• First epileptic seizure;
  • Newly diagnosed epilepsy;
  • Epileptic syndromes;
  • Risk factors;
  • Therapy


  1. Top of page
  2. Abstract
  5. First unprovoked seizures (n = 926)
  8. Appendix

Summary:  Purpose: We describe first unprovoked seizures and newly diagnosed epilepsies at initial presentation, with a special emphasis on epilepsy syndromes, in a large cohort recruited in the mid-1990s in France.

Methods: The French Foundation for Research on Epilepsy set up a network to conduct a prospective study of patients with newly diagnosed unprovoked seizures. Information was provided by 243 child or adult neurologists. Four neurologists classified each case according to the International League Against Epilepsy (ILAE) criteria. First-seizure patients and patients with previously undiagnosed seizures were compared.

Results: Between May 1, 1995, and June 30, 1996, 1,942 patients aged from 1 month to 95 years were identified: 926 (47.7%) with a single seizure and 1,016 (52.3%) with newly diagnosed epilepsy. All but 17 patients had EEGs. In the first-seizure and newly-diagnosed-epilepsy groups, neuroimaging studies were performed in 78.2 and 68.3% of patients, and medication prescribed in 54.1 and 89.6%, respectively. There were significant differences between the two groups with respect to age at onset and diagnosis, sex, etiology, several specific syndromes, as well as the type and presentation of initial seizure. In patients for whom the first seizure was convulsive, only sex, multiple seizures in a day or status epilepticus, and cryptogenic localization-related syndrome differed between the two groups.

Conclusions: Approximately half of patients who first came to attention for an unprovoked seizure already met epidemiologic criteria for epilepsy. There were significant differences between the types of patients with a first seizure and those with newly diagnosed epilepsy. One or several seizures at diagnosis did not influence the diagnostic assessment of the patients but had a strong influence on the initiation of treatment.

For a long time, epileptic disorders suffered from lack of precise definitions, convenient and reliable laboratory investigations, and epidemiologic data. Important changes occurred in the last decade. The International League Against Epilepsy (ILAE) proposed a classification of epilepsies (1) that allows identification of coherent entities, which, in turn, provide information about the expected course of the disorder and may help to determine appropriate courses of treatment (i.e., choosing among a dozen available drugs and surgical procedures). An accurate diagnosis is important and requires better knowledge on the part of the clinician and the use of more sensitive ancillary tests, in particular, an EEG and, in many instances, neuroimaging procedures. The ILAE proposed guidelines for epidemiologic studies (2) and several recent studies allow us to estimate the incidence and prevalence of seizure disorders and epilepsy specifically (3), as well as to appreciate their worldwide distribution (4,5). The heterogeneity of the epilepsies remains an issue (6).

The French Foundation for Research on Epilepsy has set up a nationwide network of child and adult neurologists (Observatoire de l'Epilepsie) to perform a multicenter, prospective, and observational study (CAROLE; i.e., Coordination Active du Réseau Observatoire Longitudinal de l'Epilepsie) of patients with newly diagnosed unprovoked seizures in the mid-1990s. This included patients with either their first unprovoked seizure or patients newly diagnosed with epilepsy who had a history of previously unrecognized or undiagnosed seizures.

This report focuses on characteristics of the cohort at the time of initial diagnosis, with a special emphasis on the use of the International Classification of Epileptic Syndromes (1). Two groups were compared with regard to clinical characteristics, initial presentation, diagnostic procedures, and management: first unprovoked seizures versus newly diagnosed epilepsy.


  1. Top of page
  2. Abstract
  5. First unprovoked seizures (n = 926)
  8. Appendix

Case ascertainment

As a result of personal contacts, 22 specialists throughout France and one in Geneva (Switzerland) agreed to coordinate the study. Over a 6-month period in 1994, regional meetings were organized, and 430 specialists were contacted. Among them, 243 volunteered for the study (list attached), including 43 child neurologists (80% of them with a hospital setting), nine pediatricians with special expertise in neurology, and 191 adult neurologists (127 had a hospital-based practice, 34 had a private practice, and 82 had a combined private and hospital-based practice). In hospitals, physicians included both in- and outpatients. In private practice, most specialists provided care for children as well as adults, but rarely infants (i.e., younger than 1 year), who are typically seen in a hospital setting.

Physicians were asked to refer to the study any patient older than 1 month with at least one unprovoked epileptic seizure first diagnosed between May 1, 1995, and June 30, 1996, and who was likely to be followed up during a minimum of 2 years. Patients with a previous clear diagnosis of unprovoked epileptic seizure(s) were excluded, as were patients with acute symptomatic seizures and patients who were considered unlikely to comply with follow-up procedures (e.g., alcoholic, homeless). We believe, but cannot confirm, that the recruitment was complete.

Index seizure was defined as the seizure that led to the diagnosis. It could be a single seizure in a day, multiple seizures occurring in a 24-h period, or status epilepticus. Status epilepticus was defined as a single seizure of >30-min duration or a series of seizures that lasted over a >30-min period during which function was not fully regained between ictal events. Multiple seizures in a day and an episode of status epilepticus were considered a single episode (2). A case of a single seizure or multiple seizures in a day and status epilepticus that day was coded as status. For this report, the first seizure patients who experienced subsequent seizures during the inclusion period remained in the first seizure group.

The investigators completed a case registration form (CRF) that contained a series of structured closed-ended questions and additional room for comments. Open comments were allowed to support conclusions or observations and to elaborate on information not adequately covered by the preformatted questionnaire. The questionnaire covered the medical history, family history, date, circumstances surrounding the index seizure, duration and description of this seizure, as well as previous seizures when present. The ICBERG questionnaire (7) was used for obtaining adequate descriptions of seizures to permit classification according to ILAE criteria (8). In the mid-1990s, the semistructured interview for seizure classification was not available in France (9). Other items were physical examination, results of EEG recordings and brain imaging, referral patterns after the index seizure, and sociodemographic data.

The CRFs were sent to a consultant society in medical economy (CEMKA) where all data were coded and computerized. Based on medical baseline records, which included the results of tests that were ordered at the time of diagnosis as well as the initial information gathered by the investigators, classifications were performed independently by each member of the study panel (four neurologists) and then compared. In case of insufficient data at inclusion, additional information was requested to assure adequate classifications. In case of interrater disagreement, cases were discussed in conference, and a consensus classification was reached in all cases.

Confidentiality was maintained according to all applicable laws. The study received approval from the French National Ethics Committee (Commission Nationale Informatique et Liberté).


Seizures were classified according to the International Classification of Epileptic Seizures (8), based on clinical manifestations and on data from ancillary tests. Thus a secondarily generalized partial seizure could be either a generalization after a simple or complex partial seizure or an apparently generalized tonic–clonic seizure (GTCS) when there were EEG or anatomic signs sufficient to indicate localization. Convulsive seizures without any clear evidence of a partial or generalized onset were labeled unclassified seizures. When someone had both myoclonic seizures and GTCS in a single day, the seizure type was classified as myoclonic. For patients who had simple or complex partial seizures and secondarily generalized partial seizures in the same day, the seizure type was classified as partial with secondary generalization.

Etiologic factors were classified into two broad categories according to the Classification proposed by the ILAE Commission on Epidemiology and Prognosis (2). The category of symptomatic unprovoked seizures was subdivided into remote symptomatic seizures owing to conditions resulting in a static brain damage and symptomatic seizures owing to progressive central nervous system disorders. Unprovoked seizures of unknown etiology were classified into two subheadings. Idiopathic etiology refers to syndromes that are characterized by both particular clinical features and specific EEG abnormalities. Cryptogenic etiology is used in reference to unprovoked seizures that do not conform to the criteria for the symptomatic or idiopathic categories.

Syndromes were classified according to the International Classification (1). A single seizure without any EEG or neuroimaging abnormalities was considered an isolated seizure (syndrome 4.2 in the International Classification). This seizure could be partial, generalized, or unclassified. Conversely, a single seizure with epileptiform EEG abnormalities or an epileptogenic lesion on neuroimaging was considered the first seizure of epilepsy and so was classified in an epilepsy syndrome. A documented (by history, examination, or imaging) cerebral insult was mandatory to place a patient in the category of “symptomatic localization-related syndrome” (syndrome 1.2). Focal (lateralized or localizing) EEG abnormalities (that did not conform to the criteria for the idiopathic category) without any documented cerebral insult were classified as a single seizure in the category of “cryptogenic localization-related syndrome” (syndrome 1.3). In the newly-diagnosed-epilepsy group (i.e., patients having experienced more than one seizure), this syndrome encompassed partial seizures with or without EEG focal abnormalities. Symptomatic and cryptogenic localization-related syndromes were also classified with an attempt to define the lobe of seizure origin: temporal, frontal, occipital, parietal, and unlocalized, when it was impossible to get a clear electroclinical correlation or when EEG abnormalities were multifocal. Patients in the category of “undetermined epilepsy” (syndrome 3.2) had either apparently generalized seizures or seizures without unequivocal focal or generalized features and had neither EEG nor imaging specific findings.

Statistical analysis

For categoric variables, the comparison of percentages was made using χ2 analysis or Fisher's exact test. For continuous variables (age at first seizure and age at index seizure), the distributions were not normal. Comparison of means between the two groups (first seizure and newly diagnosed epilepsy) was therefore performed by using the Mann–Whitney rank sum test. The statistical software Statview (version 5.0) was used for statistical analysis.


  1. Top of page
  2. Abstract
  5. First unprovoked seizures (n = 926)
  8. Appendix

A total of 1,953 patients was entered in the study. Of these, 11 were ineligible (one had a neonatal convulsion, nine had been previously diagnosed, and one had been included twice by two different investigators). In all, 1,942 patients (aged from 1 month to 95 years) with the diagnosis of unprovoked seizure(s) were included in the cohort: 926 (47.7%) with a single seizure and 1,016 (52.3%) with more than one seizure at initial presentation. All but 17 patients had EEGs. One or more imaging studies was performed in 1,418 (73.0%) of the patients. Drug therapy was prescribed for 1,411 (72.6%) of the patients. At the time of their initial presentation, 61% of the patients were hospitalized. Characteristics of the first-seizure and newly-diagnosed-epilepsy groups are separately shown in each table.

First unprovoked seizures (n = 926)

  1. Top of page
  2. Abstract
  5. First unprovoked seizures (n = 926)
  8. Appendix

The median age (25–75th percentiles) at the time of the first seizure (also the index seizure for this group) was 19 years (9–39 years). Nearly 60% of the patients were male (Table 1).

Table 1.  Baseline characteristics of 1,942 patients with newly diagnosed seizures
CharacteristicSingle seizure at diagnosis (N = 926)>1 seizure at diagnosis (N = 1016)p Value
  • CNS, central nervous system.

  • Median age (and 25–75th percentiles) at first seizure: 19 years (9–39 yr) in single sz group, 15 years (7–29 yr) in >1 sz group. Median age (and 25–75th percentiles) at index seizure: 19 years (9–39 yr) in single sz group, 17 years (8–35 yr) in >1 sz group. p Value for comparison between single seizure and >1 seizure at diagnosis.

  • a

     Median age.

  • b

     Seizure while asleep versus seizure while awake and unknown category.

  • c

     Multiple seizures in a day and status epilepticus versus one seizure in a day.

  • d

     Each category versus all else (e.g., simple partial versus all else, complex partial versus all else.

  • e  χ2 of 2 degrees of freedom (after exclusion of unclassified category).

  • f

     One case of absence status.

  • g

     Two cases of multiple seizures in a day (myoclonic seizures followed by a generalized tonic–clonic seizure).

  • h

     Generalized seizure but undetermined focal or generalized onset.

  • i

     Because of incomplete files.

Age at first seizure    <0.0001a
 1 mo–15 yr37740.754353.4 
 16–24 yr15516.716816.5 
 25–59 yr30633.021721.4 
 >60 yr889.5888.7 
Age at diagnosis     
 1 mo–15 yr37740.745745.00.047a
 16–24 yr15516.718618.3 
 25–59 yr30633.026926.5 
 >60 yr889.510410.2 
Sex    0.0001
First seizure     
 Sleep/awake state    <0.0001b
  Seizure while awake69474.984683.3 
  Seizure while asleep22624.415615.4 
 Number    <0.0001c
  One seizure in a day76382.4100498.8 
  Multiple seizures in a day12313.390.9 
  Status epilepticus404.330.3 
   Secondarily generalized31033.515715.5<0.0001d
   Infantile spasm00272.7 
Etiology    0.0003e

Partial seizures occurred in 46.2% of the group, generalized seizures in 31.9%, and unclassified seizures (generalized convulsive seizures but undetermined focal or generalized onset) in 21.9%. GTCS (31.2%) were the most common of the generalized seizures. There were only three patients who had absence or myoclonic seizures in the first-seizure group. All three reflected unusual situations, one individual with absence status epilepticus and two patients with myoclonic seizures followed by a GTCS in a single day.

Based on the initial diagnosis and evaluation, 228 (24.6%) patients had idiopathic etiology, and 535 (57.8%) had cryptogenic etiology. The remaining 163 (17.6%) had symptomatic etiology. Of these, the most frequent risk factors were pre- and perinatal insult (4.4%), cerebrovascular disease (3.9%), and head injury (3.2%) (Table 2).

Table 2.  Distribution of risk factors in symptomatic unprovoked seizures
Risk factorsSingle seizure at diagnosis (N = 926)>1 seizure at diagnosis (N = 1016)
  1. CNS, central nervous system.

Symptomatic of static conditions   (remote symptomatic)14515.715615.4
 Head injury303.2282.8
 Cerebrovascular disease363.9313.1
 CNS infection or infestation50.570.7
 Pre- and perinatal414.4363.5
  Malformation of brain161.7161.6
  Neonatal encephalopathy91.060.6
  Mental retardation161.7141.4
 Alcohol related30.330.3
 Postencephalopathic states0020.2
 More than one of these10.120.2
Symptomatic of progressive   CNS disorders181.9222.2
 CNS tumors161.7111.1
 Autoimmune diseases0010.1
 Metabolic diseases0010.1
 Degenerative diseases10.170.7

A specific epilepsy syndrome was assigned for slightly more than half of the patients with a single seizure: 80 (8.6%) idiopathic localization-related, 149 (16.1%) symptomatic localization-related, 93 (10.0%) cryptogenic localization-related, 148 (16.0%) idiopathic generalized, 13 (1.4%) symptomatic generalized, and two (0.2%) undetermined (Table 3). The remaining 441 (47.6%) were classified as cases of isolated seizure. Of these, three patients had neither EEG nor neuroimaging.

Table 3.  Distribution of epileptic syndromes
SyndromeSingle seizure at diagnosis (N = 926)>1 seizure at diagnosis (N = 1016)p Value
  • p Value for comparison between single seizure and >1 seizure at diagnosis: each group versus all else (e.g., idiopathic localization related versus all else, symptomatic localization related versus all else).

  • C/T, centrotemporal; GTCS, generalized tonic–clonic seizure.

  • a

     One case of absence status.

  • b

     Two cases of multiple seizures in a day (myoclonic seizures followed by a generalized tonic–clonic seizure).

  • c

     Because of incomplete files.

1. Localization related32234.848247.4 
 1.1 Idiopathic808.6484.70.0007
  Benign childhood epilepsy with C/T spikes667.1393.8 
  Childhood epilepsy with occipital paroxysms141.580.8 
  Primary reading epilepsy0010.1 
 1.2 Symptomatic14916.113713.5NS
 1.3 Cryptogenic9310.029729.2<0.0001
2. Generalized16117.434333.8 
 2.1 Idiopathic14816.027827.4<0.0001
  Benign myoclonic epilepsy in infancy0090.9 
  Childhood absence epilepsy00919.0 
  Juvenile absence epilepsy1a0.1303.0 
  Juvenile myoclonic epilepsy2b0.2545.3 
  Epilepsy with GTCS on awakening91.060.6 
  Other idiopathic generalized epilepsies11512.4757.4 
  With specific modes of seizure precipitation212.3131.3 
 2.2 Cryptogenic or symptomatic00393.8 
  West syndrome00252.5 
  Lennox–Gastaut syndrome0000 
  Epilepsy with myoclonic–astatic seizures0030.3 
  Epilepsy with myoclonic absences00111.1 
 2.3 Symptomatic131.4262.6NS
  Nonspecific etiology121.3212.1 
  Specific syndromes10.150.5 
3. Undetermined whether focal or generalized20.217717.4 
 3.1 With both generalized and focal seizures20.240.4 
 3.2 Without unequivocal generalized or focal features0017317.0 
4. Special44147.600 
 4.1 Acute symptomatic (provoked) seizures0000 
 4.2 Isolated seizures44147.600 
9. Unclassified****00141.4 

All patients with an idiopathic (localization-related or generalized) syndrome had at least one EEG (Table 4).

Table 4.  Ancillary tests (at least one EEG; at least one neuroimaging study) and therapeutic policy according to epileptic syndrome
SyndromeSingle seizure at diagnosis>1 seizure at diagnosis
N% EEG% Neuro- imaging% AEDN% EEG% Neuro- imaging% AED
  • AED, prescription of antiepileptic drug; GTCS, generalized tonic–clonic seizure.

  • a

     One case of absence status.

  • b

     Two cases of multiple seizures in a day (myoclonic seizures followed by a generalized tonic–clonic seizure).

  • c

     Three patients without any EEG and neuroimaging.

  • d

     Because of incomplete files.

  • e

     Comparison between single seizure and >1 seizure at diagnosis: EEG (p = 0.46), Neuroimaging (p < 0.0001), AED (p < 0.0001).

1. Localization related32298.879.565.848299.080.987.3
 1.1 Idiopathic8010055.022.54810060.466.7
 1.2 Symptomatic14997.388.689.313797.194.294.9
 1.3 Cryptogenic9310086.065.629799.778.187.2
2. Generalized16110065.264.034399.746.195.6
 2.1 Idiopathic14810063.563.527810040.395.3
  Benign myoclonic epilepsy in infancy0910033.3100
  Childhood absence epilepsy09110013.2100
  Juvenile absence epilepsy1a1001001003010040.0100
  Juvenile myoclonic epilepsy2b10050.01005410050.094.4
  Epilepsy with GTCS on awakening910044.444.4610066.7100
  Other idiopathic generalized epilepsies11510066.166.17510064.093.3
  With specific modes of seizure precipitation2110057.152.41310046.261.5
 2.2 Cryptogenic or symptomatic03910074.497.4
  West syndrome02510092.096.0
  Epilepsy with myoclonic–astatic seizures0310066.7100
  Epilepsy with myoclonic absences01110036.4100
 2.3 Symptomatic1310084.669.22696.265.496.2
  Nonspecific etiology1210083.366.72195.261.995.2
  Specific syndromes1100100100510080.0100
3. Undetermined whether focal or generalized210050.010017799.478.084.2
 3.1 With both generalized and focal seizures210050.0100410075.0100
 3.2 Without unequivocal generalized or focal features017399.478.083.8
4. Special: Isolated seizuresc44198.682.141.70
9. Unclassifiedd01410057.185.7

A neuroimaging study was performed in 78.2% of the patients (computed tomography (CT) scan only, 57.9%; magnetic resonance imaging (MRI) only, 6.5%; CT scan + MRI: 13.8%). This rate varied according to the epileptic syndrome: 55.0% for idiopathic localization-related, 63.5% for idiopathic generalized, 82.1% for isolated seizures, 86.0% for cryptogenic localization-related, and 88.6% for symptomatic localization-related (Table 4).

Drug therapy was prescribed in 54.1% of the patients. This rate also varied according to the epileptic syndrome: 22.5% for idiopathic localization-related, 41.7% for isolated seizure, 63.5% for idiopathic generalized, 65.6% for cryptogenic localization-related, and 89.3% for symptomatic localization-related (Table 4).

Newly diagnosed epilepsy (n = 1,016)

The median age at first seizure was 15 years (7–29 years), and at diagnosis, 17 years (8–35 years). Slightly <50% of the patients were male (Table 1). The time between the first seizure and the diagnosis of epilepsy ranged from 1 day to 52 years, with a median of 7 months (1.5–30 months). Diagnosis was made mainly within the first year (58.7%), and then 12.0% within the second year, 7.6% within the third year, 3.7% within the fourth year, 3.1% within the fifth year, and 14.9% at >5 years after the initial seizure. Reasons for previously undiagnosed seizures included no access to medical care, unawareness of the nature of the events on the part of the patient and hence no effort to have them diagnosed, and events that were brought previously to medical attention but not diagnosed as seizures.

First seizure type was partial in 48.1% of the group, generalized in 39.9%, and unclassified in 12.0%. The index seizure type was partial in 48.2% of the group, generalized in 39.8%, and unclassified in 12%. Seizure types that usually recur before diagnosis (e.g., absences, myoclonic seizures, atonic seizures, and infantile spasms) were present in 248 (24.4%) of the patients. The first seizure and the index event were different in 154 (15.2%) of the patients. For instance, a GTCS occurred in patients with absence seizures or myoclonic jerks (68 cases), and simple or complex partial seizures evolved to secondarily generalized partial seizures (66 cases). In 61 (6.0%) of the remaining patients, the first seizure and the index seizure were the same, but another seizure type occurred between them. In all, 78.8% of patients had one, 19.6% had two, and 1.6% had three different seizure types.

Based on the initial diagnosis and evaluation, 326 (32.1%) patients had idiopathic etiology, and 500 (49.2%) had cryptogenic etiology. Among the remaining, 12 (1.2%) were unclassified because of insufficient or conflicting information, and 178 (17.5%) had symptomatic etiology. Of these, the most frequent risk factors were pre- and perinatal insult (3.5%), cerebrovascular disease (3.1%), and head injury (2.8%; Table 2).

An epileptic syndrome was assigned for 1,002 (98.6%) of the patients: 48 (4.7%) idiopathic localization-related, 137 (13.5%) symptomatic localization-related, 297 (29.2%) cryptogenic localization-related, 278 (27.4%) idiopathic generalized, 39 (3.8%) cryptogenic or symptomatic generalized, 26 (2.6%) symptomatic generalized, and 177 (17.4%) undetermined whether focal or generalized (Table 3). The remaining 14 (1.4%) were unclassified because of insufficient or conflicting information.

All patients with idiopathic (localization-related or generalized) epilepsy had at least one EEG (Table 4).

A neuroimaging study was performed in 68.3% of the patients (CT scan only, 42.9%; MRI only, 12.2%; CT scan + MRI, 13.2%). This rate varied according to the epileptic syndrome: 40.3% for idiopathic generalized, 60.4% for idiopathic localization-related, 65.4% for symptomatic generalized, 74.4% for cryptogenic or symptomatic generalized, 78.0% for undetermined whether focal or generalized, 78.1% for cryptogenic localization-related, and 94.2% for symptomatic localization-related (Table 4).

Drug therapy was prescribed in 89.6% of the patients (Table 4). Only in some of the idiopathic epilepsies was treatment prescribed in <80% of patients, idiopathic localization-related (66.7%), or idiopathic generalized with specific modes of seizure precipitation (61.5%).

First seizure versus newly diagnosed epilepsy

Most differences between the two groups were statistically significant (Table 1). Age at first seizure (p < 0.0001) and age at diagnosis (p = 0.047) were younger in the newly-diagnosed-epilepsy group. There were more male patients (p = 0.0001) and more generalized convulsive (either primarily or secondarily generalized or unclassified whether focal or generalized at the onset) seizures (p < 0.0001) in the first-seizure group. The distribution of the etiology was different between the first-seizure and newly-diagnosed-epilepsy groups (p = 0.0003). The proportion of patients with a symptomatic etiology was virtually identical in the two groups. There were, however, more patients with an idiopathic etiology in the newly-diagnosed-epilepsy group and more patients with a cryptogenic etiology in the first-seizure group.

There were more idiopathic localization-related syndromes (p = 0.0007) in the first-seizure group, more cryptogenic localization-related (p < 0.0001) and more idiopathic generalized (p < 0.0001) in the newly-diagnosed-epilepsy group (Table 3).

Equal proportions of patients in the two groups had EEGs (p = 0.46) regardless of the syndrome (Table 4). Neuroimaging studies were less frequently performed (p < 0.0001) in the newly-diagnosed-epilepsy group, particularly in idiopathic generalized epilepsies (40.3 vs. 63.5% in the first-seizure group). Overall, patients in the newly-diagnosed-epilepsy group were more often treated (p < 0.0001).

We separately analyzed the 1,238 patients with a generalized convulsive seizure (either primarily or secondarily generalized or unclassified whether focal or generalized at the onset) as the first event (Table 5): 803 (86.7%) in the first-seizure group and 435 (42.8%) in the newly-diagnosed-epilepsy group (p < 0.0001). Only four significant differences persisted between the two groups (Table 6): more male patients (p = 0.0048) and more multiple seizures in a day or status epilepticus (p < 0.0001) in the first-seizure group, more localization-related cryptogenic syndromes (p < 0.0001) and more institutions of treatment (p < 0.0001) in the newly-diagnosed-epilepsy group.

Table 5.  Distribution of epileptic syndromes by whether the first seizure was a convulsive generalized seizure (partial secondarily generalized, generalized tonic–clonic, generalized but undetermined focal or generalized onset)
SyndromeSingle seizure at diagnosis>1 seizure at diagnosis
Total NConvulsive seizureTotal NConvulsive seizure
  • GTCS, generalized tonic–clonic seizure.

  • a

     65 GTCS and 108 unclassified seizures (generalized seizures but undetermined focal or generalized onset).

  • b

     83 secondarily generalized seizures, 148 GTCSs, and 190 unclassified seizures (generalized seizures but undetermined focal or generalized onset).

  • c

     Because of incomplete files.

  • d

     Comparison between single seizure and >1 seizure at diagnosis: convulsive seizures versus nonconvulsive (p < 0.0001).

1. Localization related32222469.648215732.6
 1.1 Idiopathic805163.8481837.5
 1.2 Symptomatic14911275.21374532.8
 1.3 Cryptogenic936165.62979431.6
2. Generalized16115696.934310330.0
 2.1 Idiopathic14814598.02788932.0
  Benign myoclonic epilepsy in infancy0900
  Childhood absence epilepsy09100
  Juvenile absence epilepsy1003000
  Juvenile myoclonic epilepsy200541222.2
  Epilepsy with GTCS on awakening9910066100
  Other idiopathic generalized epilepsies115115100756384.0
  With specific modes of seizure precipitation212110013861.5
 2.2 Cryptogenic or symptomatic03900
 2.3 Symptomatic131184.6261453.8
3. Undetermined whether focal or generalized2210017717598.9
 3.1 With both generalized and focal seizures221004250.0
 3.2 Without unequivocal generalized or focal features0173173a100
4. Special: Isolated seizures441421b95.50
9. Unclassifiedc01400
Table 6.  Baseline characteristics of the 1,238 patients of whom the first seizure was a convulsive generalized seizure
CharacteristicSingle seizure at diagnosis (N = 803)>1 seizure at diagnosis (N = 435)p Value
  • G/F, generalized or focal. Median age (and 25–75th percentiles) at first seizure: 22 years (12–40 yr) in single sz group, 19 years (12–36 yr) in >1 sz group. Median age (and 25–75th percentiles) at index seizure: 22 years (12–40 yr) in single sz group, 23 years (14–40 yr) in >1 sz group. p Value for comparison between single seizure and >1 seizure at diagnosis.

  • a

     Median age.

  • b

     Seizure while asleep versus seizure while awake.

  • c

     Multiple seizures in a day or status epilepticus versus one seizure in a day.

  • d

     Chi-square of 2 degrees of freedom.

  • e

     Each category versus all else (e.g., localization-related idiopathic versus all else, localization-related symptomatic versus all else).

  • f

     Generalized seizure but undetermined focal or generalized onset. In the >1 seizure group, the median time between the first seizure and the index seizure (and 25–75th percentiles) was 8.9 months (2.3–37 months). Diagnosis was made mainly (54.3%) within the first year, and then 10.9% within the second, 9.2% within the third, 4.6% within the fourth, 3.5% within the fifth year, and 17.5% >5 years after the initial seizure.

Age at first seizure    NSa
 1 mo–15 yr28635.616537.9 
 16–24 yr14918.69922.8 
 25–59 yr29436.612829.4 
 >60 yr749.2439.9 
Age at diagnosis    NSa
 1 mo–15yr28635.613330.6 
 16–24 yr14918.69822.5 
 25–59 yr29436.615535.6 
 >60 yr749.24911.3 
Sex    0.0048
First seizure     
 Seizure while asleep19524.310524.1NSb
 Multiple seizures in a day or status epilepticus11314.130.3<0.0001c
 Type    NSd
  Partial secondarily generalized30138.615736.1 
  Generalized tonic–clonic29036.115635.9 
Etiology    NSd
 1.1 Localization-related idiopathic516.4184.1NSe
 1.2 Localization-related symptomatic11213.94510.3NSe
 1.3 Localization-related cryptogenic617.69421.6<0.0001e
 2.1 Generalized idiopathic14518.18920.5NSe
 2.2 Generalized cryptogenic or symptomatic0000 
 2.3 Generalized symptomatic111.4143.2 
 3.1 Undetermined with both generalized and focal sz20.220.5 
 3.2 Undetermined without unequivocal G/F features0017339.8 
 4.2 Isolated seizures42152.400 
 At least one EEG79498.943399.5NS
 At least one neuroimaging study63479.033777.5NS
 Antiepileptic drug prescription42452.837586.2<0.0001


  1. Top of page
  2. Abstract
  5. First unprovoked seizures (n = 926)
  8. Appendix

The CAROLE study is the largest of the most recent studies to have applied the current ILAE criteria for seizures (8), syndromes (1), and epidemiologic definitions (2). As this study included both first seizures and newly diagnosed epilepsy, it serves as an important bridge between the first seizure and epilepsy literature and permits direct comparisons, within a single study, between these two types of clinical scenarios. The fact that 52.3% of the patients had already had two or more undiagnosed seizures at diagnosis is in keeping with other literature. In the British National General Practice Study of Epilepsy (NGPSE), only 252 (44.7%) of the 564 patients with definite epileptic seizures were registered at the time of their first seizure (10). In a Swedish prospective incidence study, the initial seizure was the same as the seizure at time of initial diagnosis in 110 (68.8%) of the 160 adults with unprovoked seizures (11). In the Rochester study, time from the first seizure to diagnosis was >6 months in 50% and >2 years in >30%(12). In another Swedish prospective incidence study of newly referred adults, 16% of the patients came to medical attention >1 year after the first epileptic event (13). The issue of delayed diagnosis is considered in another article (14).

Prospective population-based incidence studies are frequently held up as a gold standard for conducting this type of research. We note that the fact that many patients do not present immediately for diagnosis has implications for a potential bias that might occur in prospective incidence studies, such as the Gironde study performed in southwest France (15), in which the inclusion criterion is based on the date of the first seizure and not the date of the diagnosis. In this circumstance, some syndromes could be disproportionately excluded because many diagnosed during the inclusion interval will have had their onset before the interval, yet many whose onset is during the interval will not be diagnosed until after the recruitment period is complete. From a pragmatic standpoint, incidence surveys frequently suffer from small numbers of cases and very small numbers of cases with specific syndromes (15,16), depending on the size of the source population. A survey larger than those already made (3) would be extremely difficult and costly both in time and in money. That explains the attraction of multicenter studies collecting a large number of cases: 3,469 reported in the Lombardy study (17), and 1,195 in the NGPSE (10,18).

To what extent is the CAROLE study representative of patients with newly diagnosed seizures? Of course, our results (or those of most other studies for that matter) cannot be extended to patients not coming to medical attention, and to patients not included because of a dubious follow-up. However, in France as in the United States (19), specialists' visits are not limited to tertiary centers in which an underestimation of mild cases is likely (17,20). Although a population-based approach could have been used, we note that a chief purpose of this study was to examine in detail the seizure types, syndromes, and etiology in this cohort. Even if a panel of highly qualified specialists makes these determinations, such assessments and conclusions are based on the data collected by the practitioners who refer the patients to the study. Nonspecialist practitioners who are unfamiliar with the more recent advances in understanding and classifying seizure disorders may miss important details. The information directly collected by specialists in this study is more likely to provide the clinical details necessary for accurate classification according to ILAE criteria. Because of the French healthcare system (almost all inhabitants benefit from the National Health Insurance), individuals not asking for medical care are few. Once having come to medical attention, it is rare for a patient with a seizure not to request a consultation with a specialist. We suspect that the number of investigators (n = 243) in this study and their various practices (child or adult neurologists, hospital-based and/or private practice) are sufficiently representative of the management of a first diagnosed seizure in children or adults in our country.

One way in which the CAROLE study most likely is not entirely representative of the population at large is in a probable underrepresentation of the elderly. The age distribution at the time of the first unprovoked seizure is clearly different from that observed in incidence studies (21). In France, elderly patients tend to go to general practitioners and geriatricians, as found in the Gironde study (22). With general practitioners as informants, 24% of the patients with definite seizures in the NGPSE were older than 60 years (10,18) versus 9% in this study. The relative underrepresentation of elderly might also contribute to the lower proportion with remote symptomatic etiology, 15 vs. 21% (24% with the cases of neurologic deficit at birth) in the NGPSE (10,18). Remote symptomatic etiology increases with age, as was documented by several studies: 18% in a cohort of children with newly diagnosed epilepsy (19), 26% in a cohort of children with two or more seizures prospectively monitored from the time of their first unprovoked seizure (23), 31% in an Icelandic incidence survey (24), 35% in the Rochester study between 1935 and 1984 (25) and between 1980 and 1984 (16), 39% of the patients with unprovoked seizures in the Geneva study (26), 39% in a Finnish incidence survey (27), and 49% in a Swedish incidence survey (13), both in adults.

A key focus of this study is the epileptic syndrome. Seizure types are simply symptoms and give few indications regarding the possible paradoxic effect of certain antiepileptic drugs (AEDs) (28). They also do not tell one much about the underlying cause and prognosis (29). Classifications were fundamentally based on ILAE proposals (1,2). “An epileptic syndrome is an epileptic disorder characterized by a cluster of signs and symptoms customarily occurring together”(1). However, we do not consider that the terms epilepsy and epileptic syndrome are synonymous (30). Epilepsies are characterized by occurrence of at least two unprovoked seizures (2), whereas a syndromic diagnosis may be possible after a single epileptic event (1). Examples are many: a single sylvian seizure in a child with centrotemporal spikes on the EEG; a nonconvulsive status epilepticus with bilateral spike–waves; and a temporal lobe seizure with a temporal cavernoma. Thus, 485 (52.4%) of the group who presented with a first seizure met the criteria that permitted them to be assigned to an epilepsy syndrome even though they did not yet meet the epidemiologic criteria (2) for epilepsy (more than two seizures).

The use of the ILAE classification system, however, strongly depends on adequate evaluation of the patient. Required tests include, but are not limited to, an EEG, which should be routinely performed, and neuroimaging, for which the criteria are more controversial and tend to vary across countries. The small but definite possibility of finding a serious and treatable abnormality is, at least in the United States and in France, considered adequate motivation for using neuroimaging fairly routinely. How patients are evaluated, in turn, depends on the health care system and whether the necessary ancillary tests are routinely available. Our knowledge of seizures and epilepsy and treatment has advanced tremendously in the past several years. To take advantage of this new knowledge, it is necessary that a patient's epilepsy be adequately characterized. “Speciality care for patients with epilepsy must become standard of care”(31). For example, among 300 consecutive adults and children having presented with unexplained seizures, King et al. (32) diagnosed localization-related or generalized syndromes in only 47% with clinical information versus 77% with EEG findings and 81% with MRI. It would be almost impossible to classify specific syndromes without at least an EEG. In the CAROLE study, as in recent U.S. cohorts (19,23) and in the prospective Dutch Study of Epilepsy in Childhood (33), the very high rate of ancillary tests (EEG in >99% of cases, neuroimaging in 78.2% of the first-seizure group, and 68.3% of the newly-diagnosed-epilepsy group) permitted classification of seizures in 78.1% of the first-seizure group and 88.0% of the newly-diagnosed-epilepsy group; classification of syndromes in all the first seizures and 98.6% of those with newly diagnosed epilepsy; and classification of etiology in all the first seizures and 98.8% of those with newly diagnosed epilepsy, with a reasonably high degree of certainty at the time of initial diagnosis.

We suspect that there is reasonably good agreement across research groups in the criteria used to classify idiopathic syndromes (the 1.1 and 2.1 categories within the syndromic classification). These syndromes are very specific, for the most part easily recognizable, and well defined with respect to several clinical factors. These factors include age at onset, seizure type, and specific EEG abnormalities (e.g., 3-Hz generalized spike and wave). Perhaps the main limitation encountered in the classification of idiopathic syndromes is the sensitivity of the initial EEG in detecting a diagnostic abnormality.

Conversely, the criteria for classifying the symptomatic localization-related epilepsies, a large group of much more diverse syndromes, are somewhat more subject to interpretation. This is in part because many other factors must be taken into account. These factors are frequently not so well defined, and the ways in which they are used to classify an epilepsy syndrome may vary substantially from one investigator to another.

There are two main divisions within the symptomatic localization-related epilepsies. The first, “symptomatic by virtue of etiology,” refers to localization-related epilepsy occurring in the context of an antecedent cerebral insult (e.g., a stroke, severe head trauma), an abnormal neurologic examination indicating cerebral dysfunction or damage (e.g., mental retardation, cerebral palsy), or neuroimaging or other evidence of a brain abnormality or condition not otherwise appreciated through history and examination (e.g., a neuromigrational defect). The identification of a cause (etiology) depends on the types of evaluations performed and their sensitivity (2). The second category is “symptomatic by virtue of localization.” This determination may be largely or entirely determined by EEG evidence without any evidence of an identifiable underlying cause. There is a fundamental conflict between the use of the term symptomatic in the Classification of Epilepsies (1) as described earlier and the Epidemiologic Guidelines, which define symptomatic as “occur(ring) in relation to a well-demonstrated antecedent condition substantially increasing the risk for epileptic seizures”(2). Thus there are individuals whose epilepsy syndrome could be classified as symptomatic strictly by virtue of localization (1) but whose etiology would be classified as cryptogenic (2).

In applying the definitions from these two Commission statements, we decided to restrict the category of symptomatic localization-related to those who met both the syndromic (1) and epidemiologic (2) criteria. Individuals with cryptogenic etiology and partial seizures, even if well localized by EEG, were classified as cryptogenic localization-related. At the same time, patients with apparently generalized or undetermined-onset seizures that had a localizing lesion from a neuroimaging study were included in the symptomatic localization-related category. This is why there were no remote symptomatic cases in the 3.2 category.

Clearly, investigators who have addressed this issue previously recognized this dilemma and resolved it differently (16,19,23,33,34). Further discussion among investigators in this field is necessary to reach a consensus on this definitional issue for which the two Commission statements are at odds.

Beyond these remarks, 57.8% of first-seizure patients in the CAROLE study and 49.2% of those with newly diagnosed epilepsy had cryptogenic etiology. This last figure is similar to the 51.7% and the 48.9% found by Berg et al. (19) and by Shinnar et al. (23), respectively. Surprisingly, a similar proportion of cryptogenic cases had been reported in pre-MRI incidence surveys (15). The cryptogenic category corresponds to relatively nonspecific, poorly defined, or heterogeneous syndromes. In the CAROLE study, 17.0% of those with newly diagnosed epilepsy were cases of “undetermined epilepsy” (syndrome 3.2 in the International Classification) and 47.6% of first seizures were cases of “isolated seizure” (syndrome 4.2). The major difference between the 4.2 and 3.2 categories was whether a patient had one or more than one seizure. The minor difference was that a patient in the 3.2 category could not have partial seizures because recurrent cryptogenic partial seizures were classified in the 1.3 category. These categories could be considered as unclassified and provisional because they depend on the number and the sensitivity of the investigations and might be reclassified during the follow-up in light of new events and/or new investigations (23,35,36). Furthermore, more sophisticated MRI can now detect small epileptogenic lesions that passed unnoticed in 1995 (37). Similar considerations may be applied to the “cryptogenic localization-related syndrome” (syndrome 1.3).

Lennox–Gastaut syndrome (LGS) is a challenging area for nosology (38). As we were not sure which diagnostic criteria had been retained, all possible cases were categorized only as symptomatic generalized epilepsy. It seems that LGS is often not apparent at diagnosis but evolves over time (15,23). Only four (0.9%) of the 462 children in the Dutch study (33), four (0.7%) of the 613 children in the cohort of Connecticut (19), and four (3%) of the 150 children in a Finnish study were classified as LGS (34).

There were substantial differences in the distribution of syndromes in the first-seizure compared with the newly-diagnosed-epilepsy groups. The differences reflect the fact that certain seizure types are characteristic of specific syndromes and that some seizure types are less likely to present or virtually never present to medical attention at the time of an initial event. For example, absence and myoclonic seizures were absent from the first-seizure group, with the exception of three patients who presented in status epilepticus or had multiple seizures in a single day with GTCS. Consequently, there were almost no cases of childhood or juvenile absence or juvenile myoclonic epilepsy in the first-seizure group, as almost all of these patients had already experienced recurrent seizures by the time they were brought to medical attention. By the same token, simple and complex partial seizures often go unnoticed or unreported the first time they occur, whereas GTCS (if witnessed) are most likely to come to medical attention.

If the analysis is restricted to convulsive seizures, only four differences between the two groups were significant. The preponderance of male patients in the first-seizure group will be the matter of another article. Convulsive seizures and multiple seizures in a day or status epilepticus obviously lead to an earlier diagnosis. The issue remains of why a third of convulsive seizures do not come to attention the first time? Some occur in sleep or when the individual is alone. The event is later recalled and diagnosed as a probable seizure in retrospect. More cryptogenic localization-related syndromes occur in the newly-diagnosed-epilepsy group because cryptogenic syndromes are different to a certain extent between the first-seizure and newly-diagnosed-epilepsy groups because of the distribution of partial seizures.

Antiepileptic medication was prescribed to 89.6% of the newly-diagnosed-epilepsy group and to 86.2% of patients with convulsive seizures in this group. By contrast, 54.1% of the first-seizure group and 52.8% of patients with convulsive seizures in this group received medication. The decision to treat or not to treat varied depending on age, etiology, and syndrome and is the subject of a separate report (39). The decision to treat after only one seizure is in accordance with the literature (40). The presence of a lesion on neuroimaging is associated with a particularly high risk of recurrence (41). EEG abnormalities are also a valuable reason for treating children (42) as well as adults (43). Because specific abnormalities are associated with an increased risk of recurrence (44), they tend to help decide whether to treat. They also will help determine which drug to start if one is started. Specific abnormalities and hence some syndromes are associated. The risk of recurrence 5 years after a first unprovoked seizure has been estimated to be 42% when there was no risk factor for recurrence present versus 67% with at least one risk factor (45).

Whether a patient presents with a single seizure or newly diagnosed epilepsy does not appear to alter the initial diagnostic evaluation he or she receives. In fact, some of the assessment may be done before the full history of possible prior seizures has been obtained. Conversely, the decision to initiate treatment was significantly and substantially influenced by whether a patient had had prior seizures.

Follow-up of this cohort will provide valuable information about the prognostic significance of the syndromic classification as assessed at initial presentation, both in the first-seizure and newly-diagnosed-epilepsy patients, and about the stability of the syndromic classification over time.

Acknowledgment: We thank Anne Berg for offering us her helpful and detailed comments on the manuscript and Geneviève Chêne for her help in preparing the statistical analysis and the text. We also extend our thanks to all the members of the Groupe CAROLE (see appendix).


  1. Top of page
  2. Abstract
  5. First unprovoked seizures (n = 926)
  8. Appendix
  • 1
    Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989;30:38999.
  • 2
    Commission on Epidemiology and Prognosis of the International League Against Epilepsy. Guidelines for epidemiologic studies on epilepsy. Epilepsia 1993;34:5926.
  • 3
    Hauser WA. Incidence and prevalence. In: EngelJJr, PedleyTA, eds. Epilepsy: a comprehensive textbook. Philadelphia: Lippincott-Raven, 1997:4757.
  • 4
    Jallon P. Distribution géographique de l'épilepsie dans le monde. Presse Med 1996;25:187680.
  • 5
    Jallon P. Epidemiology. In: MeinardiH, ed. Handbook of clinical neurology. Amsterdam: Elsevier Science, 1999:292301.
  • 6
    Sander JWAS & Shorvon SD. Epidemiology of the epilepsies. J Neurol Neurosurg Psychiatry 1996;61:43343.
  • 7
    Shorvon SD, Hart YM, Sander JWAS, et al. The management of epilepsy in developing countries: an ICBERG manual. In: International Congress and Symposium Series number 175. London: Royal Society of Medicine Services Limited, 1991.
  • 8
    Commission on Classification and Terminology of the International League Against Epilepsy. Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia 1981;22:489501.
  • 9
    Picot M-C, Crespel A, Tricot M, et al. Validity of diagnosis using the French translation of the semi-structured interview for seizure classification. Epilepsia 1999;40:164956.
  • 10
    Sander JWAS, Hart YM, Johnson AL, et al. National General Practice Study of Epilepsy: newly diagnosed epileptic seizures in a general population. Lancet 1990;336:126771.
  • 11
    Forsgren L, Bucht G, Eriksson S, et al. Incidence and clinical characterization of unprovoked seizures in adults: a prospective population-based study. Epilepsia 1996;37:2249.
  • 12
    Hauser WA & Kurland LT. The epidemiology of epilepsy in Rochester, Minnesota, 1935 through 1967. Epilepsia 1975;16:166.
  • 13
    Forsgren L. Prospective incidence study and clinical characterization of seizures in newly referred adults. Epilepsia 1990;31:292301.
  • 14
    Groupe CAROLE. Délais évolutifs des syndromes épileptiques avant leur diagnostic : résultats descriptifs de l'enquête CAROLE. Rev Neurol (Paris) 2000;156:48190.
  • 15
    Loiseau J, Loiseau P, Guyot M, et al. Survey of seizure disorders in the French Southwest. I. Incidence of epileptic syndromes. Epilepsia 1990;31:3916.
  • 16
    Zarrelli M, Beghi E, Rocca WA, et al. Incidence of epileptic syndromes in Rochester, Minnesota: 1980-1984. Epilepsia 1999;40:170814.
  • 17
    Osservatorio Regionale per l'Epilessia (OREp) Lombardy. The contribution of tertiary centers to the quality of the diagnosis and treatment of epilepsy. Epilepsia 1997;38:133843.
  • 18
    Cockerell OC, Johnson AL, Sander JWAS, et al. Prognosis of epilepsy: a review and further analysis of the first nine years of the British National General Practice Study of Epilepsy, a prospective population-based study. Epilepsia 1997;38:3146.
  • 19
    Berg AT, Shinnar S, Levy SR, et al. Newly diagnosed epilepsy in children: presentation at diagnosis. Epilepsia 1999;40:44552.
  • 20
    Loiseau P, Duché B, Loiseau J. Classification of epilepsies and epileptic syndromes in two different samples of patients. Epilepsia 1991;32:3039.
  • 21
    Berg AT, Testa FM, Levy SR, et al. The epidemiology of epilepsy. past, present and future. Neurol Clin 1996;14:38398.
  • 22
    Loiseau J, Loiseau P, Duché B, et al. A survey of epileptic disorders in Southwest France: seizures in elderly patients. Ann Neurol 1990;27:2327.
  • 23
    Shinnar S, O'Dell C, Berg AT. Distribution of epilepsy syndromes in a cohort of children prospectively monitored from the time of their first unprovoked seizure. Epilepsia 1999;40:137883.
  • 24
    Olafsson E, Hauser WA, Ludvigsson P, et al. Incidence of epilepsy in rural Iceland: a population-based study. Epilepsia 1996;37:9515.
  • 25
    Hauser WA, Annegers JF, Kurland LT. Incidence of epilepsy and unprovoked seizures in Rochester, Minnesota: 1935-1984. Epilepsia 1993;34:45368.
  • 26
    Jallon P, Goumaz M, Haengelli C, et al. Incidence of first epileptic seizures in the canton of Geneva, Switzerland. Epilepsia 1997;38:54752.
  • 27
    Keränen T, Riekkinen PJ, Sillanpää M. Incidence and prevalence of epilepsy in adults in Eastern Finland. Epilepsia 1989;30:41321.
  • 28
    Perucca E, Gram L, Avanzini G, et al. Antiepileptic drugs as cause of worsening of seizures. Epilepsia 1998;39:517.
  • 29
    Benbadis SR & Lüders HO. Epileptic syndromes: an underutilized concept. Epilepsia 1996;37:102934.
  • 30
    Loiseau P & Loiseau J. Épilepsies: vous avez dit épilepsie? Rev Neurol (Paris) 1999;155:1938.
  • 31
    Dasheiff RM and the Promotion of Speciality Care for Epilepsy Group. Speciality care for patients with epilepsy must become standard of care. Seizure 1999;8:43940.DOI: 10.1053/seiz.1999.0350
  • 32
    King MA, Newton MR, Jackson GD, et al. Epileptology of the first-seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients. Lancet 1998;352:100711.DOI: 10.1016/s0140-6736(98)03543-0
  • 33
    Callenbach PMC, Geerts AT, Arts WFM, et al. Familial occurrence of epilepsy in children with newly diagnosed multiple seizures: Dutch study of epilepsy in childhood. Epilepsia 1998;39:3316.
  • 34
    Sillanpää M, Jalava M, Shinnar S. Epilepsy syndromes in patients with childhood-onset seizures in Finland. Pediatr Neurol 1999;21:5337.DOI: 10.1016/s0887-8994(99)00031-4
  • 35
    Berg AT, Shinnar S, Levy SR, et al. How well can epilepsy syndromes be identified at diagnosis? A reassessment after two years of follow-up. Epilepsia 2000;41:126975.
  • 36
    Loiseau J, Picot M-C, Chêne G, et al. Classification and incidence of epileptic syndromes in a prospective study: reliability and causes of modification. Epilepsia 2000;39(suppl 6):226.
  • 37
    Stevens JM. New and developing MRI techniques in epilepsy. Epilepsia 1997;38(suppl 10):2832.
  • 38
    Dulac O & N'Guyen T. The Lennox-Gastaut syndrome. Epilepsia 1993;34(suppl 7):S717.
  • 39
    Groupe CAROLE. Traitement des crises épileptiques nouvellement diagnostiquées: une expérience française. Rev Neurol (Paris) 2001 (in press).
  • 40
    Beghi E & Ciccone A, and the First Seizure Trial Group (FIRST). Recurrence after a first unprovoked seizure: is it still a controversial issue? Seizure 1993;2:510.
  • 41
    Semah F, Picot M-C, Adam C, et al. Is the underlying cause of epilepsy a major prognostic factor for recurrence? Neurology 1998;51:125662.
  • 42
    Shinnar S, Kang H, Berg AT, et al. Abnormalities in children with a first unprovoked afebrile seizure. Epilepsia 1994;35:4716.
  • 43
    Van Donselaar CA, Schimsheimer R-J, Geerts AT, et al. Value of the electroencephalogram in adult patients with untreated idiopathic first seizures. Arch Neurol 1992;49:2317.
  • 44
    Berg AT & Shinnar S. The risk of seizure recurrence following a first unprovoked seizure: a quantitative review. Neurology 1991;41:96572.
  • 45
    Loiseau J, Picot M-C, Jallon P, et al. Single unprovoked epileptic seizures, accompanied or unaccompanied: risk of recurrence. Epilepsia 1997;38(suppl 8):249.


  1. Top of page
  2. Abstract
  5. First unprovoked seizures (n = 926)
  8. Appendix

Observatoire de l'Epilepsie: C. Cachera, F. Chain, P. Jallon.

Unité de Coordination Nationale: M. Baulac, F. Fagnani, P. Jallon, J. Leveau, P. Loiseau, J. Motte, P. Thomas, L. Vallée.

Coordination Régionale: C. Allaire, A. Autret, M. Baldy-Moulinier, M. Baulac, M. Clanet, G. Dordain, J-L. Gastaut, M. Giroud, P. Jallon, E. Josien, P. Loiseau, C. Marescaux, P. Masnou, F. Mauguière, J. Motte, D. Parain, J. Perret, M. Revol, L. Rumbach, P. Tapie, P. Thomas, L. Vallée, M. Weber.

Investigators : C. Adam, C. Allaire, N. Attal, F. Attané, P. Aubrun, N. Ayrivié, N. Badinand-Hubert, M. Baldy-Moulinier, J. Bapst-Reiter, M-A. Barthez-Carpentier, F. Bartolomei, M. Bataillard, M. Baulac, N. Bednarek, C. Belair, M. Bénazet, F-X. Bergouignan, C. Bernard, L. Bernard-Bourzeix, F. Bertran, P. Bertrand, P. Beuriat, C. Billard, F. Billé-Turc, C. Billy, A. Biraben, A. Blanc, F. Boidein, N. Bouayed, S. Boudon, J. Bouillat, J. Boulloche, A. Bredin, O. Brocard, P. Brosset, F. Brunet-Bourgin, B. Cenraud, D. Chaigne, Y. Chaix, M-P. Chaunu, D. Chavot, P. Clavelou, S. Cohadon, N. Collombier, P. Contis, P. Convers, J. Couchot, M-A. Cournelle, S. Courtois, J-M. Croguennec, J-M. Cuisset, J-C. Cuvellier, J. d'Anglejan, G. Damon, A. Danielli, P. Daubney, J. de Bellescize, L. de Lumley, A. de Recondo, M. de Swarte, D. Deffond, T. Delangre, B. Delisse, P. Derambure, S. Derambure, P. Desbordes, F. Desfrançois, M. Destée-Warot, J. Dien, B. Doremus, M-C. Dourneau-Lethiecq, F. Dubois, B. Duché, X. Ducrocq, J. Duhurt, J. Duprey, G. Durand, A. Dusser, J-P. Escaillas, G. Fanjaud, D. Felten, C. Fischer, D. Fontan, F. Formosa, E. Foulon, A. Furby, S. Gallet, J. Galmiche, P. Garde-Arthaud, S. Garrel, J-L. Gastaut, C. Gaultier, C. Gauthier, D. Gauthier-Morel, P. Genton, G. Géraud, J-P. Girard, M. Girard-Madoux, M. Giroud, P. Gonnaud, C. Goulon-Goëau, S. Gros, M. Grosclaude, M. Gross, B. Gueguen, H. Guinot, C-A. Haenggeli, J-L. Henlin, B. Hevin, P. Hinault, P. Homeyer, C. Hommet, E. Huart, P. Huc, B. Huttin, L. Inglésiakis, J. Isnard, H. Isnard, P. Jallon, M. Jogeix, E. Josien, C. Juhel, P. Kahane, M. Kalafat, P. Keo-Kosal, A-M. Kreib, C. Kubler, E. Larrieu, J-L. Larrieu, D. Latinville, A. Le Gallou-Wittenberg, F. Lebas, P. Lebrun-Grandié, J. Leche, A. Legout, S. Legrand, M. Legroux, J-F. Lemaitre, P. Lestavel, M. Levasseur, C. Lienhard, M-O. Livet, P. Louiset, M. Lubeau, J. Lucas-Daviaud, S. Maillard, M. Maillet-Vioud, J. Mancini, M. Mann, C. Marchal, C. Marescaux, L. Martini, P. Masnou, F. Mauguière, J. Maupetit, R. Maynard, P. Menage, D. Ménard, R. Métreau, M. Milor, M-C. Minot-Myhie, Y. Moene, B. Montagne, M-E. Montelescaut, J. Moreau, O. Moreaud, J. Motte, B. Noelle, X. Olmi, J. Orbegozo, A-M. Ouvrard-Hernandez, D. Parain, A. Parsa, B. Pautrizel, J-M. Pedespan, P. Pernes, B. Perrouty, J. Petit, S. Peudenier, A-M. Picard, C. Pierrot-Deseilligny, E. Planque, C. Portha, P-M. Preux, M. Prud'homme, D. Raybaut-Guilhem, P. Rebaud, A. Regi, J-L. Regi, J. Reis, F. Rejou, C. Rémy, J-F. Renard, M. Revenu, M. Rey, C. Richelme, P. Ricou, J-P. Rigal, R. Rogez, C. Rousselle, L. Rumbach, C. Rummens, P. Ryvlin, P. Sabouraud, I. Saïkali, D. Saudeau, J-F. Savet, J-L. Schaeffer, J-L. Schaff, F. Schoenfelder, P. Schuermans, J. Senant, A. Setiey, C. Sevrin, G. Sivelle, T. Soisson, P. Soubielle, X. Soulages, S. Soulayrol, F. Tabaraud, P. Taillandier, C. Tannier, P. Tapie, V. Tarel, D. Taussig, F. Thedrez, P. Thomas, C-L. Tournier, J-D. Turc, L. Vallée, C. Vanhulle, J. Vaunaize, A. Verier, D. Vernay, J-M. Visy, C. Vongsouthi, J. Vrigneaud, E. Waubant, A. Weichlein, O. Weill, L. Zai, F. Ziegler, C. Zix.

Monitoring: CEMKA (F. Fagnani, M. de Zélicourt).

Financial support: Unité Thérapeutique SNC, Parke-Davis-France.