The international classification of the epilepsies and epileptic syndromes

Epilepsy Research 41 (2000) 223 – 234 www.elsevier.com/locate/epilepsyres

The international classification of the epilepsies and epileptic syndromes An algorithm for its use in clinical practice Giuseppe Rinaldi a, Michele M. Zarrelli a,*, Ettore Beghi a,b, Francesco Apollo a, Michele Germano c, Piero Di Viesti a, Pasqualino Simone a a Di6isione Neurologica, Ospedale ‘Casa Sofferenca’ IRCSS, 71013 San Gio6anni Rotondo, Italy Centro Regionale per l’Epilessia, Ospedale ‘San Gerardo’, Monza and Istituto ‘Mario Negri’, Milan, Italy c Ser6izio di Neuropsichiatria Infantile, Ospedale ‘Casa Sollie6o’ IRCSS, 71013 San Gio6anni Rotondo, Italy b

Received 10 February 2000; received in revised form 20 May 2000; accepted 27 May 2000

Abstract An algorithm has been structured as a guided reading of the international league against epilepsy (ILAE) syndromic classification to be used in clinical practice by less experienced physicians in newly diagnosed patients. The algorithm followed the original structure of the classification, which identifies major syndromic groups, subgroups, and specific syndromes. Validation required two raters, a resident and a board-certified neurologist, to apply the algorithm with different techniques (direct or recorded interview, medical record consultation) to 19 children and 18 adults with epilepsy with information available at the time of diagnosis. The two raters’ diagnoses were compared with those of the caring physicians, and cases where disagreement arose were discussed in conference to achieve consensus. The k statistic was used as a measure of inter-rater agreement. Caring physicians and both raters agreed in 51% of cases. Substantial agreement (k=0.75) was obtained between the resident and the neurologist on major diagnostic groups and subgroups, mostly in adults. Agreement with the caring physician was slightly more satisfactory for the resident (k =0.67) than for the neurologist (k= 0.60). Agreement was better with direct or indirect interview than with record consultation, and improved further after discussion. Agreement was obtained after discussion in 32% of cases, in some of which the caring physician agreed on the resident’s diagnosis. Agreement was less satisfactory for specific syndromes. On this basis, an algorithm of the ILAE classification is a fairly reliable instrument only for making a broad syndromic classification of epilepsy at the time of diagnosis. The limits of the algorithm tend mostly to reflect the intrinsic limitations of the classification itself. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Epilepsy; Classification; Epileptic syndromes; Diagnosis

* Corresponding author. Tel.: + 39-882-4106534; fax: + 39-882-453861. 0920-1211/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 0 - 1 2 1 1 ( 0 0 ) 0 0 1 4 7 - 9

224

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

1. Introduction The Classification of the epilepsies and epileptic syndromes of the international league against epilepsy (ILAE) (Commission, 1989) is the most widely recognized instrument for the diagnostic definition of epilepsies in clinical practice. This classification stratifies epilepsies in four major classes on the basis of the presumed site of origin of the seizures (localization-related or generalized) and the etiology of the seizures (idiopathic or cryptogenic or symptomatic). Within each class, further attempts are made to classify patients in more detailed subgroups with reference to selected variables such as age, family history, brain development, intellectual function, specific etiology, EEG findings, and outcome. The published classification has been repeatedly assessed in series of pediatric and adult patients, with variable results. The proportion of properly classified cases range from 34 to 97% (Loiseau et al., 1991; Manford et al., 1992; Shah et al., 1992; Eadie, 1996; OREp, 1996). The lowest classification rates may reflect less experience with the current classification of the epilepsies, particularly in unspecialized centers. An alternative interpretation is that in its present form, this classification is poorly applicable. To verify whether and to what extent the ILAE classification of the epilepsies could be used by less experienced or inexperienced investigators, an algorithm was devised on the basis of published material (Commission, 1989). The aim of this study was to illustrate this algorithm and test its validity and reliability.

2. Material and methods The algorithm (see Appendix A) was constructed starting from the major headings of the classification with reference to the presumed site of origin of the seizures (partial, generalized, partial and generalized, uncertain whether partial or generalized, situation-related). A case was then classified under one of the subheadings on the basis of the presence or absence of specific clinical, electrophysiological and neuroimaging fea-

tures. For example, to be classified as idiopathic partial epilepsy, a case had to present all the following features: age less than 16 years, no history of antecedent illnesses, absence of intellectual deficit, absence of detectable causative lesions, EEG with normal background activity and focal high-voltage epileptiform abnormalities, and (in the presence of follow-up data) remission of seizures. When one or more of the above features were absent, the case was classified among the cryptogenic partial epilepsies or, in the presence of causative lesions, among the symptomatic partial epilepsies. To be included in the latter group, a patient should have at least one among causative lesion, neurological and/or intellectual deficit, and history of antecedent illness. To facilitate the use of the algorithm, a practical guide (available upon request) was devised which included a schematic representation of the classification, with reference to the major headings, subheadings, and specific syndromes. Using the algorithm, a case was classified in steps with as much precision as possible using the information available at the time of assessment. In the first step, a case was classified according to a major syndromic group (e.g. localization-related epilepsy). The second step was classification in a syndromic subgroup (e.g. idiopathic epilepsy with age-related onset). During the third step, the individual syndrome was identified (e.g. benign childhood epilepsy with centro-temporal spikes). If at any further step, the classification was impossible, the case was classified as ‘unknown’. For each patient included in the validation, the caring physician was asked to provide his best syndromic classification, using this three-step procedure of the published form of the classification (Commission, 1989). Validation involved applying the algorithm to a number of children and adults with newly diagnosed epilepsy seen as in- or out-patients, with information available at the time of the diagnosis. The patients were recruited by the neurologist or child neurologist responsible for the care of epilepsy (FA, MG, PDV). These patients’ diagnostic work-up was done as usual in clinical practice and carefully reported in a medical

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

record. Thus, no attempt was made to modify current practice in the management of epilepsy. This choice reflected the need to test the algorithm in the earliest and most unfavorable situation (i.e. when less data were available), but at the stage when a first classification of the patient was needed. To verify the validity of the instrument in the hands of physicians from different scientific backgrounds, the patients were assessed separately by two raters, a resident with a 4-year neurological training and an unrelated board-certified neurologist also expert in epilepsy. Interrater agreement was tested in the whole study population and in different settings reflecting variable modes of application of the classification (direct interview, sound-recorded interview, and consultation of the medical record). The cases used for the validation process were assessed with one of the three techniques. Each case was examined by the resident and the neurologist using the same technique, except for the patients undergoing a recorded interview who were interviewed by the resident and assessed by the neurologist. k-Statistics for two raters (Fleiss, 1981) were used as a measure of inter-rater agreement. As suggested by Landis and Koch (1977), agreement was considered ‘almost perfect’ with k \ 0.80, ‘substantial’ with k being 0.80 – 0.61, ‘moderate’ with k being 0.60–0.41, ‘fair’ with k having the value of 0.40–0.21, ‘slight’ with k ranging 0.20 – 0.00, and ‘poor’ with k B 0. The k-statistic was calculated for the first and second step combined, as a measure of agreement on major syndromic groups and subgroups, and on the first, second and third step combined, as a measure of agreement on the specific syndromic categories. All cases on which there was no agreement at any level were discussed by the two raters with the caring physician. For cases in which disagreement was caused by an erroneous application of the classification by the caring physician, the kstatistic was recalculated on the basis of the consensus obtained between the physician and the resident (less favorable application of the algorithm).

225

3. Results The sample population included 37 children and adults (13 males and 24 females aged from 9 months to 66 years), chosen among patients attending the Neurology Department and the Child Neurology Unit at San Giovanni Rotondo hospital. The main demographics of the sample are set out in Table 1, with the syndromic classification made by the caring physician at the time of diagnosis. Cryptogenic partial epilepsies were the commonest syndromic pattern and temporal lobe the most frequent site of the seizures. Symptomatic partial epilepsies and generalized idiopathic epilepsies followed. Complete agreement was obtained (i.e. on all the categories of the classification) between the caring physicians and both raters for 19 cases (51%), nine of them adults and ten children. There was agreement between the resident and the caring physician for 21 cases (57%) and between the neurologist and the caring physician for 20 (54%). The neurologist and the resident agreed on 30 cases (81%). After discussion of the 18 cases on which agreement was not obtained, all the investigators (caring physicians and raters) achieved a consensus on 12 (nine children and three adults) (Table 2). In five of these, the caring physicians agreed to change the diagnosis to agree with the resident’s assessment. In three other cases, all the investigators modified their original diagnosis. There were six cases (all adults) for which agreement was not obtained even after discussion, although the caring physician revised the diagnosis in one patient and the neurologist in another (Table 3). With reference to the major syndromic groups and subgroups, agreement between the resident and the caring physician was substantial (k= 0.80) for the direct interview and fair (k = 0.35) for the record consultation (Table 4). Agreement between the caring physician and the neurologist was moderate (k= 0.58) on the direct interview and substantial (k = 0.69) on the recorded one. Substantial agreement was found between the resident and the caring physicians for all the patients. The neurologist agreed substantially with the caring physician for the adults and moderately

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

226

for the child neurology patients. Slight-to-moderate improvement was obtained after consensus in all the settings and for the children and the adults separately. Compared with the concordance between the raters and the caring physicians, agreement between the two raters before reaching a

consensus was better in any setting. The two raters had almost full agreement on the adults and substantial agreement on the children. The result was less satisfactory when the steps of the classification were considered altogether (Table 4). Agreement between the resident or the

Table 1 General characteristics of the sample Case

Agea

Patients from the 1 8y 2 4y 3 11 y 4 15 y 5 14 y 6 12 y 7 16 y 8 13 y 9 5y 10 5y 11 6y 12 4y 13 9m 14 6y 15 5y 16 10 y 17 5y 18 18 m 19 3.5 y

Sex

Syndromic classificationb

Child Neurology Department F Localization-related, cryptogenic, occipital F Localization-related, cryptogenic, temporal M Childhood absence epilepsy M Generalized, symptomatic, nonspecific etiology (other) F Localization-related, cryptogenic, frontal M Benign childhood epilepsy with centrotemporal spikes F Juvenile absence epilepsy F Localization-related, cryptogenic, occipital M Localization-related, symptomatic (mental retardation), amygdalo-hippocampal F Childhood absence epilepsy F Benign childhood epilepsy with centrotemporal spikes F Localization-related, cryptogenic, temporal F Generalized, symptomatic, nonspecific etiology (other) M Childhood absence epilepsy F Benign childhood epilepsy with centrotemporal spikes F Epilepsy without unequivocal generalized or focal features F Benign childhood epilepsy with centrotemporal spikes M Localization-related, symptomatic (leucoencephalopathy) M Epilepsy with myoclonic-astatic seizures

Patients from the Neurology Department 20 39 M Localization-related, cryptogenic, temporal 21 60 F Localization-related, cryptogenic, frontal (motor cortex) 22 29 F Localization-related, cryptogenic, temporal 23 29 M Localization-related, cryptogenic, latero-temporal 24 60 F Localization-related, symptomatic (metastatic breast cancer), frontal (motor cortex) 25 14 F Juvenile absence epilepsy 26 43 M Localization-related, symptomatic (mesial sclerosis), temporal 27 29 F Localization-related, cryptogenic, temporal 28 18 M Generalized, idiopathic (other not defined) 29 15 F Juvenile absence epilepsy 30 21 F Generalized, idiopathic (other not defined) 31 40 M Localization-related, cryptogenic, frontal (motor cortex) 32 21 M Generalized, symptomatic, nonspecific etiology (other) 33 17 F Generalized, symptomatic, nonspecific etiology (other) 34 54 F Localization-related, cryptogenic, temporal 35 51 F Localization-related, symptomatic (multiple undefined hyperintense lesions), frontal 36 44 F Localization-related, symptomatic (single undefined hyperintense lesion), frontal 37 66 F Localization-related, symptomatic (lacunar stroke), temporal a

y, Years; m, months. Syndromic classification made by the caring physician at the time of the diagnosis. c Code number of the algorithm. b

Codec

1.3.6 1.3.3 2.1.4 2.3.1.3 1.3.4 1.1.1 2.1.5 1.3.6 1.2.3.1 2.1.4 1.1.1 1.3.3 2.3.1.3 2.1.4 1.1.1 3.2 1.1.1 1.2 2.2.3 1.3.3 1.3.4.7 1.3.3 1.3.3.2 1.2.4.7 2.1.5 1.2.3 1.3.3 2.1.8 2.1.5 2.1.8 1.3.4.7 2.3.1.3 2.3.1.3 1.3.3 1.2.4 1.2.4 1.2.3

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

227

Table 2 Syndromic classification (code diagnoses) in patients in whom a previous disagreement among raters was followed by a consensus after discussion Casea

Caring physician

Resident

Neurologist

Consensus

2 4 8 10 11 12 14 15 16 26 33 36

1.3.3 2.3.1.3 1.3.6 2.1.4 1.1.1 1.3.3 2.1.4 1.1.1b 3.2 1.2.3 2.3.1.3 1.2.4

2.1.8 1.3.3 1.3.3.1 2.1.4 1.3 1.3 1.3.3 1.3.3 1.3.3 1.2 2.1.8 1.2.4

3.2.2 1.3 1.3.3 3.2.2 3.2.2 1.3 1.3.3 1.3.3 3.2.2 1.2 2.1.8 1.3.3

1.3 1.3.3 1.3 2.1.4 1.3 1.3 1.3.4 1.3.4 3.2.2 1.2 2.1.8 1.2.4

a b

Refers to the code reported in Table 1. Diagnosis made by the caring physician on the basis of his own experience, although no follow-up data were available.

Table 3 Syndromic classification (coded diagnoses) in patients for whom disagreement among raters persisted after discussion Casea

23 25 28 29 32 37 a

Caring physician

Resident

Neurologist

Before discussion

After discussion

Before discussion

After discussion

Before discussion

After discussion

1.3.3.2 2.1.5 2.1.8 2.1.5 2.3.1.3 1.2.3

1.3.3.2 2.1.5 2.1.8 2.1.5 3.2.2 1.2.3

1.3.6 2.1.6 3.2.2 2.1.8 1.3.5 1.3.3.2

1.3.6 2.1.6 3.2.2 2.1.8 1.3.5 1.3.3.2

1.3.6 2.1.6 3.2.2 2.1.8 3.2.2 1.3.3.2

1.3.6 2.1.5 3.2.2 2.1.8 3.2.2 1.3.3.2

Numbers as in Table 1.

neurologist and the caring physician was moderate for the direct interview and fair for the other two settings. The caring physicians and both raters had fair agreement for adults and moderate agreement for children. The change in diagnosis by the caring physician after consensus significantly improved the indices of agreement for all the settings, in children and adults. Better agreement was found between the two raters at any setting, for adults (almost perfect) and for children (substantial). Concordance was similar between the two raters when the classification based on major syndromic groups and subgroups (Table 4) was compared with that based on the specific syndromes.

4. Discussion An algorithm has been devised to permit practical application of the international classification of the epilepsies at the time of diagnosis in different clinical contexts. Using this instrument, physicians with different levels of medical training substantially agreed on a series of children and adults interviewed either directly or indirectly from a review of the clinical records. Agreement was better for children than adults and was unchanged when the main syndromic groups and subgroups were assessed separately. Agreement tended to be less satisfactory when the diagnosis was based on review of the clinical

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

228

cian. This was the case of five patients for whom there was subsequent agreement on the resident’s diagnosis. Second, an incomplete diagnostic assessment is another potential source of disagreement. However, in this study, the diagnostic work-up was considered with reference to current clinical practice and there are no published reports on the validity of an intensive diagnostic investigation nor any practical guidelines for the diagnosis of epilepsy. Third, the limits of the algorithm may reflect the intrinsic limitations of the classification of the epilepsies. Evidence comes from the relatively high proportion of patients (16%) for whom different diagnostic interpretations persisted even after discussion of the cases. Even though the limited usefulness of the classification may reflect here the restricted amount of data available at the time of the first diagnosis, the different distribution of the epileptic syndromes and the high proportion of unclassified cases in different populations (Loiseau et al., 1991; Manford et al., 1992; Oka et al., 1995) may

records. This reflects the concept that a diagnosis based on a decision tree, which depends on the fulfillment of precise inclusion/exclusion criteria, relies heavily on the completeness of the record. Although we have no comparative findings from a simple reading of the published text of the classification (Commission, 1989), we expect better results from the use of the algorithm on this basis, as previous attempts to test inter-rater agreement on the classification of epileptic seizures (Commission, 1981) provided contrasting and often unsatisfactory findings (Bodensteiner et al., 1988; Nordli et al., 1997). In this study, agreement was moderate on the same cases between physicians using the algorithm and the caring physicians (k = 0.49). This may depend on the limits of this instrument, which may be greater when a full syndromic classification (all steps included) is attempted in adults (k=0.37). However, other interpretations are possible on this point. First, a patient could be erroneously classified even by the caring physi-

Table 4 Measures of inter-rater agreement (k-statistic) before and after discussion After discussiona

Number of cases Before discussion Caring physician vs. resident Major diagnostic groups and subgroups Direct interview 0.80 Recorded interview Medical record 12 0.35 13 0.66 All Nb patients 19 0.68 All CNcpatients All cases 37 0.67 Syndromes Direct interview Recorded interview Medical record All Nb patients All CNc patients All cases

Caring physicians vs. neurologist

Resident vs. neurologist

0.58 0.69

0.81 0.81

0.86

0.38 0.55 0.63 0.60

0.55 0.82 0.67 0.75

0.62 0.74 0.82 0.79

25 15d

0.53

0.67 0.27

0.78 0.83

0.71

12 18 19 37

0.35 0.37 0.57 0.49

0.37 0.37 0.53 0.44

0.63 0.86 0.70 0.78

0.58 0.53 0.70 0.67

a Diagnostic changes included only cases in which the caring physician accepted the diagnosis made by the resident before discussion (see text for further explanation). b N, neurology. c CN, child neurology. d Direct interview was also recorded in 15 cases.

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

also reflect different interpretations of the classification itself. Substantial concordance between different raters (resident, experienced neurologist) and between these and the caring physicians was achieved mainly for direct or recorded interviews. This suggests that interviewing patients, as is currently done in clinical practice, is a satisfactory source of findings to be used in a decision tree. As expected, the inter-rater agreement tended to decrease when comparing major syndromic groups and subgroups with individual syndromes. This again may be a limitation of the classification itself. However, these findings suggest that a more rational use of the algorithm may have to be limited to broad classification of the patient. The concordance was better for children than adults, as shown by the higher statistical indices and by the complete agreement reached after discussion of the cases. The greater disagreement rate for adults may partly depend on the characteristics of the patients examined and the level of medical training of the caring physicians. However, intrinsic difficulties in the classification of a case must also be considered, which include the precise localization of the epileptogenic site, the presence (and relative weight) of factors suggesting partial or generalized epilepsy, and the contributions of specific etiologic factors. The higher concordance between the caring physician and the resident, compared with the board certified neurologist, is worth mentioning. This is not unexpected (Bodensteiner et al., 1988) and suggests that more experienced observers are more likely to rely on their own judgement and tend less to follow classification criteria closely. In conclusion, the algorithm for the classification of the epilepsies presented here proved a fairly valid and reliable instrument only for broad syndromic classification of epilepsy at the time of diagnosis. The small number of patients and assessors prevents us from drawing general conclusions from our data. However, the limits of the algorithm partly reflect the intrinsic limitations of the classification itself. Improvement of the classification, which is now being considered (Blume et al., 1997), may lead to the design of an even

229

better decision tree for use by physicians with different levels of medical training.

Acknowledgements The authors are indebted to Dr G. Avanzini, Dr J. Engel, Dr G. Gobbi, Dr W.A. Hauser, Dr L.M. Specchio, and Dr C.A. Tassinari for their helpful criticism, to J. Baggott for editorial assistance and to S. Franceschi for secretarial help.

Appendix A. Algorithm for the classification of the epilepsies and epileptic syndromes Note: epileptic seizures must be classified using the criteria of the international classification of the ILAE. From an examination of the medical records, the patient has: 1. partial seizures (go to Section A.1); 2. generalized seizures (see Section A.2); 3. partial and generalized seizures (see Section A.3); 4. uncertain seizures (partial or generalized at origin) (see Section A.3.1); and 5. situation-related seizures (see Section A.4). A.1. Patient has partial seizures. Patient may ha6e localization-related (LR) epilepsy Patient has: 1a 1b 1c 1d

age at onset in childhood no antecedent illnesses absence of intellectual deficit EEG with normal background and localized HV repetitive spikes increased by sleep 1e absence of demonstrable lesions 1f spontaneous remission

“

“

yes/no yes/no yes/no yes/no

yes/no yes/no

If the criteria 1a through 1f are satisfied, patient may have idiopathic LR epilepsy (see Section A.1.1); if the criteria 1b and/or 1c and/or 1e are not satisfied, patient may have symptomatic LR epilepsy (see Section A.1.2); and

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

230 “

if the criteria 1b, 1c or 1e are satisfied and at least one of the remaining criteria (1a, 1d, 1f) is not satisfied, patient may have cryptogenic LR epilepsy (Section A.1.3).

A.1.1. Idiopathic LR epilepsy Patient may have 1.1.1

1.1.2.

1.1.3 1.1.4

Benign childhood epilepsy with centro-temporal spikes (Annex 1, Section 1.1.A) childhood epilepsy with occipital paroxysms (Annex 1, Section 1.1.B) primary reading epilepsy (Annex 1, Section 1.1.C another epilepsy with seizures precipitated by specific modes of activation (Annex 1, Section 1.1.D)

yes/no

yes/no

yes/no yes/no

A.1.2. Symptomatic LR epilepsy To be included in this category, patient must have at least one of the following: Demonstrable causative lesion Neuro/intellectual deficit Antecedent illness

yes/no yes/no yes/no

1.2.1

yes/no

1.2.1.1

1.2.1.2

1.2.2

chronic progressive epilepsia partialis continua of childhood (Kojewnikow’s syndrome (Annex 1, Section 1.2 B8) Rasmussen’s syndrome (Annex 1, Section 1.2.B81) a particular form of Rolandic partial epilepsy (Annex 1, Section 1.2.B82) another epilepsy with seizures precipitated by specific modes of activation (Annex 1, Section 1.1.D)

yes/no

yes/no

yes/no

1.2.3

temporal lobe (Annex 1, yes/no Section 1.2.A) To be included in this section, patient must have simple partial seizures (autonomic, psychic, sensory, epigastric sensation) or complex partial seizures (motor arrest with oro-alimentary automatisms) with gradual recovery and post-ictal amnesia 1.2.3.1 amygialo-hippocampal yes/no (Annex 1, Section 1.2.A1) 1.2.3.2 lateral temporal (Annex 1, yes/no Section 1.2.A2) 1.2.4 frontal lobe (Annex 1, yes/no Section 1.2.B) To be included in this section, patient must have seizures as shown in one of the categories indicated below 1.2.4.1 supplementary motor (An- yes/no nex 1, Section 1.2.B1) 1.2.4.2 cingulate (Annex 1, Secyes/no tion 1.2.B2) 1.2.4.3 anterior frontopolar (An- yes/no nex 1, Section 1.2.B3) 1.2.4.4 orbitofrontal (Annex 1, yes/no Section 1.2.B4) 1.2.4.5 dorsolateral (Annex 1, yes/no Section 1.2.B5) 1.2.4.6 opercular (Annex 1, Sec- yes/no tion 1.2.B6) 1.2.4.7 motor (Annex 1, Section yes/no 1.2.B7) 1.2.4.8 Kojewnikow’s syndrome 1.2.4.8.1 Rasmussen’s syndrome yes/no (Annex 1, Section 1.2.B81) 1.2.4.8.2 Rolandic partial epilepsy yes/no (Annex 1, Section 1.2.B82) 1.2.5 Parietal lobe (Annex 1, yes/no Section 1.2.C) To be included in this section, patient must have simple partial seizures, predominantly sensory with visual phenomena and/or asomatognosia and/or vertigo or disorientation and/or receptive or conductive aphasia and/or lateralized genial sensations 1.2.6 occipital lobe (Annex 1, yes/no Section 1.2.D)

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

To be included in this section, patient must have simple partial (mostly visual) or secondarily generalized seizures. Complex partial seizures reflect spread beyond the occipital lobe A.1.3. Cryptogenic LR epilepsy A patient is included in this Section if the criteria 1b, 1c and 1e are satisfied and at least one of the remaining criteria 1a, 1d, 1f is not satisfied. Patient with cryptogenic LR epilepsy has presumably symptomatic partial seizures with lack of etiologic evidence. 1.3.1

Kojewnikow’s syndrome yes/no (Annex 1, Section 1.2.B8) 1.3.1.2 a particular form of yes/no Rolandic partial epilepsy (Annex 1, Section 1.2.B82) 1.3.2 another epilepsy with yes/no seizures precipitated by specific modes of activation (Annex 1, Section 1.1.D) 1.3.3 temporal lobe (Annex 1, yes/no Section 1.2.A) To be included in this section, patient must have simple partial seizures (autonomic, psychic, sensory, epigastric sensation) or complex partial seizures (motor arrest with oro-alimentary automatisms) with gradual recovery and post-ictal amnesia 1.3.3.1 amygialo-hippocampal yes/no (Annex 1, Section 1.2.A1) 1.3.3.2 lateral temporal (Annex 1, yes/no Section 1.2.A2) 1.3.4. frontal lobe (Annex 1, yes/no Section 1.2.B) To be included in this section, patient must have seizures as shown. in one of the categories indicated below 1.3.4.1 supplementary motor yes/no (Annex 1, Section 1.2.B1) 1.3.4.2 cingulate (Annex 1, Section yes/no 1.2.B2) 1.3.4.3 anterior frontopolar (Annex yes/no 1, Section 1.2.B3) 1.3.4.4 orbitofrontal (Annex 1, yes/no Section 1.2.B4)

231

1.3.4.5

dorsolateral (Annex 1, yes/no Section 1.2.B5) 1.3.4.6 opercular (Annex 1, Section yes/no 1.2.B6) 1.3.4.7 motor (Annex 1, Section yes/no 1.2.B7) 1.3.5 parietal lobe (Annex 1, yes/no Section 1.2.C To be included in this section, patient must have simple partial seizures, predominantly sensory with visual phenomena and/or asomatognosia and/or vertigo or disorientation and/or receptive or conductive aphasia and/or lateralized genital sensations 1.3.6 occipital lobe (Annex 1, yes/no Section 1.2.D) To be included in this section, patient must have simple partial (mostly visual) or secondarily generalized seizures. Complex partial seizures reflect spread beyond the occipital lobe

A.2. Patient has generalized seizures Patient may have a generalized (G) epilepsy. Patient has: Normal inter-ictal neurological findings EEG with normal background and when present, generalized, bilateral, synchronous, symmetrical epileptiform discharges increased by slow sleep No etiological factors No previous and/or current epileptogenic disorders Normal neuroradiological investigation “

“

yes/no yes/no

yes/no yes/no yes/no

If all the above criteria are satisfied, patient may have idiopathic G epilepsy (see Section A.2.1); if one or more of the above criteria are not satisfied and an etiologic factor is not documented, patient may have symptomatic or cryptogenic G epilepsy (see Section A.2.2);

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

232 “

“

if one or more of the above criteria are not satisfied and a nonspecific etiology is documented, patient may have a symptomatic G epilepsy of nonspecific etiology (see Section A.2.3.1); if one or more of the above criteria are not satisfied and a specific etiology is documented, patient may have a symptomatic G epilepsy of specific etiology (see Section A.2.3.2).

A.2.1. Idiopathic G epilepsy Patient may have 2.1.1

2.1.2 2.1.3

2.1.4 2.1.5 2.1.6 2.1.7

2.1.8 2.1.9

benign neonatal familial convulsions (Annex 1, Section 2.1.A) benign neonatal convulsions (Annex 1, Section 2.1.B) benign myoclonic epilepsy in infancy (Annex 1, Section 2.1.C) childhood absence epilepsy (Annex 1, Section 2.1.D) juvenile absence epilepsy (Annex 1, Section 2.1.E) juvenile myoclonic epilepsy (Annex 1, Section 2.1.F) epilepsy with generalized tonic–clonic seizures on awakening (Annex 1, Section 2.1.G) other G idiopathic epilepsies not defined above epilepsies with seizures precipitated by specific modes of activation (Annex 1, Section 2.1.H)

yes/no

yes/no yes/no

yes/no yes/no

2.2.1.1

2.2.2 2.2.3

2.2.4

yes/no

yes/no yes/no

west syndrome (Annex 1, yes/no Section 2.2.A) symptomatic west syndrome yes/no (Annex 1, Section 2.2.A1)

cryptogenic west syndrome (Annex 1, Section 2.2.A2) Lennox–Gastaut syndrome (Annex 1, Section 2.2.B) epilepsy with myoclonic– astatic seizures (Annex 1, Section 2.2.C) epilepsy with myoclonic absences (Annex 1, Section 2.2.D)

yes/no yes/no yes/no

yes/no

A.2.3. Symptomatic G epilepsy (Annex 1, Section 2.3) To be included in this Section, patient must have bilateral asymmetrical EEG features, less rhythmic than those of Section 4.2.1 and clinical and/or neuropsychologic and/or neuroradiologic signs of diffuse encephalopathy. A.2.3.1. Symptomatic G epilepsy of nonspecific etiology. Patient may have: 2.3.1.1

2.3.1.2

yes/no

A.2.2. Symptomatic or cryptogenic G epilepsy Patient may have 2.2.1

2.2.1.2

2.3.1.3

early myoclonic encephalopathy (Annex 1, Section 2.3.1.A) early infantile epileptic encephalopathy with suppression burst (Annex 1, Section 2.3.1.B) other symptomatic epilepsy not defined above

yes/no

yes/no

yes/no

A.2.3.2. Symptomatic G epilepsy of specific etiology. Patient may have: 2.3.2.1 malformation 2.3.2.1.1 Aicardi syndrome (Annex 1, Section 2.3.2.Al) 2.3.2.1.2 lissencephaly-pachygyria (Annex 1, Section 2.3.2.A2) 2.3.2.1.3 tuberous sclerosis (Annex 1, Section 2.3.2.A3) 2.3.2.1.4 Sturge–Weber syndrome (Annex 1, Section 2.3.2.A4) 2.3.2.1.5 hypothalamic hamartoma (Annex 1, Section 2.3.2.A5)

yes/no yes/no yes/no yes/no yes/no yes/no

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

2.3.2.2 inborn errors of metabolism 2.3.2.2.1 early myoclonic encephalopathy (neonatal) (Annex 1, Section 2.3.2.B1) 2.3.2.2.2 phenylketonuria (infantile) (Annex 1, Section 2.3.2.B21) 2.3.2.2.3 phenylketonuria with biopterin deficiency (infantile) (Annex 1, Section 2.3.2.B22) 2.3.2.2.4 Tay-Sachs and Sandhoff disease (infantile) (Annex 1, Section 2.3.2.B23) 2.3.2.2.5 early infantile ceroid-lipofuscinosis (infantile) (Annex 1, Section 2.3.2.B24) 2.3.2.2.6 pyridoxine dependency (infantile) (Annex 1, Section 2.3.2.B25) 2.3.2.2.7 late infantile ceroid-lipofuscinosis (childhood) (Annex 1, Section 2.3.2.B31) 2.3.2.2.8 infantile Huntington’s disease (childhood) (Annex 1, Section 2.3.2.B32) 2.3.2.2.9 juvenile Gaucher disease (child/adolescent) (Annex 1, Section 2.3.2.B41) 2.3.2.2.10 juvenile ceroid-lipofuscinosis (child/adolescent) (Annex 1, Section 2.3.2.B42) 2.3.2.2.11 Lafora disease (child/adolescent) (Annex 1, Section 2.3.2.B43) 2.3.2.2.12 degenerative progressive myoclonic epilepsy (child/ adolescent) (Annex 1, Section 2.3.2.B44) 2.3.2.2.13 dissynergia cerebellaris myoclonica (child/adolescent) (Annex 1, Section 2.3.2.B45) 2.3.2.2.14 cherry red spot myoclonus syndrome (child/adolescent) (Annex 1, Section 2.3.2.B46)

yes/no yes/no

yes/no

yes/no

2.3.2.2.15 Ramsay–Hunt syndrome with mitochondrial myopathy (child/adolescent) (Annex 1, Section 2.3.2.B47) 2.3.2.2.16 adult ceroid-lipofuscinosis (adult) (Annex 1, Section 2.3.2.B51)

233

yes/no

yes/no

A.3. Partial and generalized seizures yes/no

Patient may have: 3.1.1

yes/no 3.1.2 yes/no 3.1.3 yes/no 3.1.4

neonatal seizures (Annex 1, Section 3.1.A) severe myoclonic epilepsy in infancy (Annex 1, Section 3.1.B) epilepsy with continuous SW during slow-wave sleep (Annex 1, Section 3.1.C) acquired epileptic aphasia (Annex 1, Section 3.1.D) other undetermined epilepsy not defined above

yes/no yes/no

yes/no

yes/no

yes/no

3.1.5

yes/no

yes/no

A.3.1. Uncertain seizures (partial or generalized at origin)

yes/no 3.2.1 3.2.2

sleep grand mal other (spec…)

yes/no

yes/no A.4. Special syndromes yes/no

yes/no

In order to be included in this section, patient must have partial or generalized seizures with specific characteristics and/or modes of onset, as shown below:

yes/no

4.1 4.1.1 4.1.2 4.1.3

situation-related seizures febrile seizures yes/no isolated seizures or statusyes/no epilepticus seizures after acute metabolicyes/no or toxic event (spec…)

234

G. Rinaldi et al. / Epilepsy Research 41 (2000) 223–234

References Blume, W.T., Berkovic, S.F., Dulac, O., 1997. Search for a better classification of the epilepsies. In: Engel, J. Jr, Pedley, T.A. (Eds.), Epilepsy: a Comprehensive Textbook. Lippincott-Raven Publishers, Philadelphia, pp. 779–789. Bodensteiner, J.B., Brownsworth, R.D., Knapik, J.R., Kanter, M.C., Cowan, L.D., Leviton, A., 1988. Interobserver variability in the ILAE classification of seizures in childhood. Epilepsia 29, 123 – 128. Commission on Classification and Terminology of the International League Against Epilepsy, 1981. Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia 22, 489–501. Commission on Classification and Terminology of the International League Against Epilepsy, 1989. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 30, 389 – 399. Eadie, M.J., 1996. The I.L.A.E. classification of the epilepsies applied retrospectively to 1902 patients. Epilepsy Res. 25, 277 – 284. Fleiss, J.L., 1981. Statistical Methods for Rates and Proportions, second ed. Wiley, New York. Landis, J.R., Koch, G.G., 1977. The measurement of observer agreement for categorical data. Biometrics 33, 159–174.

Loiseau, P., Duche´, B., Loiseau, J., 1991. Classification of epilepsies and epileptic syndromes in two different samples of patients. Epilepsia 32, 303 – 309. Manford, M., Hart, Y.M., Sander, J.W.A.S., Shorvon, S.D., 1992. The national general practice study of epilepsy. The syndromic classification of the International League Against Epilepsy applied to epilepsy in a general population. Arch. Neurol. 49, 801 – 808. Nordli, D.R. Jr, Bazil, C.W., Scheuer, M.L., Pedley, T.A., 1997. Recognition and classification of seizures in infants. Epilepsia 38, 553 – 560. Oka, E., Surnio, I., Ohtsuka, Y., Ohtahara, S., 1995. Neuroepidemiological study of childhood epilepsy by application of the international classification of epilepsies and epileptic syndromes (ILAE, 1989). Epilepsia 36, 658 – 661. OREp (Osservatorio Regionale per l’Epilessia), 1996. ILAE classification of epilepsies: its applicability and practical value of different diagnostic categories. Epilepsia 37, 1051 – 1059. Shah, K.X., Rajadhyaksha, S.B., Shah, V.S., Shah, N.S., Desai, V.G., 1992. Experience with International League Against Epilepsy Classifications of Epileptic Seizures (1981) and Epilepsies and Epileptic Syndrome (1989) in epileptic children in a developing country. Epilepsia 33, 1072 – 1077.