously untreated patients with newly diagnosed epilepsy (median age 29 years, range 9–93 years). The first outcome paper was published in the New England Journal of Medicine in 2000 and has now been cited more than 3000 times [
]. The headline results are outlined in Table 1. Most of the patients were originally referred by their primary care physicians with just 8% coming from the hospital accident and emergency department. Response to first, second or third AED schedule were 50%, 13% and 1% respectively. Only 3% of this patient population became seizure free on polypharmacy. Interestingly, among patients who did not respond to the first AED, the percentage who subsequently became seizure free was much smaller when failure was a consequence of lack of efficacy (11%) than due to poor tolerability (41%) or an idiosyncratic drug reaction (55%).
There was a significant linear trend in the proportion of patients with uncontrolled epilepsy relative to an increasing number of pre-treatment seizures (p < 0.001). Among the patients with inadequate seizure control on the first well tolerated AED, those who received a second monotherapy (n = 35; 17%) had similar seizure free rates to the those treated with duotherapy (n = 42; 26%) [
]. More of the latter patients became fully controlled when the combination involved a sodium channel blocker with a drug possessing multiple mechanisms of action (30%) compared with other combinations (7%, p = 0.05). Most of the patients attaining complete control of their seizures with their first AED did so at modest or moderate dosing [
]. The commonest daily dosage ranges were 400–600 mg for carbamazepine, 600–1000 mg for sodium valproate and 125–200 mg lamotrigine. Most withdrawals due to poor tolerability also occurred at or below these dose levels (carbamazepine 98%, sodium valproate 100%, lamotrigine 75%).
4. Second analyses
The second set of analyses involving 780 patients was undertaken in 2003 [
]. Overall, 64.4% of this population remained seizure free, with 5.4% controlled on 2 or more AEDs. Response rates with the first, second and third treatment schedule were 50.4%, 10.7% and 2.5%, respectively, with only 0.8% patients becoming seizure free with subsequent AED trials. The prognosis was marginally better in patients with idiopathic (n = 222, 66% seizure free) than cryptogenic (n = 314, 57% seizure free) or symptomatic (n = 244, 56% seizure free) epilepsies. Outcomes versus total pre-treatment seizure numbers, duration of epilepsy prior to starting treatment, and seizures reported in the 3 months before starting treatment are illustrated in Fig. 1. This last observation demonstrated a significant association with increasing seizure density (p < 0.001). This relationship was also observed with seizure numbers 6 and 12 months before the diagnosis of epilepsy was made and treatment started. Interestingly, total seizure numbers prior to treatment initiation did not predict outcome [
Fig. 1Outcomes by (a) number of pre-treatment seizures, (b) duration of epilepsy prior to starting treatment and (c) seizure numbers in the 3 months before starting treatment. Numbers within bars represent percentages. Percentages on top of bars represent patients with refractory epilepsy (uncontrolled and relapsed).
Prognosis appeared better in patients 65 years of age and above (85% remission, p < 0.001) and in adolescents (65% remission, p < 0.01), compared to the remainder of the population (55% remission) [
]. In patients failing initial monotherapy, subsequent response to a combination of 2 AEDs (27%) was no different from that with alternative monotherapy (32%).
Fig. 2Daily doses of carbamazepine, lamotrigine and sodium valproate in patients achieving remission.
]. Many patients in each category, including some with mesial temporal lobe epilepsy, had no further seizures after starting AED therapy. Patients with post-traumatic epilepsy had the worst outcomes (35% seizure free), while those with underlying cerebral atrophy (71% seizure free) or cerebrovascular disease (70% seizure free) did best. Remission rates in patients with cortical dsyplasis (60%), hippocampal atrophy (50%) and primary brain tumours (52%) appeared little different from those with other focal epilepsies.
Table 3Outcomes in newly diagnosed focal epilepsies.
N
Remission
Immediate responder
Post traumatic epilepsy
65
23 (35%)
13 (20%)
Cerebrovascular disease
63
44 (70%)
25 (40%)
Cerebral atrophy
42
30 (71%)
17 (40%)
Primary brain tumours
25
13 (52%)
5 (20%)
Cortical malformations
15
9 (60%)
5 (33%)
MLTE with hippocampal atrophy
14
7 (50%)
3 (21%)
Vascular lesions
7
4 (59%)
2 (29%)
Immediate responders had no further seizures after starting treatment. MTLE = mesial temporal lobe epilepsy.
]. The seizure free rate with sodium valproate was superior to that with lamotrigine (66% versus 45%, p = 0.073), particularly in patients with juvenile myoclonic epilepsy (75% versus 39%, p = 0.014). Interestingly, a history of febrile convulsions was associated with a reduced likelihood of remission (p = 0.032) in this analysis.
In a separate analysis including all 780 newly diagnosed patients, a number of factors were predictive of pharmacoresistant epilepsy (Table 4) [
]. Probably the most interesting of these was the observation that the presence of psychiatric comorbidities increased the likelihood of subsequent refractoriness of the epilepsy (odds ratio 2.17, 95% confidence intervals 1.33–3.55; p = 0.002). Other predictive factors for poorer outcomes, with the possible exception of recreational drug use, were already recognised as potential markers of pharmacoresistance. These included family history of epilepsy, prior febrile seizures, traumatic brain injury and high pre-treatment seizure numbers. Recreational drug use could be a marker for imperfect adherence.
Table 4Predictors of refractory epilepsy in 780 newly diagnosed patients using a multivariate analysis.
Out of these 780 patients, 117 were aged 65 years or older (median age 73 years; range 65–92 years). Overall, 73 (62%) became seizure free for at least 12 months on their first AED with 30 (26%) failing to respond and 14 (12%) not tolerating the treatment [
]. Following pharmacological manipulation, a total of 93 (79%) patients attained remission, with 87 (93%) receiving monotherapy and 6 (7%) on duotherapy. No individual AED was more likely to confer seizure freedom than any other. Patients not attaining remission had a higher pre-treated seizure burden than those subsequently becoming seizure free (p < 0.01). This better prognosis for older people may reflect lower lesional epileptogenicity and genetic predisposition.
A separate analysis of 890 newly diagnosed patients explored mortality data comparing them with a chronic epilepsy cohort and 2 age- and sex-matched Scottish comparison groups [
]. Mortality data were obtained from the General Registry Office for Scotland and causes of death ascertained from death certificates and primary care and health authority records. Newly diagnosed patients had a 42% increase in standard mortality ratios (SMR) compared with the control groups (SMR 1.42, 95% confidence intervals 1.16–1.72). Increased mortality was recorded in patients who had not responded to treatment, with no increase in risk observed in those attaining seizure freedom. In the chronic epilepsy cohort, there was more than double the expected number of deaths (SMR 2.05, 95% confidence intervals 1.83–2.26). The incidence of sudden unexpected death in epilepsy (SUDEP) was 1.08 and 2.46 per 1000 patient years in the newly diagnosed and chronic epilepsy cohorts, respectively. The greatest excess in mortality was reported in patients younger than 30 years [
]. Median follow up was 7.2 years (range 2–26 years). The overall remission rate was 68.4% with 6.4% of this population remaining seizure free on more than one AED (Table 1). Four specific patterns were identified. Pattern A patients (n = 408, 37.2%) became seizure free immediately or within 6 months of starting treatment continuing in remission throughout follow up, while pattern B patients (n = 246, 22.4%) showed a delayed response for more than 6 months after starting treatment before subsequently becoming seizure free throughout follow-up. In some of these patients the problem was identifying the most suitable drug and in others some time required to pass before the patient fully accepted the diagnosis and was wholly adherent with the drug regimen. Pattern C included patients who demonstrated fluctuation around periods of seizure freedom lasting at least a year followed by relapses (n = 172, 15.7%). Most had 1 (n = 140) or 2 (n = 17) such episodes, but a few, had 3, 4 or even 5 relapses. Patients with pattern D (n = 272, 24.8%) never attained seizure freedom for any 12 month period, although numbers and severity of seizures may have been improved in some. Overall seizure freedom was attained by only 52% of the 188 patients, who were followed up for more than 10 years.
Most patients who became seizure free did so on their first (49.6%), second (13.3%) or third (3.7%) AED schedule, although a few patients (1.8%) became fully controlled on their 4th, 5th, 6th or even 7th regimen, a few of these on monotherapy. There was a higher probability of attaining seizure freedom in patients receiving 1 compared to 2 drug regimens, and 2 compared to 3 regimens (p < 000.1) [
]. The difference was greater among patients with symptomatic or cryptogenic than idiopathic epilepsy. Some of the uncontrolled patients did not receive further drug schedules after failing 2 or 3 trials and so there was potential for subsequent improvement in their outcomes. There were a range of factors contributing to the reason why many of these patients did not become seizure free [
Uncontrolled epilepsy is not necessarily the same as drug-resistant epilepsy: differences between populations with newly diagnosed and chronic epilepsy.
]. This observation also applied to the second regimen.
6. Fourth analyses
The last set of analyses, including 1795 newly diagnosed patients, is still ongoing, and the results have not yet been finalised and submitted for publication [
]. The cut off point was 31st October 2014 allowing, as with all the other data sets, a minimum follow up time of 2 years. The median follow-up period was 11 years (interquartile range 7–16 years). Epilepsy was classified as focal in 78% of patients and generalised in the remaining 22% following the newer ILAE classification system [
]. Perhaps counter-intuitively, the seizure free rate with this analysis was 64%, which is identical to that from the first analysis undertaken nearly 20 years before (Table 1). The drop in the number of patients controlled on one AED (55%) and the rise in those responding to combinations (9%) can be explained by the natural history of treated epilepsy with some patients responding to the addition of mechanistically dissimilar drugs, while others, who were previously controlled, relapsed over time [
]. Seizure free rates with the first 7 successive AED schedules in this analysis were 46%, 12%, 4%, 0.6%, 0.5%, 0.3% and 0.6% respectively, mirroring the results from all previous analyses [
]. Patients who did not achieve at least 1 year of seizure freedom on their first AED had 1.73 odds of uncontrolled epilepsy for each subsequent AED schedule tried (95% confidence intervals 1.56–1.91).
7. Discussion
A number of observations on the prognosis of epilepsy in adolescents and adults have been made from the database from the Western Infirmary in Glasgow, Scotland over the past 34 years from 4 series of analyses. The first paper published in the New England Journal of Medicine in 1990, is still widely cited [
]. This short review summarises the results from this and all subsequent publications include some as yet unpublished data. Important observations include the optimal response to the first, second and perhaps third regimen which contributed to the development of the ILAE definition of pharmacoresistant epilepsy [
]. This was rapidly followed by the observation that many if not most seizure free patients with newly diagnosed epilepsy will respond to a modest or moderate dose of any appropriate AED [
]. Another important observation was the relevance of high seizure density up to a year before the diagnosis in predicting subsequent pharmacoresistant epilepsy.
A further interesting finding was the contribution of psychiatric comorbidities to poorer outcomes in newly diagnosed epilepsy [
]. The patterns of response to AED therapy clearly delineates 2 major populations of patients with around 60% who are easy to control and 25% who are pharmacoresistant de novo [
]. The remainder have relapsing/remitting epilepsy, which is responsible for the observation that with time fewer patients appear controlled on monotherapy with an increasing greater number responding to AED combinations. Overall patients with generalized epilepsy tended to do better than those with focal seizures. The bottom line to this series of observational studies is that overall outcomes in adolescents and adults with newly diagnosed epilepsy have changed little over the past 30 years despite the introduction in the UK of 14 AEDs with a range of disparate mechanisms of action [
]. The reason is likely to be due to the way drugs are developed in animal seizure models. They look chemically and mechanistically different but were designed to do the same thing!
8. Conclusions
Has the probability of achieving seizure freedom increased significantly in the last 2 decades? Regrettably, the answer is “no”. Quo vadis? What we need to develop are antiepileptogenic drugs and not antiseizure agents. Seizure data should be collected prospectively ideally using an objective electronic counting system [
]. Lastly, all patients should have genome analysis undertaken to look for markers of response to individual AEDs with distinct modes of action.
Conflict of interest
Martin Brodie serves on the scientific advisory boards of Eisai Ltd., UCB Pharma, GlaxoSmithKline, Lundbeck, Bial, GW Pharmaceuticals and Takeda. He is on the speakers’ bureau for Eisai Ltd., UCB Pharma, GlaxoSmithKline, Lundbeck, Sanofi Aventis, and Abbott and has accepted travel grants for scientific meetings from Eisai Ltd., UCB Pharma and Lundbeck.
Commission on classifications and terminology of the International League against Epilepsy: proposal for revised classification of epilepsies and epileptic syndromes.
Uncontrolled epilepsy is not necessarily the same as drug-resistant epilepsy: differences between populations with newly diagnosed and chronic epilepsy.