Contact Us

Ilae Facebook Ilae Twitter

Definition of Epilepsy

Request for Comments

An ILAE task force, under the leadership of Robert Fisher, has created a new definition of seizures and epilepsy that will become the League's working definition. As part of the process, we are seeking the broad input and guidance of our members and community from around the world to refine the document that we in the epilepsy community will use going forward as our definitions.

Please read the paper, especially with regard to how it will be used, and send us your comments. We ask you to send us your thoughts about the definitions before October 1. All comments will be posted below, and at the end of the comment period, a task force will be appointed by the Executive Committee to review the additional thoughts from our community.

Read the definition document.
Please send your comments to who will post them here on a weekly basis for the entire community to consider.


4 October, 2013

Thank for the opportunity to comment upon the ILAE proposal for new definitions for seizures and epilepsy. "At least 75% risk" is very difficult to estimate in many clinical situations and it will also hard to explain to the patients. A greater than 50% risk would be easy to handle.

The term "no longer present" is also hard to explain and can we really know that the risk is 0%. Remission is something that most people will understand and it is used in many other clinical settings.

Eva Kumlien

4 October, 2013

A disorder implies a functional disturbance, not necessarily lasting; whereas, a disease may (but not always) convey a more lasting derangement of normal function. Many heterogeneous health problems, for example cancer or diabetes, are comprised of numerous sub-disorders and are still considered to be diseases.The term "disorder is poorly understood by the public and minimizes the serious nature of epilepsy. The Task Force thought that epilepsy, as an enduring predisposition to seize, is best considered to be a disease.

Typical childhood absence epilepsy (T-CAE), have the following unique characterizations: a) onset at specific age, b) remittance in adulthood; c) no long-term effect on brain development, brain function or intelligence. Unfit expression of specific genes appearing at explicit stages should be considered an important etiology of T-CAE.

Evidence from animal models shows that the circuit among neurons of the reticular thalamic nucleus (RT), thalamic relay neurons (TC), and neo-cortical pyramidal cells involved in the formation of sleep spindles, is the mechanism of spike-wave discharges and absence seizures. T-type calcium channels are involved in the formation of low-threshold calcium spikes (SWD) in neurons, and lead to burst firing and rhythmic oscillations. Mice with an overexpression of the cacna1g gene have elevated T-type currents in their TC neurons and these mice express a typical absence seizure phenotype.

There are three T-type calcium channel genes: CACNA1G, CACNA1H, and CACNA1I. In the brain, different T-type calcium channel genes are co-expressed in a single neuron, like CACNA1H and CACNA1I co-expressed in RT neurons, where they are thought to be responsible for generating SWD in absence epilepsy. Additionally, the same sub-type of T-type calcium channels may have different alternative isoforms, which appear to be uniquely individual. These unique isoforms have specific biophysical properties like the voltage-dependency of channel activation, inactivation, deactivation and recovery from inactivation. It is extremely complicated when variable amounts of different T-type calcium channel splice variants are co-expressed in one neuron because very subtle modifications in the input of the T-type calcium current can drastically affect the physiological output of the thalamus neurons. It is therefore reasonable to speculate that some members of our global population have temporary exceptionally "strong T-type calcium currents because of their distinctive alternate transcription of T-type calcium channel genes. This has endowed them with the unique distinction of easily induced oscillation of synchronization in the thalamocortical circuit, and they could be positively inclined to development of typical absence seizures. Although this concept need further be proved. In regard this possibility, using "disorder to define "pure typical childhood absence epilepsy might be better than suing "disease because this sub-type of seizure might be the consequence of temporary unbalanced expression of genes such as T-type calcium channel genes, which would be adjust to normal with aged following the epigenetic modification of genes.

With best regards.
Yucai Chen

4 October, 2013

Comments from Paolo M. Rossini & Catello Vollono

3 October, 2013

Sorry to be late. I agree for this defintion.

Thank you,
Raidah Al-Baradie

3 October, 2013

Thanks so much for the opportunity to do such an important work. So sorry for delay. I think this definition surely gives us some advances.

I have only a small question about the classfication of epilepsy as disease or disorder. Disease implies there is damagement of normal structure, which means symptomatic epilepsy according to classfication of epilepsy. But idiopathic epilepsy, like which is from genetic reasons can not underlying these structural pathological state. So I would like to suggest whether it should be a disease or a disorder depends on classification of epilepsy itself.

Finally, so sorry for delay and thanks again for hard work of this new definition.

Warm regards
Xiaoting Hao

2 October, 2013

Thank you very much for the opportunity to comment the excellent work of the ILAE task force. I think that the three definitions are appropriate, only propose some modifications:

1. At least two unprovoked seizures occurring more than 24 hours apart;

2. One unprovoked seizure and a probability of further unprovoked seizures equal to 75% or more;

3. At least two seizures in a setting of reflex epilepsy.

I think that "remission" is a better term but should be used differentially according to etiology and epileptic syndrome.

I also think the definition of "oligo-epilepsy" should be the next task.

Best regards,
Iván O. Espinoza Quinteros

Return to top

2 October, 2013

We welcome very much the proposed definition of epilepsy, in particular because it indirectly emphasizes that a diagnosis actually implies a prognostic statement.

In our opinion the ILAE correctly states that we have insufficient support for defining "probable epilepsy", while at the same time the task force acknowledges the need for such a definition in the future.

It remains surprising that epilepsy is one of the last areas of neurology without probalistic criteria for diagnosis. Very recently, the ILAE has released its requirements for the diagnosis of psychogenic non-epileptic seizures, including a probalistic hierarchy that is based on the strength of clinical, neurophysiologic and combined information.

We believe that that epilepsy community deserves equally clear and practical criteria. As said in a number of comments below, we will ultimately need operational definitions of definitive – probable – possible etc. epilepsy. Only large clinical studies will show us where the most relevant uncertainty comes form. We suspect that incomplete observation or description is the most relevant factor in first fits and in early epilepsy. This will not be resolved, but is explicitly dealt with when describing spells as definite, probable or possible epileptic seizures.

To be accepted by neurologists, the scheme must be roughly in line with subjective impression of certainty and with. This is possible by assessing each individual item or sets of criteria in diagnostic studies using the STARD guidelines.

We are confident with the definition of epilepsy by a recurrence risk of 75%. It's time to STARD estimating the confidence intervals.

Bert U. Kleine and Helenius J. Schelhaas

1 October, 2013

I am grateful for the opportunity to comment on this document from the perspective of a paediatric neurologist seeing patients with suspected or diagnosed epilepsy.

The authors should be congratulated on developing an operational clinical definition reflecting clinical practice and the everyday challenges faced in diagnosis, including how we communicate with people with epilepsy and their families.

Endorsing a diagnosis of epilepsy after a single seizure in certain circumstances reflects modern clinical practice where the clinician seeks at first consultation to describe the seizure type(s), a syndrome, and an aetiology. This may not be possible, but with history, EEG and imaging data a syndromic diagnosis with or without a related aetiology can be made at the first consultation . In paediatric practice it is not unusual to see a child with Epilepsy with Centro-temporal Spikes or a child with a familial genetic epilepsy after their first seizure. It seems appropriate to be able to say that this individual has epilepsy. Even with a confident syndromic diagnosis we may not be able to say in specific syndromes that there is a 75% recurrence risk. Being able to give a diagnosis of epilepsy will inform management and communicate better the risk of further seizures to the individual and family.

Like many discussants I have an immediate reaction to absolute figures such as the 75% figure quoted however the authors use "high probability" with ≥75% in brackets and I think this is reasonable as broad guidance given the epidemiological data presented in the paper. A comment that a 75% risk may not apply to all epilepsy syndromes might be included.
I agree that the issue of treatment and diagnosis are related but different issues. I do not think that allowing a diagnosis in these defined circumstances will lead to individuals being treated inappropriately.

I prefer the proposed terminology of "Epilepsy no longer present" than in remission or cured. "In remission" to physicians and people with epilepsy implies that the epilepsy is in abeyance but may come back at any point. It also implies that the individual and physician needs to be actively monitoring for the recurrence of symptoms. Children and their parents appreciate being told that when they are 16 or 18 and asked to complete an employment form that it is likely they will not have to write that they have epilepsy if, at present, they have a self limited childhood epilepsy syndrome. I advise them that they will only have to mention epilepsy if the form asks if they have ever had epilepsy.

I prefer the term disorder to disease. Will save the ILAE having to rename the journal recently taken under its wing.

The authors have distilled a highly complex area into straightforward operational definitions, which reflect modern clinical practice.

Sameer Zuberi

1 October, 2013

Thank you for giving us an opportunity to comment on the proposed new definition of seizure and epilepsy. We welcome this new definition, that helps the clinician to make firm diagnosis even with single seizures in some situations. The corroborative evidence from imaging and or EEG examinations are given the weightage of a second seizure in such instances. It is not clear, how we should handle seizures with small evanescent neurocysticercal lesions, that often do not require long term AEDs (but would fulfill the criteria one seizure and positive MRI).

We also feel that the criteria for remission need to be amplified further.

Sanjeev V Thomas

1 October, 2013

Thank you for the opportunity to contribute with this definition. Please find below some thoughts and suggestions:

-Epilepsy is a neurological disease characterized by recurrent seizures. These seizures are separated by a period of normal neurological and cognitive function independently of the duration of such period.

-Epileptic condition is a neurological state when a seizure is associated with a brain injury and/or evoked by an internal or external stimulus and poses a risk of recurrency.

The void in our understanding the fundamental mechanism of ictiogenesis, or epileptogenesis, makes difficult to provide a precise definition (conceptual, operational, etc.) and in order to update this operative definition some current insights from basic and clinical research should be considered.

In addition, the definition should be, clear, precise, objective, concise and easy to understand not only by epileptologists, but also by all physicians so they can clearly communicate the operative definition with patients and health care personnel.

Alberto E. Musto

Return to top

1 October, 2013

We thank the authors for this job and to give the opportunity to comment on the proposed new definition.

Disorder vs disease: In pediatric epilepsy is well known when epileptic seizures are a disease and when they are a disorder. Epilepsy is a disorder when only predisposition is the causative factor, such as in absences, Rolandic epilepsy and Panayiotopoulos syndrome. It is no different from what happens in neuro-mediated syncope, no one would say that those who have a predisposition to syncope have a disease but certainly they suffer from a disorder. Conversely, epilepsy is a disease when it is the expression of an anatomical-structural pathological state, and in these cases is not epilepsy to be a disease but the underlying pathological condition, such as metabolic pathology, neurodegenerative conditions, genetic syndromes. In these cases, seizures are only one of the symptoms, associated with others (cognitive delay, motor dysfunctions etc. etc) all expression of the underlying pathology. If you call epilepsy as disease or disorder depends on the context in which epilepsy occur, as pediatric epilepsies teach that epilepsy is far from being a single condition, but it is the expression of different underlying dysfunction ranging from simple predisposition the most severe structural anatomical disease.

Definition: recurrence risk
General consideration is that this definition is focused more on the needs of epidemiologists and forensic scientists than clinicians.
Since there are no epidemiological data that precisely point out the percentage of recurrence risk we think that is difficult to apply to the diagnosis of epilepsy only if there is a recurrence risk higher than 75%.

By a clinical point of view this definition is not useful since we have to give all suggestions regarding risks of successive seizures after the first one. We do not think that a hypothetical recurrence risk of 40% is really different from a risk of 75% percent, in both situations we have to inform patients and give advises about preventative measures. The important think is that this definition does not influence the therapeutic decisions.

Complex Febrile Seizures. Regarding complex febrile seizures, personal data do not justify epidemiological work of the past; in the previous case series were probably included patients with epilepsy related to genetic etiology with onset with complex febrile seizures (such as Dravet Syndrome), thus invalidating the results. So, the risk of subsequent epilepsy should be recalculated in the light of recent data (If there are). In our experience the occurrence of complex febrile seizures (repeated febrile seizures in 24 hour or during the duration of the febrile illness and duration commonly defined as longer than 10 minutes) do not represent a risk for future epilepsy (unpublished data) at least when clinical data indicate no other underlying condition than febrile seizures.

Finally we think it should be always emphasized, as was pointed out in the 2005 classification, that the definition of epilepsy requires the occurrence of at least one epileptic seizure and not only the presence of EEG abnormality discovered by chance.

Lucia Fusco
Nicola Specchio
Federico Vigevano

1 October, 2013

Thank you for the opportunity to comment on the proposed changes to the definition of epilepsy.

Item 2 in Table 2 focuses on an operational definition for diagnosing epilepsy based on a single unprovoked seizure and a probability of further seizures similar to the general recurrence risk after two unprovoked seizures (approximately 75% or more).  In addition to the clinical implications, this new definition has significant implications for epilepsy research and clinical trials. As presented, the definition relies on the individual physician's discretion to estimate the risk of recurrence. In the absence of a clear evidence base, this has the potential to introduce subjectivity and risk of people without epilepsy being referred into clinical trials. In trials aimed at evaluating a drug's potential effectiveness for newly diagnosed epilepsy, inclusion of patients who do not have significant risk of further seizures could inflate the placebo response and decrease the likelihood of demonstrating a statistically significant difference between an investigational drug and placebo. The definition may also have implications for comparative effectiveness research where comparability of populations across clinical trials must be demonstrated.  If this definition is endorsed for clinical use, it may be necessary to reflect the change in enrollment criteria proposed for clinical trials to ensure consistency between practice and trials as well as harmonization across compounds in clinical development. We would like to propose the formation of a collaborative working group involving members of the ILAE, industry and others to develop guidelines for implementation of these new definitions into clinical trials.

Thank you again for soliciting feedback from the broad epilepsy community and the hard work that has gone into creation of these recommendations.

Respectfully submitted on behalf of UCB, Inc.,
Kelly Simontacchi

1 October, 2013

I would like to congratulate Task Force for this proposal of a new definition of epilepsy, and I also would like to thank for the opportunity to comment the document. I agree with most of the statements and concepts illustrated in the paper, but I have some concerns that are shared by many other colleagues:

1) I prefer to use the term "disorder" instead of "disease" for epilepsy. There are specific cases of "epileptic diseases", but the term "disease" for epilepsy, in a general view, sounds misleading. There would be also consequences in stigma and social acceptance of epilepsy as a "disease", especially for persons with seizures not due to known cerebral lesions, and with normal intellectual abilities.

2) It is very difficult in the everyday clinical practice to calculate a high probability (≥75%) of future seizures in a patient who presented a single unprovoked seizure. The Task Force gives some examples of ≥75% recurrence risk, but this item could lead to subjective and disputable definitions of epilepsy. A more precise effort on epidemiological literature is needed before to propose this statement in a new operational definition of epilepsy that will be used worldwide both by expert epileptologists and general neurologists. In the meanwhile, I would propose to maintain the old diagnostic criteria ("at least two unprovoked seizures occurring more than 24 hours apart").

Maurizio Elia

Return to top

1 October, 2013

I read with interest your proposal for "an operational definition of epilepsy". Thank you for this huge deal of work and for initiating this discussion. I have some remarks from a pediatric clinical point of vue:

  1. The major point that puzzled me was about the main objective of giving the diagnosis of epilepsy after a first unprovoked seizure and not to be able to wait for the second episode. I cannot see many situations where this is of major impact on the patient follow up without going into a useless listing.
  2. The over 75% of risk is a difficult issue to make by a clinician and needs to some extent an additional expertise that might not be available since after a first seizure, all over the world, patients cannot (and fortunately in a way) access to a specialized center. So this expert opinion on the 75% risk of probability a second seizure or the 25% risk of not doing another episode might not be available. The risk from my point of vue is confusion between this concept of diagnosis (diagnosis with a good analysis of the risk factors that might be on EEG, on MRI, on genetic results... or from the syndromes per se for instance IS after the first salve is to be considered as epilepsy) and the treatment. My worry is that in the absence of all of these additional analysis criteria, clinicians will play "the safe side" and use in excess this definition to do "preventive therapy" in all patients after a first seizure.
  3. In addition, although you precised very clearly the difference between diagnosis and treatment, the example of cancer is a bit puzzling. What you are announcing to a patient is that he presents epilepsy, it means at least for me that a recurrence is highly probable and that he presents a disease. The first question of the physician and the patient (or the family) is how to prevent these seizures and continue a normal life. In Pediatrics, it is worthy to know that Rolandic epilepsy for instance (and even Febrile seizures) are much more treated with AEDs in first line and secondary centres than in tertiary expert centres since all of this risk benefit evaluation is directly correlated to expertise and that convincing parents "not to treat" such a terrifying event needs the intervention of an expert. Giving the tiquette of epilepsy after the first seizure could increase the risk for an unuseful therapy.
  4. The high recurrence risk calculation seems asking too much since the authors stated in the paragraph (high recurrence risk) that "it is important to note that a single seizure plus a lesion or plus epileptiform EEG spikes does not automatically satisfy criteria... Data must be available to support an approximate 75% risk.....when data are insufficient to demonstrate more than 75% risk...." This is really a very complicated paragraph and from my clinician point of vue, only in some syndromes or with some etiologies we have these data (Infantile spasms, Dravet syndrome... Do we have it in TSC with multiple tubers? Do we have it with porencephaly, post anoxic ischemic insult...?

My fear is that there will be an confusion between diagnosis and treatment issues and the example of cancer does not seem the best since the relation between treatment and diagnosis is not the same.

The point is that we might have the diagnosis of epilepsy (and the stigma of epilepsy ...and may be the treatment of epilepsy) for at least 25% of patients who will not develop epilepsy (?).

For the definition of the "remission", I have a problem to understand the subtle difference between remission and no longer present. Is it my problem of understanding this or is there a need to renew definitions and use new words to say the almost same thing. I believe that remission is an accepted medical concept and I would prefer to use it.

The other major point is that such definition od epilepsy after a single unprovoked seizure opens wildly the way for other emerging problems that are more difficult to handle. The word "probable epilepsy" will be as predicted by the authors the next step! How do authors believe that the community (physicians and patients) will deal with it? Again, we will need personal expertise, what will be probable for the first will be possible for the others....

Finally, operational definition for me is something that is easy to handle, easy to widespread and does not need sophisticated expertise. I believe that the title of this excellent proposition might be changed since it could not be as operational as authors would like it to be and it will not be used easily by first and may be even second line physicians.

These are my thoughts and I thank again the task force for this interesting proposal. Authors presented in a very clear way how an expert could consider epilepsy after the first unprovoked seizure and does not wait always until a second episode to evoke the diagnosis of epilepsy.

Rima Nabbout

1 October, 2013

The Task Force are to be congratulated on their thorough and thoughtful paper. I find it to be very clearly written and highlights the areas where this operational definition is difficult.

With respect to the terminology: epilepsy as "disease" rather than "disorder":
I instinctively agree with the view that epilepsy should be called a "disease" rather than "disorder", a more nebulous term, in order to emphasise the seriousness of the problem.  I understand that this may be problematic for some patient groups, perceived as increasing the stigma of epilepsy. However, if we consider the definition purely in medical terms, the term "disease" appears more appropriate.  There are many diseases that run an intermittent course and are never truly cured, and there are others that remit spontaneously. Thus the natural course of epilepsies does fit into the general pattern of what we would call a "disease". The aetiologies may be different, but the epilepsy itself as a pathological condition of the brain characterised by spontaneous occurrence of seizures deserves a "disease" status.

With respect to definition of what constitutes epilepsy:
Point 1 of Table 2 - The classic definition of epilepsy is maintained by the new proposal – this is not controversial.

Point 2 of Table 2 - The novel proposal is to diagnose epilepsy after a single unprovoked seizure when the risk of recurrence is estimated to be 75% or more, i.e. similar to the risk of recurrence after 2 unprovoked seizures.

I think this is conceptually correct and supported by some epidemiological data, but I share the concern of many colleagues that the risk of recurrence would be difficult to quantify in some, probably many, individual  cases, i.e. there may be operational issues in applying this definition.

Having said this, I am in favour of keeping this definition with the caveat that it can be used where there is clear epidemiological evidence that in certain clinical settings the risk of recurrence is approaching/exceeding this threshold. If the risk is unknown, because we simply have no data, I would go along with the recommendation of the Task Force to acknowledge this lack of evidence and use the classic definition (i.e. diagnosis not made until a second spontaneous seizure occurs). We have a precedent in McDonald's diagnostic criteria for MS where the diagnosis can be made on the basis on radiological MRI progression, not just using classic Poser's clinical criteria.

The threshold of 75% is somewhat arbitrary,  but a degree of arbitrary decision on something that is a risk continuum is always going to be difficult and if we are to use this new definition we have to set the threshold somewhere. The 75% figure seems reasonable, if we accept that any figure over 50% would be arbitrary. The Task Force recommends "approximately" 70-75% and for most patients and clinicians it would translate into a "high probability" of seizure recurrence.  This is arbitrary, but extrapolated from epidemiological  data on recurrence after 2 spontaneous seizures.

However, it is important that the Task Force specify the time period to which this risk applies (e.g. recurrence of seizures during the next 4 years). Whilst in the clinical practice these "shades of grey" may be overcome by a careful discussion with patients (who will always tend to wish to avoid further seizures and who will not like taking medication unless necessary) this diagnostic concept may pose some operational difficulties in medico-legal or occupational setting.

The diagnosis of epilepsy is separate from the decision to start treatment. In our current practice, we would only start treatment if the risk of recurrence after the first seizure is thought to be high (e.g. brain tumour or JME with supportive EEG). The modified definition of epilepsy would not change our practice with respect to these high risk patients. The potential danger is that more patients (with lower risk of recurrence) would be diagnosed and then treated with antiepileptic drugs unnecessarily after the first seizure purely because their risk of recurrence was overestimated.

In order to assist with therapeutic decisions, it would be helpful  if, in addition to the examples of specific cases included in the, the Task Force could produce the state of art compilation of epidemiological data on risk of seizure recurrence in certain situations, stating that in some settings the data are not available.

Point 3 Table 2 - I agree with the definition proposed

The issue of "epilepsy no longer present":
I would prefer the term "in remission" in reference to cases that have been seizure free for 10 years and off antiepileptic drugs. The exception should probably be made for benign (age-dependent) epilepsy syndromes of childhood that have passed the applicable age and I wonder whether a shorter time seizure-free and off antiepileptic drugs should apply to these patients (? 5 years).
My main reservation about the term "epilepsy no longer present" is the difficulty in proving the "absence of epilepsy" and the fact that the risk of epilepsy will probably be never truly zero for the majority of patients.

Kasia Sieradzan

Return to top

1 October, 2013

Commendable job by the Task Force on proposing the new operational definition of Epilepsy. Some of my thoughts on the proposed definition are:

1. The diagnosis of seizure may not be secure at its first occurrence. It may have not been witnessed. It is not uncommon in clinical practice to (mis) label the first spell as seizure only to be clear about the true nature of spell when it has recurred. EEG is frequently done after seizure to assess the risk. A false positive EEG in such situations may unnecessarily bias the risk assessment.

2. The enduring tendency for recurrence can only absolutely certain when second seizure has occurred. Having a high risk (>70-75%) also means that there is 25-30% risk of not having future seizures in lifetime. So these patients will have to wait for at least for 10 years to come clear.

3. Inter-rater variability in the the risk assessment is another factor that needs to be considered in the definition. Based on the new proposed definition, a patient may be diagnosed with epilepsy by one physician but other physician may chose to call it a single seizure.

I sincerely thank ILAE for giving me the oppurtunity to offer my comments on the issue

Jatinder Singh Goraya

1 October, 2013

Thank you for the opportunity to comment on the proposed new definition by the ILAE Task Force.

Psychiatric aspects of epilepsy do not have properly clarified and defined etiology and pathogenesis. So they probably can not be involved in the definition, but may be considered in a more detailed explanation. The classification of neuropsychiatric disorders in epilepsy (2007) is a new begining of considering epilepsy with psychiatric aspects as a part of it.

Best Regards
Slavoljub Vlajin

1 October, 2013

I commend and congratulate the task force on the huge efforts put into expanding the default definition of epilepsy into a conceptual definition and keeping the default definition in place as fall back. This will really achieve a lot of things pointed out so well be the task force.However, the thoughts that come to mind are the following...

1. In non age dependent syndromes when do we say epilepsy is no longer present? It may be difficult to define a time period so probably 10 years since last seizure seems appropriate as an arbitrary figure  based on current data.

2 ."Disease" signifies more stigma than "disorder" in most cultures and the use of epilepsy in the context of disease may be counterproductive to the old stand taken for several decades such as  the Global campaign or "bringing epilepsy out of the shadows " ...especially in some situations such as marriage and employment.

This is of great relevance to the developing world where the scenarios, the knowledge , attitudes and practice(KAP) about  epilepsy is drastically different from the developed world and we would be better downgrading the impact of the diagnosis as "disorder"(with an inherent meaning that it can be controlled) rather than a "disease" for which we currently state in this very proposed definition ,that do not talk of cure but only "disease no longer present".

It may imply to the laymen the person harbours disease which is "no longer  present "and can  resurge anytime.

What's in a name? It does not change reality of recurrence but changes stigma and perception of person with epilepsy, family and society a lot!

The group should probably be more globally representative to bring in many diverse situations and cross cultural issues especially from developing countries.

3.However definition and hence treatment in a few of the conditions with  high risk of recurrence after a single unprovoked seizure seems appropriate in a few situations. 

After a single unprovoked seizure, defining risks that may not be known for majority of conditions due to poor  data at a given time probably will make the definitions subjective.This may make it difficult to compare these gruops due to difference in definitions and inclusion criteria in scientific and epidemiological studies in the future.

Also older studies and new studies (pre and post new conceptual definition) will be impossible to compare.

For example, many conditions such as newly emerging autoimmune causes of epilepsy or epileptic encephlopathy may have a high but yet undefined risk of frequent seizures in long term outcome and in the acute phase may be defined as provoked seizures.This discrepancy may add to confusion not so much in clinical judgement but conducting and interpreting studies on various epilepsy subtypes and aetiologies.

Is it not easier to keep the default definition and declare a table of conditions for which the risks are more than 75% and add to this as we learn from future research...the .non neurologist ,physician ,pediatrician can follow this much better.(and these are the primary caregivers who diagnose epilepsy,the neurologist is mostly concerned with affirmation or negation of the diagnosis)?

I sincerely thank you for giving me this opportunity to read the document and allowing me to make suggestions or comments on this excellent work.

Nandan Yardi

Return to top

1 October, 2013

My comments on operational clinical definition of Epilepsy:

1.  The main reason for enlarging the definition by adding points no. 2 & 3 of the proposed operational clinical definition appears to be presumed inadequate treatment for patients with high risk of recurrence after a single unprovoked seizure. If this is so then I think we need data to show that patients with one unprovoked seizure and a brain lesion which is associated with high recurrence risk of 70-75% and not adequately treated. In my practice, I practically never find such a case and I therefore think that there is no strong reason to enlarge the definition.

2.  Adding point 2 and 3 will result in "labeling" of patients with single unprovoked seizure with a lesion like stroke or traumatic brain injury to the "label" epilepsy. This will be a great disservice to such patients because change of label from seizure to epilepsy will have psychological, legal and/or employment implications resulting in more harm than good to such patients.

3.  I do not agree with use of the word ‘disease' because by definition a disease is "A pathological condition of a part, organ, or system of an organism resulting from various causes, such as infection, genetic defect, or environmental stress, and characterized by an identifiable group of signs or symptoms".  On the contrary, it is better referred to as a ‘disorder' because different diseases and conditions can lead to epilepsy. Under the subheading "High recurrence risk", there is assignment of a task to the physician caring for patient to make a determination of high recurrence risk (75% or more) in patients with single unprovoked seizure. I think most physicians do not have the knowledge or competence to make such determination.

4.  Under the same subheading "High recurrence risk", a phrase in line 16 say "75% risk or more for another lifetime seizure". The word lifetime here contradicts the notion of 10 years seizure freedom with medication constituting the state of "epilepsy not present". It may be worthwhile to consider 10-year risk rather than life time risk.

5.  Under subheading "Implication for treatment", line 3, 4 & 5 suggests that the proposed definition will provide support to a physician who wishes to treat a patient with high recurrence risk after the unprovoked seizure. I do not think lack of such of support is preventing physicians from treating such patients. In my practice, I find that nearly all patients being treated without any hesitation by the treating physician.

6.  Under the subheading "Reflex epilepsy", line 3, 4, 5 implies justifying use of the work ‘reflex epilepsy' on the basis of enduring abnormal predisposition to the services but it ignores immense suffering of patients who will be victims of the label ‘Reflex epilepsy' rather than the current' reflex seizures. No amount of physiological justification can compensate the adverse effect of "labeling".

7. I agree with the concept of ‘epilepsy no longer present'.

Kameshwar Prasad

1 October, 2013

Thank you very much for the opportunity to read the document and to comment on the proposed new definition by the ILAE Task Force. We hope that this effort will lead to the prospect of a clear and precise definition to be agreed and used in clinical practice. Below please find our comments:

1) We think it is more correct to define epilepsy a disorder of brain function rather than a disease (as described in the "conceptual"
definition of epilepsy by Fisher et al, 2005). Disorder in fact indicates temporary or permanent derangement of electrical function of brain. Moreover seizures and epilepsy are symptom of a disease where the cause may or may not be known

2) Reflex epilepsy: we think the definition you made about the reflex epilepsy is appropriate

3) We think it should be emphasized that family history of epilepsy and genetic backbround are important to define the recurrent risk of epilepsy

4) Unprovoked seizures separated in time: we agree with the Task Force.
Brain MRI, EEG and when possible the knowledge of aetiology or pathology, allow a diagnosis of epilepsy that to be made after a single seizure if risk or recurrence is based on prior knowledge

5) We think that the term "no longer present" is more correct respect "remission".

Best regards
Renato Borgatti and Romina Romaniello

1 October, 2013

We read authors document with interest. Author's proposals are intended to change the definition of epilepsy, which according to us adds more problems than solutions.

1. 75% risk threshold renders number of conceptual and practical limitations: (a) it is a matter of separate debate in itself on why it should be 75% and not anything else; the author's basis of this is not convincing; (b) this risk is most certainly undeterminable in low-income, less trained, and resource-deprived settings such as many areas in low and middle income countries (LMIC) and is against the principle of decentralization; which is a forthright aim in epilepsy services; (c) it might be too optimistic to expect that there can be a reliable, universally acceptable mechanism that might help determine this risk threshold, given the complexities of each patient, of populations, of epilepsy, and in the light of numerous risk and prognostic factors (both at individual and population levels). It is important to mention here that putting 75% risk condition would mean that underlying clinical conditions would always be known and present, which is not always the case. Today more than 50% of cases do not have an etiological explanation; (d) current treatment practice doesn't have any correlation with treating only those who have 75% risk, those with lesser risk are also treated; (e) there is no correlation of this risk to time period even though authors referred on page 4: the risk is 40-52% over a period of four years; (f) there is no explanation about those who have >75% risk of seizure but haven't yet developed any visibly seizures.

2. Traditionally, epilepsy is said to occur once two unprovoked seizures have noticeably taken place, which seems to be a fair threshold for diagnosis, universally. Occurrence of two seizures is also important because repeat occurrence of a seizure (i.e. second seizure) is a kind of "confirmatory test" that the disease has now occurred in a person or person is now diseased with epilepsy. It is very likely that one seizure theory (±75% risk) would make inclusion of acute seizures as epilepsy, much more likely (e.g. from neurocysticercosis).

3. The utilizers of the definition of epilepsy are not only specialist, i.e., epileptologists, but general practitioners, paramedical health staff, patients, and general public as well. There is a possible learning from the definitions of hypertension or diabetes, which even general public is able to understand, determine, and relate easily. Proposed definition of epilepsy doesn't seem to match this requirement, and is complicatingthe assessment, which might contribute in turn into a mystification of this condition.

4. Numerous community based studies have been carried out on the prevalence, incidence, outcome and aetiology of epilepsy in LMICs with lack of uniform methodology. Such studies make it difficult to obtain comparable data. A change in definition will further complicate the data and collection and the comparability of data. In addition, the classification of epilepsy by aetiology has also been changed. The intention is to group epilepsy in presumed genetic, structural/metabolic and unknown categories. Without sophisticated additional investigations, it is virtually impossible to classify epilepsy in to these categories which is leading again in LMIC to an increased lack of comparable data, and hence to a blurring of the picture of the burden.

5. As others have already mentioned, our aim should for the moment be to first identify precise recurrence risks (in different patient, population, comorbid, and risk factor scenarios and their likely benefit from anti-epileptic treatment. Then, the second step should be to modify the definition of epilepsy, as per need and results obtained, and not vice-versa.

Some of our other remarks are already covered in the sentiments expressed by other colleagues. It is evident to us, that the proposed definition is unlikely to be applied uniformly in clinical practice of developed and developing countries, and that in LMIC the difficulties will increase to come to a definite diagnosis of people with epilepsy. Hence, treatment gap is likely to increase. Overall, we feel that it might be useful to keep in mind the issues that arise in areas with inadequate investigation and high treatment gap. LMICs are already suffering from high gaps in epilepsy care and these changes would put an enormous additional burden on them if the definition of epilepsy would be changed as proposed. Finally, it would have been more appreciable to have members from LMICs and other stakeholder groups as well to have their concerns and requirements taken into account while defining epilepsy.

Devender Bhalla (France), Nadir Bharucha (India), Peter Odermatt (Switzerland), Pierre-Marie Preux (France)

Return to top

1 October, 2013

In general, I find this new draft welcome. It is a thorough and thoughtful report, and the task force is to be congratulated with the result. I have some comments:

I am a bit uneasy with the term "disease" and will prefer the broader and less exact term "condition" as the spectrum of epilepsy (or better: epilepsies) is so broad and heterogenous.

Whether two unprovoked seizures within 24 hours should be regarded as one single episode, has been debated; I remember that Camfield & Camfield, at the 2005 Bethel-Cleveland symposium, stated that such episodes had the same prognostic weight as two seizures with more than 24 between.

A very essential point is the risk of recurrence after one unprovoked seizure; from the present body of evidence, the cut-off value (75 %) is well chosen, but may still sound a bit arbitrary.

Also, the term "unprovoked" is not straightforward; where to draw the line? Alcohol withdrawal and sleep deprivation may have quantitative aspects as well as qualitative; how much should be required to call a seizure provoked? In, for example, juvenile myoclonic epilepsy, these factors are operating at an intensity far below the threshold for most normal people, whereas massive sleep deprivation will be able to provoke seizures in some persons who never have seizures.

But this problem applies equally to the 2005 definition. It could b an independent issue at a future congress.

"Epilepsy no longer present": this paragraph seems to be somewhat a play with words.  The proposed term "Epilepsy no longer present" has one advantage, namely that it is a purely descriptive term of the patients' present condition: no seizures during the last (10) years off AEDs. "Remission" and "cure" have some and different  interpretative bearings on the future - "cure" can mean something like a guarantee that no more seizures will occur, while "remission" is less committing.  In  "remission", the duration can be set at the discretion of the investigator; shorter periods (2-5 years) are commonly used, and in often includes patients still in treatment.

Among conditions with (deceptive) remissions one could mention mesial temporal lobe epilepsy in children; some of those patients can have remissions long enough to letting them obtain a drivers' license.

In the new EU regulations, a drivers' license for taxi, bus and trucks can only be issued if the applicant has been seizure-free for at least 10 years off antiepileptic treatment. Hence, this requirement is in accordance with the term "Epilepsy no longer present", and it makes very good sense.

I would therefore agree with the term "Epilepsy no longer present", though it is a bit clumsy. But it has an emphasis on prognosis which the term "remission" lacks.

"Epilepsy no longer present" = seizure-free for ≥10 years without treatment.

The most problematic aspect of the new definition is the fact, that its implementation – as the authors rightly point it out – requires access to modern hi-tech investigations, e.g. video-EEG and neuroimaging incl. high-resolution MRI. Also here, there are quantitative aspects; as imaging data now play a role in the diagnosis of epilepsy, an increase in strength from 1.5 to 3 or 7 T MRI may make a significant difference in number of patients with epilepsy! So, even in developed countries, differences in figures may emerge.

As the authors point it out, the old definition could still be the default model, and the new one serve as a superstructure.

Jørgen Alving

1 October, 2013

Thank you for the opportunity to comment this important work. Here are several topics that I consider relevant for reflexion:

From a clinical perspective, I think that it is relevant to consider the recurrence of a single seizure, but I share the concern with the dificulty to ascertain the risk and probably it would be usefull to give a table with the recurrence risk accepted for the most common conditions. Although it is virtually impossible to be completelly exaustive, a general table could help the clinicians in most of the situations.

The term disease could be usefull for patients, but the stigma associated should be taken in consideration. I consider ten years without medication and without seizures a very long period, when we think about the legal implications for the patients. I prefer the term remission.

From a research perspective, since epilepsy is a manifestation of a neuronal/circuitry disfunction, it would be very useful to start to individualize the molecular changes and possibly associated biomarkers that can be predictive of a reduced seizure threshold and/or seizure recurrence. Gathering the epidemiological data with the data from basic neuroscience is relevant for a correct definition and ultimately for a targeted and mechanism based treatment.

With my best regards,
Sofia Duarte

1 October, 2013

We agree with the proposed definition which seems more practical and allows flexibility to the practicing physician:

1. In clinical practice, the term "two unprovoked seizures" is inadequate in some circumstances such as some symptomatic epilepsies and epileptic syndromes. This definition will allow the diagnosis of epilepsy after the first seizure in many situations. However the rate of 75% of recurrence risk is not applicable for all epilepsies.

2.Reflex epilepsy:
The idea to individualize the reflex epilepsy is very interesting; it cannot fit the definition of unprovoked seizures (by definition)
I suggest to add the term "reflex" seizure in this context

3. Epilepsy no longer present:
This idea is acceptable; however, I think that the terms "remission" already used in practice seem to be more adequate than the term "no longer present" for epilepsy. The recurrence may be very delayed in some circumstances and the prognosis is difficult to predict in some patients.

Thank you for the opportunity to review and comment this great work.

Prof. Riadh Gouider, Dr Amina Gargouri, Dr Yosr Hizem

Return to top

1 October, 2013

Thank you for asking feedback on this very important Definition. I have these few comments/suggestions:

- Implications for treatment: I agree that diagnosis and treatment are two distinct entities. But this may not be easy in case of epilepsy, which is very different from other diseases, in the sense that initiating treatment is a big dilemma in many patients. A lot of epilepsy is treated by the primary care physicians who rely on the formal definition of epilepsy to make treatment decisions easy for them. It may be important for them to have as less ambiguity as possible. Leaving the definition too open ended and asking them to calculate the risk of recurrence (more than 75% or not) may be putting too much responsibility on these doctors. It may be prudent to strike a balance between going for the most accurate description, and keeping it simple and useful. We may at sometime need to enumerate criteria (Clinical/EEG/Imaging/Etiology) which denote a higher risk of recurrence after a single seizure & can be considered equivalent to a ‘second seizure'.

- Unprovoked seizures separated in time: We can also contemplate keeping a maximum time limit between two unprovoked seizures to qualify for a diagnosis of epilepsy. To avoid confusion, one option is to keep this interval equal to the period after which we call the epilepsy ‘no longer present'. The hypothesis to be considered can be- if seizures recur after decades, it is more likely that the pathology is different from the initial one.

Dr Param S. Kharbanda

30 September, 2013

I wish to appreciate the efforts of the committee on this great job.

The definition is appropriate but it does not involve syndromes.

Theophine Chinwuba Okoye

30 September, 2013

Thank you for your call for comments.

The 2005 definition of Epileptic seizure should insert the word "neurological" before the signs and symptoms as in: Epileptic seizure is a transient occurrence of neurological signs and symptoms due to an abnormal single or synchronous discharge of neuronal activity in the brain.

Epilepsy is a disease or disorder of the brain characterized by an enduring predisposition to generate epileptic seizures with the consequences of neurobiologic, cognitive, psychologic and social problems.

The present operational definition qualifies the definition with suggested critera as follows: (1) 2 unprovoked seizures in a scenario of absent brain insult(time is not important) (2) one unprovoked seizure in a scenario of present or previous brain insult; and (3) unprovoked seizure in reflex epilepsy. Absence of seizures without medications for at least 10 years diminishes the likelihood of epilepsy but may not be compared to the general population without history of epilepsy.

Thank you for your kind attention.

Maria Lina Renales

30 September, 2013

Sorry for the delay in responding. I agree that the definition of epilepsy should be revisired. I agree most with the comments by Dr William Tatum sent in on 25/9/13. Thanks.

Raymond Azman Ali

30 September, 2013

1) I agree with the new statements concern definitions regarding one seizure/two seizures/ten years seizure free without AED
and no risk of epilepsy syndromes; but let me say an additional comment:

2) Abnormal discharges can be attributted to: a) excitatory neurons, or b) inhibitory neurons.

This terms enlight the understanding of epileptic firing itself.

Many thanks to ILAE members for give me the opportunity of take note of operational definitions.

Jorge Ure

30 September, 2013

Abstract: "This revised definition of epilepsy brings the term in concordance with common use by most epileptologists."
This would require adding "adult". Indeed, the whole concept of cognitive or motor defect related to epilepsy has been discarded. This would be OK if there was the addition of a specific concept for epileptic encephalopathy. There is at least one situation in which epilepsy may not comprise seizures: Landau-Kleffner syndrome.

Page 2 last line: the term "seizure" should be replaced by "illness".

"Juvenile myoclonic epilepsy is known to be subject to lifelong elevated risk for seizures": This claim does certainly not apply to all instances. Same for the next sentence: little is know regarding long term course when seizures related to brain dysplasia are under control.

Olivier Dulac

30 September, 2013

Thank you
I agree with these new definitions.

Bao Ung

Return to top

30 September, 2013

I appreciate the efforts of Task Force Team. It is really useful to have new operational definition.

- from initiation of treatment point of view even after one unprovoked seizure in certain cases. We often do it. Now it justifies officially.

- Concept of "No longer seizures" is useful from psychosocial point of view specially marriage and employment point of view.

- Inclusion of reflex epilepsy in definition i.e. "Labeling" may be a stronger incentive for an individual to avoid these stimuli more consciously thus reducing frequency of seizures.

With warm regards
Pravina Shah

30 September, 2013

Thank you  for the opportunity to comment on this important topic.

The relationship between seizures or epilepsy with neurocysticercosis (NCC) is crucial and timely to be analyzed in the context of the proposal "An Operational Clinical Definition of Epilepsy," considering that NC is probably one of the main causes of seizures or epilepsy in poor-resource countries where T/C is endemic and, in turn, most of the population with epilepsy worldwide inhabits precisely in those poor-resource countries. Unfortunately, most of the studies on this topic have been addressed by distinguished and renowned researchers (not necessarily neurologists nor epileptologists), who do not deal patients with epilepsy in their day-to-day clinical practice, which has originated a distortion in the clinical approach of the above-mentioned relationship.

NCC is the best example in the medical practice that acute symptomatic seizures (or provoked seizures) shouldn't be included as epilepsy. In fact, most of patients with NCC usually debuts with seizures due to the brain inflammatory reaction originated by the death of the parasite inside the parenchymal brain. Interestingly, while the inflammatory reaction lasts (around 3-6 months) there is a high risks (40%) of seizure recurrence; however, once the inflammation disappears, the recurrence risk also decreases (less than 30%) (Carpio A, Hauser WA. Prognosis for seizure recurrence in patients with newly diagnosed neurocysticercosis. Neurology 2002;59:1730-1734). After then, the evolution of seizures is benign (there is abundant bibliography supporting this assertion), Therefore, most of the NC patients with an acute symptomatic seizures do not evolve to epilepsy (Carpio A, Fleury A. Hauser WA. Neurocysticercosis: Five New Things. Neurol Clin Pract April 2013 3:118-125)

However, NCC is not simple like that, it is a very heterogeneous and complex disease, since the parasite has many evolutionary phases in the brain. The brain lesions due to a specific evolutionary phase (not the parasitic disease per se) may be considered a "provocative factor"  (brain inflammation around a degenerative parasite) or an "enduring potentially epileptogenic abnormality" (gliosis around a calcified or death parasite). In other words, the difficulty to establish a clear distinction among provoked or unprovoked seizures in patients with NCC also turns problematic the diagnosis of NC as cause of epilepsy.

The following example shows some of the difficulties to diagnose epilepsy in patients with NCC

A patient with a parenchymal degenerative cyst, whose imaging studies show perilesional edema, has a first seizure (provoked or acute symptomatic seizure), which is well controlled with antiseizure medication. Two years later, the patient develops a new seizure (first unprovoked seizure); the imaging studies show a single calcification. Should we wait a second unprovoked seizure to diagnose epilepsy in this patient? Comment: according to the new (but not the old) definition, this patient would have epilepsy.

Kind Regards
Arturo Carpio

30 September, 2013

Congratulations to the Task Force of the International League Against Epilepsy for their good efforts to a dynamic process of epilepsy definition. The classifications are now in need of updating to reflect significant advances made in nearly all fields of epileptology. I would like to comment two main concerns about the presented manuscript:

- I disagree with the term disease for epilepsy. Epilepsy it is a disorder or clinical epileptic syndrome;

- The ILAE Commission could be made update the pediatric epilepsy classification and diagnostic concepts. The Consensus to appropriate diagnosis is crucial given the potential for lifelong consequences to the developing brain.

Finally I agree with the itelligent comment given by Orrin Devinsky, MD and Souhel Najjar, MD (Epilepsy Currents, 2011 ) which is simplified and summarised in one sentence as:

"Change is challenging. Consensus is complex. Classifying seizures and epilepsies only seems to magnify the limits of our science and semantics. The International League Against Epilepsy (ILAE) Commission on Classification and Terminology should be congratulated on their provocative and valuable contribution. Let a new conversation in an endless discussion begin".

Best regard,
Aurea Nogueira de Melo

30 September, 2013

Thank for the opportunity to comment upon the ILAE proposal for new definitions for seizures and epilepsy.
I have read this important paper with interest.

1) "One unprovoked seizure and a probability of further seizures similar to the general recurrence risk after two unprovoked seizures (approximately 75% or more)" and "This operational definition makes no attempt to enumerate the conditions that would increase risk for a second unprovoked seizure above the threshold cited above."

Proposing such a definition leaves it to individual physicians to consider the recurrence risk without a clear evidence base. What work can we currently draw upon to define risk recurrence as greater than 70-75? I think it is premature to introduce such a definition. I think this will introduce more uncertainty because of the lack of this evidence base. All structural lesions confer a >50% risk of recurrent seizures, which ones push it above 75%? Many of the cited pathologies have an evolving epileptogenicity;  stroke, head injury, encephalitis, resected tumour and polymicrogyria (age related changes).

2) I agree the discussion of "provoked and unprovoked seizures" in relation to reflex epilepsy is an improvement upon the previous definition.

3) "Epilepsy has traditionally been referred to as a disorder or a family of disorders, rather than a disease"

I do not think the argument for Epilepsy being considered a disease rather than a disorder is not compelling. "disorder implies a functional disturbance, not necessarily lasting" this is exactly the situation with epilepsy; age related conditions, cure with surgery, changes in the expression and nature of the condition. Using the term disorder honestly recognises the uncertainty around prognosis for the individual patient. Seizures are so often a manifestation of an underlying condition; thus specific aetiology may be tumour, dysplasia, encephalomalacia as a cause of the seizures, and epilepsy exists as a disorder complicating these specific pathologies.

I believe it is likely the public will not welcome the use of the term disease and considering our efforts to humanise and de-stigmatise the condition, I can't see any downside in the use of term disorder.

Thank you,
Andrew Bleasel

Return to top

30 September, 2013

I appreciate the effort to reconcile the 1989 and 2005 definition of epilepsy.

A few issues could be more specifically addressed:

  1. Is status epilepticus considered equivalent to a seizure?
  2. The definition by Hauser (1991 and 1998) referred to generalized tonic-clonic seizures. The current proposal makes no reference to seizure type, but it is unlikely that the risk for recurrence is identical for various seizure types. It might be useful to address and incorporate this issue in the definition.
  3. It would be useful to discuss conditions where a first unprovoked seizure has a > 75% risk for recurrence and those where it does not. The various conditions with sufficient evidence to place them in one or the other category could be listed. E.g. according to recent articles by Josephson, the risk for 2nd seizure in patients with cavernous malformations is >90% whereas in arteriovenous malformations, the risk is < 50%.

Stephan Schuele

30 September, 2013

I congratulate the commission for this important work and I very much appreciated the call for comments to it.

Actually I was quite satisfied with the previous conceptual definition (Fisher et al 2005) and namely with  the concept of "enduring propensity" to generate seizures. How to translate it in the daily practice was up to the physician to decide based on the available information and personal experience. The commission made a considerable effort to overcome the judgment subjectivity specifying some reasonable criteria (two or more closely spaced seizures, estimated high, i.e. >75% recurrence risk). The intention is praiseworthy, but the result seems to me little helpful as the estimation of the risk percentage still leaves room to subjectivity.

Other points.

  • Disease vs disorder: I strongly support disorder in view of the  heterogeneity of the epilepsies.
  • Epilepsy no longer present: this is a merely a descriptive definition, moreover as latin people used to say definition non fit per non. I would prefer epilepsy in remission.
  • Implications for treatment: I found it very appropriate the statement that the diagnosis of epilepsy does not lead automatically to the decision to initiate treatment.

Best wishes
Giuliano Avanzini

30 September, 2013

Thank you to the authors for this work.

I strongly favour "disorder" rather than "disease". Epilepsy is a disorder that can be caused by different diseases.Using the term disease might be confusing.

I see as a difficult challenge to clarify the 75% chance of another seizure. There are criteria that we can get from clinical studies but that are difficult to use in clinical practice. I would agree, though on an effort of finding literature citations that might help in the decisions in different etiological/syndromic settings.

I find the separate category for reflex epilepsy a potentially confusing entity.

It would be sufficient to underline that they are enduring epileptic conditions that can be classified, not to be confused with other conditions such as acute symptomatic seizures.

Maria Paola Canevini

30 September, 2013

I thank the Task Force for the opportunity to comment this document, that surely introduces some advances in the definition of epilepsy such as the concept that at a certain point "epilepsy can be considered no longer present". My comments are:

Disease or disorder: my impression is that in some situation the term "disorder" is more appropriate, for instance in some childhood age-related syndromes, in which we can admit a "functional disturbance, not necessarily lasting", in subjects otherwise neurologically normal and that will not bear long-lasting consequences of the fact that they have suffered from epilepsy in childhood.  Completely different is the situation represented by other epileptic syndromes, such as for instance progressive myoclonus epilepsies, in which the term "disease" fits well since it conveys the notion of a more lasting, even irreversible, derangement of normal function. These distinctions are related also to the recognition that epilepsy is not a single disorder/disease but it covers a large group of conditions that can bear extremely variable prognosis. This latter concept, i.e. several forms of epilepsy with different prognosis,  can be very helpful  when communicating the diagnosis to a patient.

Risk assessment
The 75% recurrence risk is very difficult to apply in everyday practice. Indeed, in clinical practice, the recurrence risk is based on several  clinical data and laboratory findings (such as EEG and MRI) that can allow the epileptologist to establish whether the risk of recurrence is "high". On the other hand, the 75% criteria is based on a single, although prestigious, paper published 15 years ago. If we really want to stick to a precise value, I think that the study should be updated, including a syndromic approach and updated imaging techniques.

Epilepsy no longer present
I agree in general with the definition. My only concern is about those patients that are at least 10 years off medications (therefore their last seizure was several years before) that may have an associated structural non-evolutive lesion, and because of this they cannot be told that their epilepsy is no longer present. Should these patients really be excluded from the "no longer present" category ? Indeed,  I find somewhat difficult to tell them that their epilepsy is still present.

My best regards and many thanks again to the Task Force for their efforts
Guido Rubboli

Return to top

30 September, 2013

I would like to congratulate for the great work of the ILAE task force: an operational definition is needed and it is going to be very useful in clinical practice. Here are some my comments:

1. I believe, that Epilepsy should remain a disorder rather than a disease. The word "disease" has more serious social impact than the word "disorder". "Disorder" allows more heterogeneity than the "disease" and is more applicable to epilepsy.

2. I would like to point out that the introduction of a "75 %" seizure reccurence probability as a necessary precondition for calling a state epilepsy lead to practical problems for practical doctors and is not helpful. The term "at least 75 % risk is very hard to explain to patients, students and practicians.

3. I find the term "remission" more usable than "no longer present". the purpose of defining "no longer present" may not be clear enough.

Thanks again for the opportunity to comment on this paper.
Best regards,
Ruta Mameniskiene

30 September, 2013

I appreciate the opportunity to comment on this statement, and I apologize for the delayed response . Here are my comments:

  1. I think that it is not important to introduce a untrue change about the term disease. Within epilepsy, which must still be considered a "disorder" of the CNS, are comprised a spectrum of different conditions, from occasional/symptomatic seizures to well defined epilepsies and epileptic syndromes. So epilepsy is a broad disorder of CNS and not a single disease.
  2. Symptomatic and reactive seizures do exist according to Beghi et al (please check the reference Beghi et al 2002 in references) and they do not define an epilepsy, considered as a chronic condition. On the other side, one isolated seizure within a well defined epileptic syndrome (Panayiotopoulos and focal idiopathic syndromes) is enough to define a condition of active epilepsy. It is difficult however on many occasions to define the risk of recurrence > or < 70%.
  3. From a practical point of view, the opportunity to define an epileptic event as occasional seizure or epilepsy is unimportant because it is true that some symptomatic seizures have to be treated for a long period, while some well defined epilepsies can not be treated.
  4. In reflex epilepsies there is an enduring abnormal predisposition to have seizures, so reflex epilepsies have to be considered true epilepsies.
  5. I totally agree both with the statement of "no longer present" to define a situation cured, but that it does not guarantee that it will not return, and with the statement that a "a probable epilepsy" is a confounding definition.

Best regards,
Massimo Mastrangelo

30 September, 2013

Thank you once more for the opportunity to comment on the proposal. I hope you don't mind me adding some additional thoughts. The most problematic part of the proposal from my perspective is the change in the definition to include on seizure plus a 70% risk of recurrence.

It is important to emphasise that the person carrying a diagnosis needs to know whether the diagnosis is active at a specific point in time, rather than based on a probability of seizure recurrence following a historical seizure.

So, one further thing that the committee needs to consider is that risk of recurrence is conditional upon time since the first seizure. For example, an individual may be at high risk of a recurrence immediately after a seizures, but within say after a few months of seizure freedom the risk will drop below 70%. Are we to say that this individual moved from a state of epilepsy to a state of not having epilepsy?

While I understand the desire to consider the whole epileptogenic process and broaden the definition of epilepsy, I do not think that we are ready or sophisticated enough to define and put into operation a broader definition that will be to the benefit of patients.

Best wishes
Tony Marson

30 September, 2013

Congratulations to the ILAE Task Force for all efforts done to improve the definition of epilepsy, and thank you for the opportunity to comment it. In my opinion, the authors should better justify their recommendation to use the term "disease" instead of "disorder" or "medical condition", clarifying the concepts. Considering that the terms "disease" and "disorder" are conceptually close, and that a "medical condition" is a more value-neutral and broad term that includes both, in this case I would suggest to use "condition".

Cristina R. Reschke Bandero

30 September, 2013

Congratulations to the Task force for this work.

The operational definition of epilepsy is practical and useful for clinical diagnosis.

Epilepsy can be considered alternatively as a disease or disorder but I used to prefer the term disease.

Thank a lot for the opportunity to comment on this document
Thuy Nguyen Thanh

Return to top

30 September, 2013

We all are aware that the term "epilepsy" can be open to an endless list of definitions on the basis of different points of view, different contexts in which it is used, different gradations of value given to the various clinical manifestations/implications of the pathology. Our present effort is to find a definition of epilepsy as a disease which is the most largely shared and the most easily understood in a clinical setting.

My main considerations are the following:

1. DISEASE: Epilepsy is a brain disease of acquired, genetic or unidentified aetiology, characterized by recurrent epileptic seizures as the by far predominant clinical manifestation. The risk after a first seizure of chronic recurrence of seizures is high (>75%) in the following circumstances: a. two unprovoked seizures more than 24 hours apart, b. at least two seizures in a setting of reflex epilepsy, c. others. Like any other disease, epilepsy can experience worsening, improvement, chronicity or remission.

Examples of epilepsy as a brain disease are: 1. juvenile myoclonic epilepsy (genetic epilepsy) ; 2. epilepsy secondary to a dysplasia (symptomatic epilepsy); 3) types of epilepsy of which a genetic, metabolic, structural, etc. disorder has not been identified (epilepsy of unidentified aetiology).

Two points are open to discussion:

a) the term "recurrent": in my opinion, the term could be intended as a purely clinical parameter and, as such, it can be left with the evaluation of the single individual clinician in relation to the single individual patient; of course, clinicians must be confident with those circumstances in which the risk of seizure recurrence becomes >75%; however, the concept of "risk" has the intrinsic mean of relativity and, as it happens in medicine with many other diseases, the diagnosis of epilepsy as disease is a longitudinal one; as such, "recurrence" should be evaluated clinically by taking into consideration all the clinical features of a given patient and the features of seizure presentation.

b) the term "the by far predominant clinical manifestation": when we refer to epilepsy as a disease, this has to be characterized almost exclusively by seizures; the terms "improvement", "worsening", etc. should be referred exclusively to epileptic seizures and not to other possible additional manifestations and/ or to the underlying pathology; the following examples could better explain my thought:

  1. a patient with a calcified meningioma or a dysplasia and exhibiting epileptic seizures clearly suffers from a "symptomatic epilepsy"; the frequency of seizures (which are the main clinical manifestation) could change independently of the structural lesion (which remains unmodified);
  2. if in a given patient the clinical picture is made by recurrent seizures, severe mental retardation, severe motor disorders and other signs, "epilepsy" would be a reductive, and in some way misleading, diagnosis as compared to the complexity of the complete clinical picture; in this case, the patient suffers most probably from an "epileptic/epileptogenic encephalopathy".

In both examples, the patient suffers from epileptic seizures, but only in the first case they are the only (or the "by far predominant") manifestation; in this case, therefore, recurrent seizures gain the clinical dignity of disease.

Apart from the above considerations, the terms "epileptic syndrome" (which, in my opinion, should be referred to the whole of the features characterizing only seizures and not other possibly associated signs/symptoms) and "epileptic/epileptogenic encephalopathy" deserve thoughtful reflections. These topics, however, require a prolonged reasoning and I will try to express my thoughts on future occasions.

Generally speaking, I think that, in the forthcoming definition of epilepsy, inclusion of terms like "risk" or "enduring predisposition" should be avoided: these kinds of terms, in fact, although easily understood by insight, cannot be easily defined despite all the efforts made. Additionally, in a clinical setting, a given clinical picture could be of a very hard task to be labeled with a precise diagnosis definition as it can be the result of combined and difficult-to-be-graded factors. A recent experimental study well explains the concept: "Seizure predisposition after perinatal hypoxia: Effects of subsequent age and of an epilepsy predisposing gene mutation" by Soren Leonard et al (2013).

My best greetings to all the Colleagues of the Task Force and many thanks for your extraordinary work.

Franco Pisani

30 September, 2013

In the definition of epilepsy, we just have small wonder concerning the phrase "Epilepsy is considered to be no longer present for individuals who had an age-dependent epilepsy syndrome but are now past the applicable age". For example: A child suffering Lennox Gastaut Syndrome, When she (or he) grow up and become adult but seizures keep occur. In that case, can we consider that patient is not epileptic or epilepsy is no longer present? the same wonder is for Benign Childhood Epilepsy with Centrotemporal Spikes if the seizures continue when the child pass the teenage.

We agree with all the other points in the definition.

Best Regards,
Tran Viet Luc

30 September, 2013

I agree with the proposed definition.

I thank you for involving me in this feedback program.

With best regards,
Ivanda S. Silva Tudesco

30 September, 2013

The Indian Epilepsy Society endorses the definition of seizure and Epilepsy drafted by the Task Force.

Best wishes

Sincerely yours,
Dr. Man Mohan Mehndiratta

30 September, 2013

Congratulations to the ILAE for the effort for producing this excellent and useful paper opened to colleagues for discussion. Thank you very much for offering me the opportunity to comment about this so important issue. Here are my comments:

As a basic researcher I find the term "disorder" more appropriate than "disease" when we talk about epilepsy. Besides, I would favour the term antiepileptic drug versus anti-seizure drug.

I also think it is welcome the reference to "reflex epilepsy" in the definition.

Best regards,
Clarissa Fantin Cavarsan

Return to top

30 September, 2013

I have a specific concern with regards to co-morbidities of epilepsy. The most recent IOM report on epilepsy outlined the importance of focusing on epilepsy comorbidities, such as cognitive function. Dr. Fisher's 2005 definition of epilepsy encompassed the comorbidities by stating that epilepsy not only is seizures but also cognitive, psycho-social and other comorbid problems. I think the most current definition is a departure from this and towards a seizure-centered definition of epilepsy. This needs to be addressed.

Kris Bujarski

30 September, 2013

I agree on the comments from Finland of Silallampe of the 17. September.

Peter Uldall

30 September, 2013

In my opinion the new operational clinical definition of epilepsy brings many aspects into light, although the developing of new investigations will be very important for the sake of clarity of many aspects.

For instance, the document says: "a seizure that is provoked by a transient factor acting on an otherwise normal brain to temporarily lower the seizure threshold does not count towards a diagnosis of epilepsy. The term "provoked seizure" can be considered as being synonymous with a "reactive seizure" or an "acute symptomatic seizure" (Beghi et al. 2010)". However, physician could be assisting a person with the first epileptic seizure he/she will have in life (e.g. generalized tonic-clonic seizure in juvenile mioclonic epilepsy), which has been precipitated by some factor that lowers seizure threshold (e.g. alcohol intake). Alas, in those patients with provoked seizures, doctors should always wonder if they really have "reactive seizures" or actual epileptic ones, because "identification of a provocative factor does not necessarily contradict the presence of an enduring epileptogenic abnormality".

In other part of the document we may read: "if a treating physician determines that the symptomatic lesion has generated an enduring predisposition for unprovoked seizures with a risk comparable to those who have had two unprovoked seizures (which we all agree is epilepsy), then that person too should be considered to have epilepsy". But in this case physician needs the development of investigations, which ascertain the real risk of seizure recurrence of every epileptogenic lesion, to satisfy a specific threshold risk number (70-75%); otherwise epilepsy could not be diagnosed. Furthermore, the presence of a single seizure plus a cerebral lesion is not enough to diagnose epilepsy, since all brain lesions are not epileptogenic. Besides, the physician should carefully interview patient and seizure's witnesses in order to know if cortical areas activated during the episode correspond to that where the structural brain lesion is.

Authors also explained that: "among children with a complex febrile seizure (CFS), the risk for a subsequent unprovoked seizure is less than the 75% risk needed for epilepsy diagnosis therefore; the condition of having CFSs is not defined as epilepsy, although it may presage epilepsy". Nevertheless, in patients with family history of febrile seizures in infancy or with family history of epilepsy which suggest the diagnosis in relatives of GEFS(+), the likelihood of them also having this type of epilepsy is extremely high, even in those with simple febrile seizures, and it should prompt the diagnosis of epilepsy (GEFS(+)), although the physician do not prescribe for them any antiepileptic treatment at that moment.

The inclusion in the definition of recurrent reflex seizures as epilepsy is very reasonable since, as the authors explains, "tendency to respond to such stimuli with seizures meets the conceptual definition of epilepsy, in that reflex epilepsies are associated with an enduring abnormal predisposition to have such seizures". Moreover, reflex epilepsies are included in the ILAE classification of epileptic syndromes and epilepsies since 2001.

Adriana Goicoechea Astencio

30 September, 2013

Many thanks for giving us the opportunity to comment on this important topic re clarification of some of the definitions of convulsions and epilepsy. Here below we listed our comments re some of the issues detailed in the document:

1. Disorder vs. Disease
We think that the term "disorder" could better define a heterogeneous and complex condition such as epilepsy. The derangement of normal function (in this specific case the electrical function of the brain) as well as the predisposition to seize, are not always permanent (nor enduring) and not necessarily long lasting (as adroitly mentioned in the definition document). If we think of the paediatric and adolescent ages (i.e., a relevant part of the epileptic syndromes) many forms of epilepsy are strictly confined to that period being related to the presence of (altered) specific brain receptors and/or molecular structures, which changes over time (i.e., the predisposition in these cases could - and often is - no longer enduring). In addition, at this young ages, the families involved (either as parents or as affected family members in their youth), could find more reassuring (without minimizing the serious nature of epilepsy), to have their children or themselves not to be labelled as having a disease. We also think that the current public opinion is more cultured and should understand, nowadays, the term "disorder" and the serious nature of epilepsy.

2. Two unprovoked seizures [& High recurrence risk]
One unprovoked seizure and a probability of further seizures (approximately 75% or more) to define as epileptic a patient could be difficult to apply to all epileptic patients and especially to children. As adroitly said in the paragraph on "high recurrence risk" a child with one unprovoked epileptic seizure and a focal cortical dysplasia could more easily diagnosed as having epilepsy (see the definition "epilepsy no longer present"). However, that does not hold true for many remaining categories. At the young ages, is particularly difficult to distinguish each component of the risk factors. Thus, we would prefer to restrict the overall definition of "epilepsy" to patients who have had two unprovoked seizures occurring more than 24 hours apart. The definition mentioned in the last lines of the paragraph entitled "high recurrence risk" fits well with what we would mean: i.e., some conditions (e.g., focal cortical dysplasia or CFS) may presage epilepsy. That holds true also for implications for treatment.

3. Reflex epilepsy
The definition of "reflex epilepsy" is that of an epileptic crisis, which recognizes a trigger factor but it is included in the group of the epilepsy. So we think it should have the same definition as that re provoked seizures: i.e., "At least two reflexes seizures in a setting of reflex epilepsy occurring more than 24 hours apart".

4. Epilepsy no longer present
The term "Epilepsy no longer present" seems more appropriate even though it is difficult to foresee the natural history of crises and when they will stop definitely.

5. Length of seizure free interval
The duration of a "seizure free interval" in children is dependent by the etiologic causes. In general when no lesional (e.g., anatomic, histological, receptors, proteins, etc.) causes are present in a child, the duration of free interval should be restricted to three years and no longer, without forgetting the side effects that the long lasting therapies with anticonvulsant drugs may cause in patient who could no longer have seizures.

Lorenzo Pavone & Martino Ruggieri

Return to top

30 September, 2013

The analysis we undertook for the justification of both definitions of epilepsy – the definition of 1989 and the definition of 2005 – reflects two different methodological approaches to the understanding of epilepsy: the clinical/social approach and the biological/essential one, respectively. The former is relevant for clinical doctors, as it determines the diagnostic and therapeutic algorithm. The latter is relevant for scientists, as it encourages further investigation of the nature of epilepsy (i.e., its origins and pathogenesis).

One of the most debatable of the ILAE's proposals suggests that the definition (diagnosis?) of epilepsy be applied to the situation of a single unprovoked seizure. We believe that the qualitative estimate of risk factors (< and> 75%) is not accurate, as there are no representative studies that support this indexes . Quantitative criteria that take into account such basic risk factors as genetic burdening, perinatal brain injury, cerebral cortical dysplasia, craniocerebral trauma, cerebral insult or epileptiform activity in EEG may provide a more accurate assessment of risk. The presence of the aforementioned risk factors in addition to a single episode of an unprovoked seizure plays a decisive role in the diagnostic process. In the event where only one risk factor is present, one can suggest possible epilepsy; with two risk factors – probable epilepsy; and with three or more risk factors – real epilepsy. For another version of diagnostic formulation in the cases of patients with a single unprovoked seizure, it possible to use "Risk factors of epilepsy" with the list of risk factors. We believe the problem of criptogenic epilepsy (CE) cannot be resolved with the new methods of CNS investigation (including future methods). According to our various clinical studies (1990 and 2010), the CE by genetic burdening holds consistently a place between idiopathic and symptomatic epilepsy (11% vs. 18% and 3.3-8.4%). Thus, CE represents an epileptic illness with the soft but more or less equivalent presence of both basic epileptogenic factors – genetic and exogenous.

We believe that there is no accurate distinction between the terms of "epilepsy" and "epileptic illness"(ED). Epileptic illness appears to be a wider phenomenon that embraces both epilepsy and epileptic encephalopathy regardless of whether or not the latter is accompanied by epileptic seizures (seizures type 1 and 11, as we argued before, 2010).

Lastly, there are certain forms of epilepsy (or certain types of seizures) where seizures are not a direct response to physical and biological stimuli ("mathematical epilepsy", "game epilepsy" etc). Is it correct to apply to them such terms as "reflex seizures" or "reflex epilepsy"? Perhaps the term "stimulus dependent" provides a more adequate description?

Vladimir Karlov

Karlov V.A. Epilepsy Moscow: Medizina 1990, 333 P(rus)
Karlov V.A. Epilepsy of Children and Adult Women and Men 2010: Medizina, 718 P(rus).

30 September, 2013

Thank you for sharing this important document and for the opportunity to comment it.

a. I just wonder if the task force shared its work with the WHO classification group so that the ICD 11 revision team and the ILAE can try to join forces. Here below the ICD 10, that for many countries remains the standard. In ICD 10 Epilepsy is a disorder, within the diseases of the nervous systems. The definition of epilepsy as disease might be complex where there is a DRG system and for example epilepsy follows stroke as a consequence. Would you put stroke first and epilepsy as second? Than is secondary epilepsy, and is an episodic disorder.

Should the definition of disease be approved by ILAE than a strong work should be done at WHO as well as all DRG levels to be sure that this concept is clear, shared and taken.

b. What concerns me most is the " no longer present" definition. Is ambiguous, although calls for more clarity and might have serious legal consequences in terms of invalidity pension eligibility in several countries for example. It is equal to cured than, use cured, that means finished,  terminated, problem is closed forever. No more disability benefits, as after a pneumonia!.  If  "no longer present" means absent, means that the patient is out of danger of a recurrence, can do everything, back to work, driving and in some countries can vote and can marry that he deserves to be defined "cured". Can the ILAE experts say this?

c. Can 10 years be shortened to 5 ? In any case the definition still provides some of the ambiguity it wants to avoid, so why keep so long? Given the legal implications bounded to the diagnosis,  a shorten period could be given so that issues such as return to work, driving or other similar can provided. But here could be a schizophrenia with the patients and family associations. Stigma is not disappearing with the disease, even when the disease disappears. "Cured" is stronger and as unreliable as "no longer present", and could protect patients more from discrimination as it can give a sort of security to patients as well as to his/her environment.
Is more easily sellable in certain environments where our patients have to live.

Thank you. Matilde Leonardi

30 September, 2013

Comments from Rupprecht Thorbecke

30 September, 2013

Thank you very much for your work "An operational definition of epilepsy"! It is a very valuable trial to give "a definition of epilepsy", to define the signs and symptoms giving a  certainty for a diagnosis of an epilepsy, to start (or not to start) a treatment and to make a decision on the ending of the treatment. There will be a lot of discussion about your paper because the way up to the decisions will be complicated in every individual case. At any rate, the colleagues are animated to reflect the clinical points  in greater detail before they make a decision "epilepsy yes or no" or "treatment yes or no".  Some cases may be so complex that would be is necessary to wait the ensuing course with repeated consultations until it is possible to make a diagnosis. This concerns especially young children.

On the primary level care it would be difficult to use the new operational definition of epilepsy as well as to consider starting (or not) a therapy. One of the difficult points, even for a specialist, is to find out "the risk of further seizures".

To:  "Disorder": when the patient has an epilepsy AND concomitant disturbances (reduced intellectual level, reduced memory, psychological problems...) does he have a disorder or a disease called "epilepsy"? or several different disorders / diseases?

To: "How long to treat a patient with epilepsy?" There are several types of epilepsies with myoclonic jerks and atypical absences which are difficult to treat and usually need a lifelong treatment. On the other side, it is important to find some recommendations concerning a termination of the medical treatment in seizure-free patients.

To: "Reflex epilepsy": I think the definition you made about the reflex epilepsy is appropriate.

Looking forward waiting results of  future discussions and a success of the application of "An operational definition of epilepsy", with kind regards

Ritva A. Sälke-Kellermann

Return to top

30 September, 2013

Please find my comments below for the ILAE website:

1) Epilepsy is indeed a disease rather than a disorder. But the definition should be reviewed again to exactly emphasize that "not all epilepsies are the same" for patient and public education. "The frequency and types of the seizures as well as severity of this disease change widely between the individuals and may subject to change even in the same individual."

2) It should also be emphasized that family history of epilepsy and genetic variations are important to define the recurrence risk to promote the involvement in genetic studies.

3) "At least 75% risk" is very hard to explain to the patients. Could we say only "high risk".

4) I prefer as an aprropriate medical term "remission" instead of "no longer present".

5) I also think like many others that the following phrase should be omitted or rephrased "Juvenile myoclonic epilepsy is known to be subject to lifelong elevated risk for seizures" in the light of current data.

How many patients with JME have you seen after the age of 40-50 years ?

- Baykan B, Martínez-Juárez IE, Altindag EA, Camfield CS, Camfield PR. Lifetime prognosis of juvenile myoclonic epilepsy. Epilepsy Behav. 2013 Jul;28 Suppl 1:S18-24. doi: 10.1016/j.yebeh.2012.06.036

Many thanks for your huge efforts and sharing the work with us in the right time before publishing!
All the best,
Betul Baykan

30 September, 2013

We very much welcome the effort of the ILAE to stimulate thinking on new definitions on epilepsy. The new definition requiring only one unprovoked seizure plus a prevailing tendency to have further seizures is useful from a practical standpoint. However, we foresee that the main problem will occur in low and middle income countries (LMIC) where there is a paucity of facilities for investigations. Not only will the underlying tendency to recurrence not be identified, but in addition acute symptomatic seizures such as those due to neurocysticercosis may be inappropriately included. This definition will lead to an increase the number of patients suffering from epilepsy which in turn will have an impact on the already high epilepsy treatment gap (ETG)in such areas. With this new definition the ETG will certainly increase as the denominator will be increasing.

Numerous community based studies have been carried out on the prevalence, incidence, outcome and aetiology of epilepsy in LMIC with lack of uniform methodology. Such studies make it difficult to obtain comparable data. Any change in definition will render it even more difficult to compare them at an international level. As a consequence it will be much more challenging to bring in international donors and stakeholders to support a national program since ambiguity of the data will increase.

Finally, the new definition also foresees a new classification of epilepsy by aetiology. We now have presumed genetic, structural/metabolic and unknown categories. Without more sophisticated diagnostic means, it is virtually impossible to classify epilepsy into these categories. Hence, yet again, people in LMIC will be affected by this change as their health status will not be adequately represented. Once more, it is not the fact that the new definition is not better at capturing the true nature of the condition, but rather the fact that people with epilepsy in LMIC do not have any comparable support in obtaining their diagnosis, and therefore are left out in international comparisons. As a consequence, and most probably, the quantity of resources available for their support and management will diminish.

We feel that it is crucial to keep in mind the living conditions of people with epilepsy in LMIC. Inevitably, such a change of definition of epilepsy as is proposed will lessen the chance of making a valid diagnosis and hence entitlement to treatment. Thus, there will follow an increase in treatment gap.

Nadir E. Bharucha and Peter Odermatt

30 September, 2013

Thank you for the opportunity to comment on the proposed new definition by the ILAE Task Force.

Change of the term 'disorder' to 'disease' is not suitable as it will heighten stigma. For example, to date in many developing countries 'disease' is speculated to be 'communicable' and can result in unwarranted social ostracism. 

Re: Point 2 of the suggested definition. How does one quantify the high probability (>75%) of further seizures? This would be a very arbitrary guess.  

Zarine Mogal

Return to top

30 September, 2013

Thanks for the opportunity to participate in this discussion.

I would like to comment on reflex epilepsy:

I believe that it is difficult to understand the inclusion of reflex seizures for people who are not experts in epilepsy once epilepsy definition tells about "unprovoked" seizures. I think it is worth differentiating the two situations:

1) the stimulus is not capable of generating seizures in a normal brain but it is in a hyperexcitable one (eg, flashing lights, hot water, skin friction, etc.),

2) the stimulus is capable of generating seizures in a normal brain (eg, uremia, hyponatremia).

In the first situation the abnormality is in the brain (primary hyperexcitability), so it is epilepsy, while in the second situation the brain is normal and the seizure is a response to a disruption in homeostasis, an insult, so it is not epilepsy but acute symptomatic seizure.

Mirian Guaranha

30 September, 2013

We really appreciate the effort of the ILAE task force: an operational definition was needed and it is going to be very useful in clinical practice.

Here are some comments:

- Genetically determined epilepsy is a condition lasting for life and also age- dependent epilepsies such as BECTS can last for many years. Consequently, should we believe that the child is carrier of a disease for 6 or 8 or 10 years? In cases of genetically determined epilepsies, we believe that it has to be considered as a condition characterizing the subject which may or not protract. With that and since the goal of the work is to provide an operational definition, we believe that the term "disorder" is preferable and more pragmatic since it refers to a more extensive dysfunction including neuropsychological and social deficits. We agree with collegues affirming that "disorder" includes the multifaceted aspects of epilepsy and different clinical situations with age-dependent evolutions. The use of the term "disorder" may also be more acceptable by epileptic patients and their families. It is, also, important to consider that epilepsy per se is associated with a social stigma, since it traditionally recalls fatal situations. Epileptic patients are scared by the disgraceful sensation following for example a seizure with fall or loss of consciousness, especially if occurring in social contexts. On the other hand, professionals working with parents and epileptic patients in clinical settings have the opportunity to support them using a sincere and evidence – based approach in explaining what epilepsy is and implies.

- Remission can be considered as seizure control since 5 years without anti- seizure medication. 10 years may be too much time in countries like Italy with strict forensic norms. Since nobody can predict if there will be a relapse after 5 or 10 or 15 years, from an operational point of view it could be more advantageous a limit of 5 years.

- We totally agree with what claimed by Umberto Aguglia and Edoardo Ferlazzo :
"Since it is well known that a considerable percentage of patients experiencing the first unprovoked seizure may be diagnosed as having a well defined epileptic syndrome (King et al., 1998; Jallon et al. 2001; Olafsson et al. (2005), we propose to change the second point of the operational clinical definition in table 2 as follows: "One unprovoked seizure in the context of a well-defined epileptic syndrome (or condition) (i.e.: JME, symptomatic epilepsy in patients with a static or progressive CNS disease and so on)".

Melissa Filippini and Giuseppe Gobbi

30 September, 2013

1. I liked the concept of defining Epilepsy as a disease.

2. I found it very confusing particularly for physicians outside the field o neurology to decide when to treat or not an unprovoked seizure based on this study. Therefore, I do not think it reached its main goal.

I myself have applied the concept of treating the first unprovoked seizure when an epileptogenic lesion is found in the brain MRI and/or when the EEG suggests epileptogenicity. I believe many neurologists, and particularly epileptologists, are already doing so as well. However, we still do not have this concept clearly stated. So, perhaps it would be adequate to attempt to clarify it at this time and not later on as the authors suggested. What about to expand the concepts only vaguely described here?

For example: Highly epileptogenic lesions on the brain MRI include but are not limited to: cortical dysplasias, ...and so on ... Then, do the same thing to the EEG findings, and Epilepsy Syndromes...

Best regards,
Diosely C Silveira

Return to top

30 September, 2013

Good work addressing a very practical issue. Even better is the opportunity to comment!

My main suggestion is to remove the requirement for seizures to be unprovoked. The 2005 conceptual definition stated: "The definition does not include a requirement that the seizure be "unprovoked," a feature of several prior individual definitions." Why has it now been reinstated? The current paper discusses provoked seizures but not recurrent provoked seizures and offers no clues as to why the change was made from 2005. Was it pressure from epidemiologists? I assume the intention is to avoid labelling those with acute symptomatic seizures more than 24 hours apart but within the same transient illness as having epilepsy, with the implication that long-term AED therapy is indicated, when it may not be. The problem is that some patients with undoubted epilepsy have only provoked seizures – for example, some patients with JME who have seizures only when sleep deprived (and declaring that sleep deprivation is somehow not a provoking factor does not make sense). One of the greatest gaps in our understanding of epilepsy is what makes a seizure occur on one day and not another – i.e. what provokes it? Most seizures are unprovoked as best we can tell but that is just a reflection of our ignorance. In fact, all seizures must be provoked by something, whether exogenous or endogenous.

The essence of the concept of epilepsy is that without intervention, there is a likelihood of further seizures. It doesn't matter to the patient whether we can identify a provocation or not (unless it is an avoidable provocation). Acute symptomatic seizures often recur and when they do, they have the same consequences for the patient as recurrent unprovoked seizures. Indeed, in our (Australian) latest driving regulations, recurrent acute symptomatic seizures are handled the same as recurrent unprovoked seizures.

The assumption that patients with acute symptomatic seizures are just demonstrating the response of a normal brain to a provocation cannot be true – if it were, why is it that the vast majority of patients experiencing hypoglycaemia for example, do not have seizures and why is it that those who do, tend to do so recurrently? As an illustration, I have agonised over this with several patients with diabetes in whom I could not decide whether they had epilepsy or just hypoglycaemic seizures until it became clear that this was a waste of time - they had both – they needed both AEDs avoidance of hypoglycaemia and to avoid seizures. Yet they had seizures only when hypoglycaemic. By the proposed definition, these patients do not have epilepsy. I (and they) would argue that they do! What matters is whether, without intervention, seizures will recur. If a provocation that triggered a seizure is likely to recur, then seizures are likely to recur and the patient therefore has epilepsy. This agrees with the Task Force's statement that "Epilepsy exists in a patient whose brain, for whatever reason, demonstrates a pathological and enduring tendency to have recurrent seizures." Leung (2009) found epileptiform EEGs in 15/105 patients with acute symptomatic seizures and it conferred an odds ratio of further acute symptomatic seizures of 16.

A single unprovoked seizure does not earn a diagnosis of epilepsy because the probability of further seizures is not high enough. Similarly, a single provoked seizure should not earn a diagnosis of epilepsy because the probability of further seizures is not high enough (though elevated at 32% risk of further acute symptomatic seizures and 12% risk of unprovoked seizures at 2 years in Leung's 2009 study). The seizure risk after 2 or more distinct episodes of acute symptomatic seizures is unknown but likely to be high. Other examples of recurrent provoking factors include hyponatraemia in patients with compulsive water drinking, drug intoxications and withdrawal, other metabolic derangements and mitochondrial disease.

The exclusion of provoked seizures from the definition can lead to inappropriate management, as illustrated by a patient with alcoholism, who repeatedly presented to the emergency department with seizures associated with alcohol excess or withdrawal. He was not investigated with EEG or offered AED therapy because he was not considered to have epilepsy. After a seizure in which he suffered a spinal cord injury, an EEG was done and showed generalised epileptiform activity. He was then started on AED therapy and has had no further seizures except when noncompliant. The point of this anecdote is that provoked seizures and epilepsy may coexist and interact and that specifically excluding provoked seizures from the definition of epilepsy may have adverse clinical consequences.

A problem created by removing the "unprovoked" requirement would be that patients may experience more than one acute symptomatic seizure more than 24 hours apart but within the same acute transient insult. To avoid such patients being inappropriately labelled as having epilepsy if this is the first time the insult has provoked seizures, there would need to be a separate criterion such as "recurrent provoked seizures more than one week apart" or "seizures provoked by a factor that has occurred on more than one occasion". Whether this would meet the 70-75% probability criterion is unknown but that figure is arbitrary anyhow.

Indeed, the inclusion of a separate diagnostic criterion involving reflex epilepsy in the definition document acknowledges the same arguments I am trying to make for provoked seizures. From the patient's point of view and in respect of therapy, the situation is the same as recurrent provoked seizures – i.e. something has triggered a seizure more than once and will do so again if that something recurs – whether it be reading in someone with primary reading epilepsy or hypoglycaemia in someone with diabetes. I am not of course suggesting that reflex epilepsy and provoked seizures are the same but there are compelling parallels in the reasons one is labelled epilepsy and the other should also be.

There may be sound epidemiological reasons for separating acute symptomatic seizures (as defined by ILAE in 2010) from other provoked seizures and from unprovoked seizures but these need to be weighed against the consequences for patients and their clinical management.

While I think I understand the distinction between "operational" and "conceptual", there is clearly much overlap and when this is finally published, the differences should be discussed and those terms also defined.

Re: "Probable epilepsy": Telling a patient they have epilepsy is saying "It is probable you will have more seizures." Probable epilepsy is saying "It is probable that it is probable that you will have more seizures"!

Re: the 2005 conceptual definition, I do not see why even a single seizure is required. But that's another story and I'm grateful you have read this far!!

Ernie Somerville

30 September, 2013

Thank you for the opportunity to comment on the work of the task force.

The proposed text is good progress from earlier definition. There are a few issues both raised in the document and in some comments that I think reflect uncertainty of the scope or framework of the definition. E.g. epilepsy definition in relation to a drivers license is very different from ILAE proposed definition. Treatment or reimbursement procedures do not need this definition. However, defining the field, both from patient point of view, from the professionals and from regulatory authorities is important.

Adding some constrains to the field of validity of the definition would help to come around some of the issues discussed.

Pål G. Larsson

Return to top

30 September, 2013

I am grateful to have had the possibility to read the document. It is much work done this document, and I find that it will be useful. I congratulate to a good work.

Best Regards
Bertil Rydenhag

30 September, 2013

The German society for epileptology appreciats the new definition of Epilepsy but we want to give some comments on this topic:

  1. When data are insufficient to demonstrate more than a 75% chance for another seizure, then epilepsy is not considered present until there is a second unprovoked seizure. 75% is very difficult to calculate and we are afraid, that the general neurologist will define an epilepsy in this case after the second seizure. We would be very glad, if instead of 75% the definition would give the expression "high risk" and that could be more than 50% risk of a second seizure.
  2. The time of remission to be cured from epilepsy is very long with 10 years. In Germany we have different times after which epilepsy is cured: 3 years remission in the social security statutes and 5 years in the traffic regulations. Maybe that there are good argument for 10 years remission but in general it is a burden for patients . Children with Rolandic epilepsy with a last seizure at the age of 15 will get the possibility to drive a bus with 25 years or later. Exceptions are possible but the law giver does not know anything about the special conditions of genetic epilepsies. For German legislature the time of 10 years seizure freedom are 10 years.
  3. The definition of reflex epilepsies is very helpful, but why 2 reflex seizures ? In the case of real reflex epilepsy like photosensitive epilepsies or primary reading epilepsies, the first reflex seizure will give you all the information you need. Maybe it could be helpful to speak from reflex epilepsy after the first  reflex-induced generalized tonic-clonic seizure.

Many thanks to Robert Fisher for his work about the new definition.
Thomas Mayer for the German Society for Epileptology

30 September, 2013

We would like to sincerely thank the Task Force for their efforts and to express our concurrence with their findings. Below please find our comments in response to the following paragraph:

Unprovoked seizures separated in time:
The time span between two unprovoked seizures that together qualify as epilepsy is subject to ambiguity. Seizures clustering within 24 hours confer approximately the same risk for later seizures as does a single seizure (Neligan et al., 2012). The Task Force retained the current thinking that unprovoked seizures clustering in a 24 hour period be considered to be a single unprovoked seizure for purposes of predicting recurrence risk.

It is our opinion that in reference to an unprovoked seizure or a cluster, its nature has not been specified, i.e. whether it is a self-limiting seizure or a status epilepticus. We also believe that in determining recurrence risk, the roles of a focal seizure cluster versus a series of generalized convulsive seizures distinctly vary.

Best regards, Josif Volkov & Oxana Volkova

29 September, 2013

Having this important document to be opened for comments should mark a milestone for the field of Epilepsy.

As for most diseases or disorders, the published document will have major long lasting impact for the resourceful-poor countries or medical communities (especially for the medical students, trainees, primary care physicians or paediatricians) who might not be working in pace with the "Epileptologists".

A working definition should be concise and easy to apply at a clinic setting when we are confronted with our patients. The definition should automatically carry the implications to treatment and prognosis in a clinic setting as well. Thus, our diagnosis for Epilepsy should be based on what information we had collected and what further tools are available to confirm our clinical suspicion.
Who can make the definitive diagnosis of epilepsy? —

The primary care physicians? The Paediatricians? The Paediatric Neurologists or Neurologists? The Psychiatrists? Or one needs the Epileptologist to make the final verdict?

As EEG is the only diagnostic tool we have, though not 100% sensitive or specific, one should include this in the operational definition, although acknowledging its limitations.

This is analogous for other diseases eg. for diagnosing rheumatic fever, we need ESR, ECG etc; for diagnosing ADEM or MS, we need MRI as supporting tools to add on our diagnostic sensitivity.

I still think that the concept of "Definitive----Probable-----Possible----- Definitely-Not" Model should be applied for the diagnosis of epilepsy and there should be a time-lien or life-line to make the definition viable, say, 2-, 5- or even 10 years. Table 2 is very difficult to understand and apply to a clinic setting, especially for neonates, infants or children. The 3 points- point 1 with "24 hrs", point 2 with "~75%" and point 3 of "at least 2" are just arbitrary figures.

As an analogy, a lot of confusion had been generated by the psychiatrists with the use of DSM only based on clinical interview and without a set time age limit for making the diagnosis i.e. DSM-3, DSM-4 and DSM-5 (2013) for the diagnostic criteria for Autism, Autistic disorder, Asperger Syndrome, Pervasive Developmental Disorders and Autism Spectrum Disorders. This is due to the fact that the checklist of symptoms and signs were not documented with any diagnostic tools. Actually there are no surrogate tools for making diagnosis for most neurodevelopmental disorders. However, we still have EEG as the so called "semi-gold standard" for helping us with diagnostic uncertainties by increasing the sensitivity range. We should specify the standard operating procedures to make this tool more diagnostic. Currently, I think only resource-poor countries like Africa are still in need of donations of "second hand EEG machines "to assist in the clinical diagnosis.

One can always add back a research definition for epilepsy with a broader scope as this will affect the epidemiology, management and prognosis etc.

Virginia Wong Chun-Nei

Return to top

29 September, 2013

Yes we do need a better picture on what defines epilepsy and the work done by ILAE is awesome and timely:

1. However, it would be better to mention Epilepsy as a chronic brain disorder with episodic manifestations (CDEM) with a tendency to relapse, remit and also no longer be present.This is because it should be made clear that most persons with epilepsy (PWE) have only brief episodic manifestations and are normal the rest of the time and that is it is not all prevailing and pervasive 24 x 7 phenomenon.

It is imperative that the term epilepsy be used only for witnessed account or documented events (on digital cameras/ mobile phones) as epilepsy still remains a clinical diagnosis.

2. .Single unprovoked seizure: It will be good not to define a person as having epilepsy after a single seizure knowing fully well that it may never recur in 70 % of them.

Hence in order to avoid the consequence of giving the diagnosis of epilepsy (psychosocial, stigma, costs etc) it would be best to let the single seizure be defined as a) .Single seizure (probable low recurrence ),b) Single seizure (probable high recurrence ) and term epilepsy be used only if it actually ever recurs.

This will not only help PWE but also epidemiologists to collect better robust data. It will help PWE as having epilepsy automatically prevents them applying for certain jobs, marriage prospects- specially in developing countries like India where women are disadvantaged anyway and adding epilepsy is like a double wound. in India employment and schooling discrimination still continues too. So though the clinician may know that the risk of recurrence is high for a particular patient and may start meds after talking to the PWE for a high risk PWE (single seizure+MRI substrate + - focal deficit + - EEG finding) significant impact on the patients QOL can be avoided by an early preemptive definition.

3. Long-interval seizures. It will be good to clearly state that epilepsy be considered only when acute symptomatic seizures and differing etiologies have been ruled out and the same etiology which caused seizures at young age is responsible at the later age of 70. .Hence please do clearly provide criteria when the person with two distantly placed seizures will NOT be considered to have epilepsy. As it could impact patients driving/ work /marriage prospects.

5. Epilepsy is no longer present: Should epilepsy in remission be a better term as it is difficult to predict under what circumstance and when the next seizure may happen and this could have significant impact for a young girl with epilepsy in India where she could be married off as Epilepsy no longer present and no disclosure be done there off as her parents are under the impression that epilepsy is NO LONGER PRESENT (gone). Once married she could have a seizure and this could jeopardise the marriage and be a cause for marital discord/ divorce on the ground of cheating (in a country like India). It is imperative that the task force include sensitivities of the developing world and of different cultures as the impact of epilepsy and the numbers of PWE  in these regions are high. It would have been good to have epilepsy specialists and PWE from the developing world also participating in the definition development as it impacts them most.

However the work done by the task force is exemplary and needs to be commended. We hope the suggestions will be looked at  so as to make it a more universally accepted entity.

These comments from the developing areas should be useful.

Manjari Tripathi and Sarat Chandra

29 September, 2013

Some years ago I sent the statement – concerning the definition of Epilepsy – that you will find in the atachment to Doctor Ann Berg. I am forwarding it in the hope that it can contribute to the discussion.

Yours sincerely
Jorge Eslava-Cobos

29 September, 2013

In general, I agree that this definition covers a broader number of clinical scenarios. It is a welcome addition. I do like the simplicity of the current definition, it is easy to use in the clinical setting.

Richard P. Morse

29 September, 2013

I very much appreciate the opportunity to comment on this valuable work. Here are my comments:

Disorder vs disease.  I prefer to refer to epilepsy as a disorder of brain function rather than a disease.

>75% probability of seizure recurrence might be difficult for non-specialist physicians to apply. 
A probabilities chart with relevant data  might be more functional universally.

The description for superscript 7 on page 8 is missing on page 16.

Lillian Martinian

29 September, 2013

First of all, I'd like to congratulate the Task Force on the work done so far. This is certainly a very difficult task, and the public consultation will surely bring very useful thoughts on the matter.

I believe the new definition is very clear and applicable on clinical practice. However, the criteria used for risk of recurrence may be a bit confusing and misleading. Maybe it would be better to use more straight forward criteria, such as 50%, so that patients presenting after a first unprovoked seizure would be divided in two groups: those more likely to have seizure recurrence and those more likely not to have this recurrence.

I would also like to share a collection of impressions on the new ILAE proposal for definition of epilepsy. During the last Brazilian Congress of Clinical Neurophysiology and Meeting of the Brazilian League of Epilepsy (Brazilian Chapter of ILAE), which was held this last month in Rio de Janeiro, we conducted a small survey regarding the new proposal.

During a session on epilepsy treatment, we asked to every attendee if he or she sees epilepsy patients on a regular basis, and five questions were directed to the audience:

  1. Are you acquainted with the new ILAE proposal?
  2. In your opinion, is the proposal clear enough?
  3. In your opinion, will the new proposal improve care of persons with epilepsy?
  4. Do you consider the recurrence risk criteria (75%) appropriate? If not, which criteria would you suggest?

Fifty-three attendees responded to the survey. Among all respondents, 41 were acquainted with the proposal. Nearly two thirds (36/53, or 68% of respondents) found the proposal clear enough, while 12 didn't; five neither agreed nor disagreed with the proposal. The majority (39/53) agreed that the new definition will indeed improve care of persons with epilepsy; five disagreed. Thirty-tree responded that the 75% criterion is appropriate, while 14 found it inadequate (half of whom suggested a 50% criterion).

Interestingly, among 10 respondents who identified themselves as epileptologists, only half found the new definition clear enough, and again only half agreed with the 75% criterion. It seems to me that neurologists who are also epilepsy specialists have a more critical view of the proposal, particularly regarding the 75% criterion. On the other hand, it should be reassuring for the Task Force that non-specialists found the proposal overall clear and appropriate for clinical use.

I would like to thank for the opportunity of commenting on the new proposal.

Best regards,
Luis Otavio Caboclo

Return to top

29 September, 2013

Thank you for the chance to share my opinion about such a global and important issue.

I did not find out a reflection of epileptic syndromes which may appear without seizures for example ESES and Landau-Kleffner in an operative definition. I think these syndromes which are usually diagnosed based on EEG data without clinic correlation (seizures) should be considered and added in a definition.

Vahabzadeh Ulviyya

29 September, 2013

Thank you for the efforts and also for asking opinion from the epilepsy community which is a democratic initiative which could be a good example for future. I have few comments:

1. Disorder vs. Disease
I personally can not consider epilepsy as a disease and do not want my patients to feel themselves as sick persons, I would rather prefer disorder or condition.

2. Recurrence risk
Including (approximately 75% or more) for recurrence risk in the definition seems to be inappropriate.
First it changes from one patient to another according not only to the etiology but many other factors as well.
Second, there is no enough data for each different types of underlying etiology.
Third, how shall we proceed if the chance of recurrence is less than 75%?
Moreover, presence of epileptiform EEG abnormality or epileptogenic lesion on MRI should better to be included.

3. "Remission" vs. "No longer present"
We can never be sure about recurrence even after 10 years seizure freedom even without drugs. For example patients with idiopathic generalized epilepsy may have a seizure precipitated by sleep deprivation, drugs etc. even in elderly. Therefore I would prefer remission and refrain from mentioning about 10 years in the definition. If it is needed for studies then the term cure is better and number of years may be stated.

Finally I agree with the definition given by William Tatum which is simplified and summarised in one sentence as:
Epilepsy is a disorder of the brain that consists of at least two unprovoked seizures (or 2 reflex seizures) occurring more than 24 hours apart; or one unprovoked seizure and a high probability of further seizures (similar to the effect of at least two unprovoked seizures).

Best Regards,
Cigdem Ozkara

29 September, 2013

I have to say that I am deeply troubled by the seemingly unstoppable ILAE push for redefining and reclassifying seizures, epilepsy, and epilepsy syndromes. The value of our current system is summed up by the recent comments of Simon Shorvon (Epilepsia 2013;54:1134):

"The great advantage of the traditional International League Against Epilepsy (ILAE) terminologies are that they are stable, well-established, universally used, and fully understood".

So, why change what works? This question is made especially compelling by the fact that there has been no particular quantum leap in our understanding of the fundamental mechanisms of epileptogenesis or of our knowledge of the clinical presentation of seizures, which always has represented the infinite behavioral repertoire of the brain, since the original ILAE terminologies and definitions have been put into place.

This question of definitions and terminology has become critical for those of us in Ontario, Canada who care for children and adults with epilepsy. The reason for this is that we, i.e. the epilepsy community in Ontario, currently have a unique opportunity to formulate and implement an evidence-based strategy for truly comprehensive epilepsy care across the life span for all patients in Ontario with epilepsy. This effort is being guided by the Epilepsy Implementation Task Force of which I am co-chair and which has the full support of the Ministry of Health and Long Term Care in Ontario. If we are successful in our efforts in this regard the result will be a unique, government sponsored plan of care that will transform the lives of 70,000 people in Ontario with epilepsy, both adults and children, for the better.

In order for any broad based strategy like this to work, careful, evidence based standards of care have to be developed with input from all stakeholders including neurologists, epileptologists, primary care providers including family doctors, pediatricians, internists, and nurse practitioners, as well as patients and their support organizations, hospital administrators, and government policy makers.

Taking the numbers and heterogeneity of the different physician specialties, administrators, and patient groups involved in this effort into account, the absolute imperative of terminology and definitions being 'stable, well-established, universally used, and fully understood' becomes immediately apparent. We epileptologists can debate endlessly about the arcane criteria for this or that syndrome or seizure to be classified this way or that or alter the definition of epilepsy such that it is apparent to us; however, we do not care for the vast majority of patients on this planet who have epilepsy. Those who do don't have a clue what we are talking about.

For that matter, neither do our patients or their families. Further, for the latter, i.e. patients and their families, simply the proposed broadening of the definition of who has epilepsy has enormous ramifications in terms of increasing stigma and complicating their life with school, insurance, driving, etc., not to mention calling into question every evidence based guideline and outcome study related to epilepsy now in existence.

We in Ontario have very tight timelines for our deliverables. So, as co-Chair of our Task Force, I intend to encourage my working groups to stick to the tried and true ILAE terminologies and classification system because that is what everyone understands and accepts.

I urge the ILAE to take a big step back and think clearly about the potential unintended consequences for all of our patients that promise to accompany the radical changes being considered.

O. Carter Snead III

Return to top

29 September, 2013

I greatly appreciate and applaud the expanded effort done by the members of the ILAE Task Force to put an operational definition of epilepsy. This will systematize the work of epidemiological researchers worldwide and make comparative studies more perfect. However I have some remarks:

  1. The definition states that "Epilepsy is a disease of the brain." I think it is better to be regarded as " a Paroxysmal brain dysfunction". It is caused by many diseases. This paroxysmal dysfunction is more near to the nature of epilepsy, and more optimistic to the patients and public.
  2. The 75% probability of recurrence after one unprovoked seizure is vague. It will open the door for much debate among researchers and create great discrepancies between different studies. It is better to be addressed "probable epilepsy" until (s)he fulfills the criteria of definite epilepsy, (as migraine, Multiple sclerosis".
  3. "Epilepsy is considered no longer present for those who have remained seizure free for at least 10 years off anti-seizure medication". Patients with active epilepsy were previously defined as those having their most recent seizures within the last 5 years. What about patients who have no seizures within the last >5 years but <10 years? Are still have probability of recurrence or they are considered no longer to have epilepsy? I think they could be considered patients with controlled epilepsy. If they remained off anti-seizure medication for further 5 years, epilepsy could be considered no longer present.

Wafaa M.A. Farghaly

29 September, 2013

Dear colleagues and members of ILAE.

I have read the very interesting commentaries made by several of my colleagues around the word and I feel that all have been of great knowledge to me. I think that it has been a good initiative of ILAE and I appreciate the opportunity of having a suitable scenario to share opinions. Now, I shall start with my point of view.

The definition of seizure proposed by the ILAE task force in 2005 and with which I do not totally agree is: a seizure is the manifestation of "a transient occurrence of signs and/or symptoms due to abnormal excessive or synchronous neuronal activity in the brain". This concept involves the current ideas about the origin of seizures in the brain and not only in the cortical areas, as in the case of seizures originating in hypothalamic hamartoma. That point of view is correct and is supported by the new knowledge obtained from deep intracerebral recording in patients surgically treated because of focal resistant epilepsy.

Nevertheless, according to the definition of seizure, to generate seizures, the brain discharges should be abnormally excessive and synchronized. I wonder how the clinician explores what is excessive and synchronized neuronal activity, in the clinical setting. This could be confusing and difficult to know. These characteristics cannot be evaluated in the clinical setting and the clinician must assume its existence when the diagnosis of seizures has been established. Thus, a seizure is suspected before the pathogenic mechanism is known.

On the other hand, in Landau Kleffner syndrome (LKS) seizures are scarce. In such cases, only aphasia is recognized for a long period of time. Most of the patients can be diagnosed only after any transient occurrence of signs and/or symptoms due to abnormal excessive or synchronous neuronal activity in the brain are present. Because of that, the diagnosis can be delayed.

Aphasia in LKS is not a transient symptom. Nevertheless, aphasia is a clinical manifestation of abnormal excessive activity in the brain. As aphasia in the LKS, many other symptoms are not transient, but are a long lasting expression of an abnormal excessive and synchronized neuronal activity in the brain. For example, cognitive deficits in complicated rolandic epilepsy, the drooling in the biopercular syndrome, continuous auras in temporal lobe epilepsy, limbic status and different disorders of consciousness due to focal or generalised abnormal, synchronized and excessive continuous brain activity.

Up to date, according to the 2005 ILAE definition of epilepsy, it is difficult to consider the above mentioned symptoms as epileptic seizures. That is why; I think that the concept of seizures should be limited to say "a seizure is the occurrence of signs and/or symptoms due to abnormal activation of neuronal activity in the brain (eliminating the words excessive and synchronized). It can be transient or long lasting (adds long lasting). In cases of prolonged symptoms, the nature of seizures should be suspected only until adequate correlations with epileptiform activity in the electroencephalographic recordings were recorded. The term "until" is referred to the necessity of apply special montages, prolonged recording and different activation maneuvers.

Thus, we eliminate the current ILAE inconsistency in the concept of epilepsy when it states that "the definition of epilepsy requires the occurrence of at least one epileptic seizure". The contradiction emerges when we assist a patient with a prolonged language disorder (such as aphasia) without other typical seizures. In this case we cannot diagnose an epileptic disorder until one typical and transient seizure is found. In the case of LK syndrome, it can occur after a lengthy period of time and the initiation of appropriated medications can be delayed.

Finally, I want to referrer to four other points:

  • First, we consider that epilepsy is a disease and that as other well-known diseases, different types are present. But the predisposition to present seizures constitutes a disease in itself.
  • Second, I would prefer to stop employing the terms unprovoked or provoked seizures. Thus, we avoid the possible confusion that could appear with reflex seizures that are in fact provoked seizures. So, instead, we should say epileptic seizures (without the words provoked or unprovoked. Those seizures where a clear temporary cause is demonstrated should be called symptomatic seizures.
  • Third, the risk of seizure recurrence should be more than 75% to operationally diagnose epilepsy; but I think that the operational concept should emphasize that when myoclonic, absences, spasm and dialeptic seizures are diagnosed, the recurrence risk should be evaluated differently. The reason is that these types of seizures are difficult to count and when the patients come to the outpatient setting, many them have occurred. That is why, when spasm, absences, dialeptic or myoclonic seizures are diagnosed, the long lasting predisposition to generate seizures is high and consequently, epilepsy should be diagnosed.
  • Fourth, an operational diagnosis has therapeutic implications, but it does not directly imply that treatment should be initiated. This decision is made according to the physicians' judgment and implies many other factors such as: age, sex, personal history of diseases, nutritional state, pregnancy and many others. The doctor should evaluate cautiously the individual risk-benefit ratio and the available options.

Taking into account the above mentioned clarifications, I totally agree with the concept of epilepsy offered by the 2005 ILAE task force.
In conclusion I propose the following operational definition of seizure and epilepsy.

An epileptic seizure is a transient or long lasting occurrence of signs and/or symptoms due to abnormal neuronal activity in the brain. In cases of prolonged symptoms, the nature of seizures should be suspected only when an adequate correlation with epileptiform activity in the electroencephalographic recordings is demonstrated. The term "when" refers to the necessity of applying special montages, prolonged recording and different activation maneuvers.

Epilepsy is a disorder of the brain characterized by a permanent predisposition to generate epileptic seizures, and by the neurobiological, cognitive, psychological, and social consequences of this condition. The definition of epilepsy requires the occurrence of at least one epileptic seizure and that the specific temporary or reversible factors causing acute symptomatic seizures were accurately ruled-out.

René Andrade Machado

Return to top

29 September, 2013

The task force is to be commended for the work they have done in producing this paper and for giving the community the opportunity to feedback on their recommendations. I have the following comments.

1) Disease or disorder
Epilepsy refers to a symptom complex, it is still best considered a disorder, whereas a distinct aetio-patholgical condition is a disease.
Whilst it is fair to consider epilepsy as a 'disease of the brain' that results in seizures, the reference to disease here is based on the sentence in which it is found, we cannot and should not extrapolate this to – 'epilepsy is a disease'. We need be mindful of how terminology is interpreted, used and perceived.

Defining a disease needs knowledge of a distinct aetiology and or pathology, hence we have Alzheimer's disease but memory disorder, Parkinson's disease but movement disorder, Lafora body disease but epilepsy as the disorder. The example of cancer / diabetes as diseases are given in the paper, however similar disease / disorder debates exist in these fields. Furthermore, in cancer / diabetes we have knowledge of, and they are diagnosed based on pathological findings – not a symptom complex.;

Secondly and perhaps more importantly is the increased stigma the term disease can bring; disease implies a contagion, particularly where pathology isn't known, along with a need for isolation. We should, to a great degree be guided by what our patients think; I don't think the term disorder minimizes the seriousness of epilepsy, but the term disease does have the risk of stigmatizing epilepsy further – let us stick to disorder for now.

2) The "two unprovoked seizure" definition.
I agree with the task force – to a point. Modern imaging, the EEG and our understanding of the disease's (where we have knowledge of aetiology or pathology) natural history should allow a diagnosis of epilepsy to be made after a single seizure if risk of recurrence is high based on prior knowledge.

I disagree that we should always diagnose epilepsy after a single photosensitive seizure with PPR on EEG. Sleep deprivation, inter-current illness and age related predisposition need to be taken into account. The risk of further seizure, with increasing age, lifestyle modification, resolution of inter-current illness, could be considered sufficiently low not diagnose epilepsy in a child or young adult; there may not be 'an enduring abnormal predisposition' and the term 'provoked seizure' may be more appropriate in some circumstances of photosensitivity.

3) Epilepsy is considered to be 'no longer present'.
The following are given as examples:

  1. "Individuals who had an age-dependent epilepsy syndrome but are now past the applicable age"
  2. "those who have remained seizure-free for at least 10 years off anti-seizure medicines"
  3. "after resection of their hippocampal sclerosis"

The above groups are clearly distinct in terms of the epilepsy syndrome, aetiology and treatment intervention, and shouldn't therefore come under the same operational definition. Consider:

  1. some age related syndromes, for example BECTS are resolved in adulthood.
  2. the patient with JME or frontal lobe epilepsy who is seizure free for 10 years off medication is seizure free for 10 years, but cannot be said to have epilepsy no longer present - they still have epilepsy and continue at a higher risk of a further seizure than the background population; and if a seizure then occurs do we then define them epilepsy now returned?
  3. What about those who have continued to take medication but might still have been seizure free at 10-years had they chosen to stop, in terms of neurobiology they are the same, but we are grouping them differently here.
  4. the authors state "Clinicians will have to individualize a determination of whether epilepsy is no longer present." This is too imprecise in complex healthcare environments with different levels of healthcare provision to make this proposal a useful operational definition.
  5. The term 'epilepsy no longer present', seems no less stigmatizing that 'epilepsy in remission'. Surely, preferable to say history of epilepsy, now seizure free.

So, operationally let us just call it what it is: epilepsy, syndrome / aetiology, relevant MRI or EEG finding - seizure free off-treatment x-number-of-years.

In order to progress understanding and avoid confusion we must base any new operational definitions on knowledge about aetiology or causative pathology, not symptom manifestations. In the absence of such knowledge, which remains the heart of the problem in epilepsy definitions, we should stick to common but accurately descriptive, and not overly reduced or grouped terminology.

At the same time our greatest efforts should go into understanding the diseases that underpin the disorder – epilepsy.

Thank for the opportunity to comment on this paper.
Kind regards,
Khalid Hamandi

Return to top

28 September, 2013

I wish to congratulate with the task force for this significant advance in the definition of epilepsy and, at the same time, to highlight a few points of debate.

1) disorder or disease? probably the term epilepsy encompasses a group of disorders and diseases (where the term disease usually implies a known aetiology). However I agree that the term disease is more useful to provide an impact on the media an the public.

2) how to define the 70-75% risk for developing further seizure after a first unprovoked seizure? The experienced epileptologist may forecast the risk on the basis of a number of features but the are few available data in the literature for many aetiologies. Therefore the task force could try at least to enumerate the conditions for which it is documented the 70-75% risk of having further seizures instead of describing only a few examples. Or otherwise could try to infer that any structural aetiology may give this 70-75% risk.

3) I agree with the inclusion of reflex epielpsy. A good point is also the paragraph in which it is specified that the diagnosis of epilepsy does not necessarily imply the need to commence an antiepileptic treatment.

3) Overall, this definition fits the needs of epileptologists but seems to be more difficult to apply for general neurologists.

Roberto Michelucci

28 September, 2013

I think the "operational definition" should be practical, helpful in clinical use.

I prefer the term disorder instead of disease, as described in the "conceptual" definition of epilepsy (Fisher et al, 2005). The discussion whether epilepsy is a disease or not is conceptual.

The 75% recurrence risk criterion may be difficult to apply in clinical practice. The decision in clinical practice may depend on whether risk is higher than 50% or lower. Just qualitative words such as high probability of further seizures with describing various aspects including EEG, MRI may be an alternative.

I find remission more usable than "no longer present". The purpose of defining "no longer present" may not be clear. When considering fitness to drive heavy vehicles, 10 years seizure freedom without medication may be appropriate, but otherwise too long and whether taking medication or not may not be essential for clinical practice. Five years seizure freedom with or without medication might be enough, as described in the ILAE commission report of 1997 where epilepsy in remission was defined as a prevalent case of epilepsy with no seizures for ≥5 years with or without AED treatment at the time of ascertainment (Epilepsia 1997;38:614-8). The situation has probably not changed.

Yushi Inoue

28 September, 2013

Unprovoked seizures separated in time:
The example mentioning by the paper is very extreme (two unprovoked seizures at age 1 and at age 80 – really needs also an intact memory of the pts, too /cit. P.Wolf/). But it is not rare at all when patients have unprovoked seizures in every second – fourth year (even mixed with provoked ones)See reference. I think these patients alters from the clinically active ones in their clinical/therapeutic and even prognostic aspects but on the other hand their epilepsy must somehow considered as still active disease/disorder. I am not a fan of the maintenance of the old-fashioned term oligoepilepsy but I would support the differentiation of these "more benign" subgroup from the others.

Epilepsy no longer present:
I agree that the epilepsy of a seizure-free patient of many years but still on drugs (of therapeutic dosage) could not consider "no longer present" or "remitted" (first of all because of the risk of accidental withdrawal seizure!). But I also fully agree with P.Wolf that many pts with long-term seizure freedom are unable to leave their AED (as a kind of 'psychological dependency'). The majority of them have good jobs, some of them that of extreme high level/responsibility. After some steps of tapering they want to stop with. Just all of them remain on a definite "subtherapeutical" dosage (like 150mg valproate or 100 mg carbamazepine etc per day.). I consider this treatment as an efficient 'active placebo' therapy.

So I would allow to include patients still on antiepileptic monotherapy of subtherapeutic level: ... remained seizure free for at least 10 years off or still on one and deeply subtherapeutical level (not more than 25? 33? % of DDD) antiseizure medicine, ...

Necessity of "lifelong treatment of JME":
Agreeing again with P.Wolf, I suggest to omit this statement. My practice also does not fully support it, therefore it may not be free from unnecessary iatrogenic consequences (mainly of psychological nature).

Thanks a lot for giving the opportunity to discuss this very valuable and important work of the Task Force.

Best regards,
Peter Rajna

Rajna P, Solyom A. (2011) [Oligoepilepsy: a real entity or the benign form of epileptic disorder?].Ideggyogy Sz 64:344-349.

Return to top

28 September, 2013

It is great effort by ILAE task force to simplify the definition of epilepsy for neurologists across the world. There is indeed a need for a uniform working definition to label a child with "epilepsy" versus "no epilepsy". Some of the concerns pertaining to adoption of this definition in a developing country like India are outlined below:

1. Labelling epilepsy as a "disease" would be a major concern in Indian perspective where lot of social stigma is attached to this "disorder" in our community. Acceptance of epilepsy as a disease in Indian community might hamper the parental emotions and sentiments. It is still considered a social evil and parents often hide the treatment issues in the society and fear that it might hamper the school admission, acceptance in society and among the peer groups. Moreover, the term "disease" in common parlance reflects infectious/contagious disease. This raises a major concern about the implications of adopting epilepsy as a disease.

2. We consider neurocystercosis as one of the commonest etiology for recent onset seizures in developmentally normal Indian children. The recurrence risk of 40-50% has been reported following a single episode of seizure and an active NCC on neuroimaging (Carpio A, et al. 2002). We perfectly agree with a cutoff of >75% as a high recurrence risk. This ensures that all children with NCC are not labelled as "epilepsy" after their first seizure. However on the other end, most of the epidemiological studies on recurrence risk based on the etiology have a range and single cut off value of 75% might create confusion.

3. We approve of the time interval of "24 hours" between the seizures in the Indian context also. There is often a considerable delay in seeking medical attention owing to long distance to nearest health care centre and lack of appropriate expertise at primary health care centre. Domiciliary management of seizures although repeatedly emphasised is only administered in a minority. Also there is lack of emergency helpline facilities for emergency management at home and transfer to hospital. Hence it would be appropriate to conclude that seizures recurring within 24 hours might be considered a single event.

4. We welcome the inclusion of "epilepsy no longer present" as a new terminology for individuals who had an age-dependent epilepsy syndrome but are now past the applicable age. However, in children age dependent epilepsy syndromes like early infantile epileptic encephalopathy and west syndrome often evolve to Lennox Gestaut syndrome when they go past the infancy or early childhood period, but by definition the evolutions are not considered. Generalization of term 10 year seizure free period may be inappropriate especially in children with epilepsy secondary to calcified neurocystercosis. We would prefer using the terminology of "epilepsy no longer active", that sounds more scientific rather than "epilepsy no longer present". The definition of this entity could include "off any antiepileptic medications/dietary therapy"

5. Neonatal seizure as a separate entity has not been addressed in the definition.

6. The inclusion of reflex epilepsy in definition of epilepsy not only provides clarity but also highlights the entity which often remains undiagnosed and ignored.

I appreciate the task force for opening the discussion and valuing the opinion of its members. Hope my comments are useful and any further clarifications I would be happy to address.

Sheffali Gulati

28 September, 2013

The conceptual definition (Fisher et al., 2005) states that Epilepsy is a disorder of the brain. However the operational definition affirms that Epilepsy is a disease. Obviously, if definitions have to be useful, both of them have to consider Epilepsy as the same entity.

In my opinion Epilepsy is obviously more a disease, than a disorder, irrespective of the public understand. And, even going beyond, it should be more precise to use the term Epilepsies instead of Epilepsy and, in this sense, to define the Epilepsies as a set of diseases. If we define disease as an entity with a (more or less) well defined pathogenesis, physiopathology, course and prognosis; for example, it is clear that Dravet's syndrome and mesial temporal lobe epilepsy didn't share any of these features, except the presence of seizures (also clearly different).

Sincerely yours,
Jesús Pastor

28 September, 2013

1) Epilepsy is indeed a disease rather than a disorder. But the definition should be reviewed again to exactly emphasize that "not all epilepsies are the same" for patient and public education. "The frequency and types of the seizures as well as severity of this disease change widely between the individuals and may subject to change even in the same individual."

2) It should also be emphasized that family history of epilepsy and genetic variations are important to define the recurrence risk to promote the involvement in genetic studies.

3) "At least 75% risk" is very hard to explain to the patients. Could we say only "high risk".

4) I prefer as an aprropriate medical term "remission" instead of "no longer present".

5) I also think like many others that the following phrase should be omitted or rephrased "Juvenile myoclonic epilepsy is known to be subject to lifelong elevated risk for seizures" in the light of current data.

How many patients with JME have you seen after the age of 40-50 years ?

Ref: Baykan B, Martínez-Juárez IE, Altindag EA, Camfield CS, Camfield PR. Lifetime prognosis of juvenile myoclonic epilepsy. Epilepsy Behav. 2013 Jul;28 Suppl 1:S18-24. doi: 10.1016/j.yebeh.2012.06.036.

Many thanks for your huge efforts and sharing the work with us in the right time before publishing!

All the best
Betul Baykan

28 September, 2013

Thank you for the very interesting, clear and informative document. I am grateful to have the possibility to give my comment based on my simple experience on clinical ground. Personally I see positive aspects in the new proposed operational definition. I would like to underlie the important consequence mentioned in the article of "sensitizing clinicians about the need to give greater consideration to the risk of recurrence after a single unprovoked seizure, and making the clinicians more comfortable in initiating treatment after some unprovoked seizures".

It is a daily experience (I work as paediatric neurologist in a Tertiary Hospital in Zambia - Africa) to see children with severe neurological sequelae due to prolonged or repeated seizures and status epilepticus who could have been prevented by a more attentive initial diagnosis and management.

I believe that in a setting like ours in which people do not have an easy and fast access to health facilities (ER), the use of DZP rectal/ or MDZ at home is restricted by the local legislation and economical constrains, and the incidence of childhood symptomatic epilepsy due to perinatal injuries and CNS infections is very high, it would be beneficial to increase the awareness among physicians of the risk of recurrence of unprovoked seizures, especially in children with other risk factors.

In terms of decision taking on when starting treatment, factors like adherence and accessibility to AED, risk- benefit of the treatment and other socio-psychological factors are to be taken into account but a higher level of awareness about the risk of recurrence of seizure (especially SE) I believe could help improving management and outcome of many children in developing countries.

Kind regards,
Ornella Ciccone

Return to top

28 September, 2013

Collective comments on the "An Operational Clinical Definition of Epilepsy"

1, We appreciate very much to Dr. RS Fisher's team for their good efforts to a dynamic process of epilepsy definition, which we think is a significant progress;

2, We agree to define epilepsy as a disease of central nervous system rather than a disorder of brain dysfunction, though in Chinese language no difference between the two words;

3, We think that using of the definition should not be limited only for clinical doctors, but also to satisfy the practical use in all other areas, such as in scientific researches, in medical management, patients and their carers, etc.

4, Although "provoked and unprovoked" seizures are not easy to be understood, sometimes there is no clear boundary between them; we could hardly give another word which is much better than "Provoked" to describe the characteristics of seizure, and this word has been used for long time, so we accept this word to be used in the definition.

5, "What would be that potential risk factor, which is associated with a probability of future seizure higher than 75%?" is very hard in some clinical situations, especially for doctors who are not specialists in epilepsy. It, however, is acceptable, because we understand that it's just a conceptual definition meaning that patient has strong potential to have additional seizures. The percentage 75% is too accurate to master clinically, we suggest to use another way to emphasize the high probability of future seizure.

6, Concerning "reflex epilepsy", the new definition doesn't claim clearly that two seizures must occur more than 24 hours apart, same as unprovoked seizures?

7, About "Epilepsy no longer present": How about the patient after epilepsy surgery? Patient's seizure has high tendency to recurrent as the duration after the surgery gets longer, even if the patient has not been on medication for over ten years. Can we name the patient's epilepsy "no longer present"?

With kind Regards,
Shichuo Li, on behalf of an expert team from China Association Against Epilepsy

28 September, 2013

I agree in general with the new definition. In my opinion is difficult to say a disease in a child with typical absences. For example is more easy to talk about a disorder because this syndrome has a good prognosis and the word disorder is more appropiate.

Best regards,
Santiago Galicchio

28 September, 2013

Thank you for asking for comments. Although the current definition of epilepsy has flaws (as all definitions do), I wonder if change is currently necessary in the absence of any fundamental advance in understanding of epilepsy? Furthermore, any change has the potential was unintended consequences which can outweigh the benefits. If I may be permitted to offer the following brief observations to add to the debate:

1. The importance of the conceptual basis of any definition:
A change in definition by the ILAE is a serious matter. It is a weighty decision which in my view should be undertaken only when there has been a sea change in the understanding of the conceptual or physiological basis of epilepsy. The best definitions (in any disease) are those which reflect physiological process.

Similarly, I wonder if "recurrence" conceptually the right parameter on which to base a definition. Recurrence depends on a whole raft of factors, many of which apply uniquely in individual situations, and in many different situations the definition will be no better than the current definition. It seems a potentially superficial parameter which to feature in an official definition of Epilepsy.

2. What is unprovoked:
The concept of "unprovoked" is very difficult. Most seizures are "provoked" in some way. Surveys of patients (and they afterall should know!) have shown that approx. 40-90% of their seizures are, in their view, provoked, and a quarter of patients consider all of their seizures to be provoked. Provoked seizures occur whether or not there are also structural, genetic, metabolic, immunological or other addition causes. Causation is multifactorial, and where the line drawn about what is provoked, and what is not, is completely arbitrary. Definition should avoid arbitrariness.

3. Operational .v. conceptual:
The ILAE proposed an excellent new definition in 2005 which was stated to be a consensus and "practical and operational definition applicable both in medical and nonmedical settings". The new proposal has reclassified the 2005 proposal as a "conceptual" definition, and proposes a new "operational (practical) definition". In my view the 2005 definition was both practical and operational.

4. Downsides to change:
Changes in definition can cause confusion, especially amongst non-specialists in clinical practice, in published guidelines and regulations (many of which would be invalidated at a stroke by a new definition of epilepsy), in legal and legislative and regulatory circles, and in scientific discourse and research.

Furthermore, too frequent a change in the official version, without sufficient theoretical justification, risks trivializing a process on which much intellectual authority of the ILAE rests.

Simon Shorvon

28 September, 2013

The proposed definition does seem to be an improvement, bringing into the definition cases with well-known high recurrence rates.

May I suggest the term 'triggered' for seizures regularly caused by photic stimulation, music, mental calculation, fine finger manipulation, reading, or other physiological activities. The association of such 'triggers' with epilepsy syndromes is clear, but clinicians and teachers often have to spend time on distinguishing between provoked seizures, due to sleep deprivation, and provoked seizures due to photic stimulation. If such 'triggers' as I have referred to were to be given a separate collective name, such as 'triggers', it would improve clarity of discourse and understanding. The presence of triggers, easily understood as triggers, would then ead to a diagnosis of reflex epilepsy or generalised epilepsy (JME for example).

John Willoughby

P.S. I made a comment previously (6 August) in the absence of the full definition document that I have now read. It was very helpful, and I would not have made my original comments had I been able to read it. John W.

Return to top

28 September, 2013

Thank you for the possibility to comment on your manuscript.

I have the following comments to make:

1. Single unprovoked seizure with high risk for another

a. The possibility to diagnose epilepsy at an earlier time point (one unprovoked seizure +75% recurrence risk for another seizure) may be conceived in the light of a certain analogy to other neurological diagnosis such as Multiple Sclerosis (MS) and Clinically Isolated Syndrome (CIS). Here, CIS (one clinical attack + e.g. certain fulfilled neuroradiological criteria) offers the advantage to start immunomodulatory therapy early and hence lower the risk of further relapses which is a benefit for the patient. However, in your chapter "implications for treatment", you explicitly state that diagnosing epilepsy after a single seizure with high recurrence risk may not initiate treatment ("Diagnosis of epilepsy and a decision to treat are 2 related but different issues"). So, would an earlier epilepsy diagnosis only mean earlier support by health insurances or access to social infrastructure? If not, please cite in your treatment chapter epidemiological studies about risk reduction of seizure recurrence and long term outcome by early anticonvulsive treatment.

b. The identification of medical conditions with 75% recurrence risk for another seizure is very hard as you state in your manuscript. Ideally this approach should result in long lists of medical conditions with proved elevated recurrence risk. Does a 75% recurrence risk however take into account the genetic heterogeneity of patients? Is the seizure recurrence risk always the same for different patients of a certain age group suffering from the same medical condition (e.g. brain injury, stroke, encephalitis)? There appear to be modifier genes even in patients with symptomatic /structural & metabolic epilepsy, as it has been shown that relatives of patients with acquired brain injuries and epilepsy had more frequently epilepsy (1.8%) compared to relatives (0.5%) of non-epileptic controls (Berkovic SF et al. Human epilepsies : interaction of genetic and acquired factors. Trends Neurosci 2006;29 :391-7 and references therein). These results suggest that a 75% recurrence risk will possibly be not applicable to an individual patient, as this number will likely correspond to a mean recurrence risk in patients (of what age group, adults?) with a certain structural/metabolic alteration.

c. Apart from the seizure recurrence risk of a lesion in a certain age group, the ontogenetic properties of a certain structural/metabolic abnormality and its seizure recurrence risk are not well known. In other words, seizure recurrence risks have to be formulated not only for certain types of structural/metabolic abnormalities but probably also for age groups.

2. Chapter "two unprovoked seizures"
As you state, "unprovoked is (...) an imprecise term" and "concepts of "unprovoked" and "provoked" therefore, do not clearly distinguish epileptic from reactive seizures". This is true but the issue is not resolved in the current manuscript. Should both expressions be kept and decisions on e.g. driving ability (fitness to drive) be made as before? That is, should driving ability be banned for 6 months after an unprovoked seizure versus 2 months after a provoked one (as done in several European countries)? Or should the discrimination between provoked and unprovoked be cancelled and driving ability be banned for e.g. 4 months after a first seizure?

Kind regards,
Heinz Krestel

28 September, 2013

Thank you for the opportunity to express my comments on the new definition of epilepsy.

1. Will the definition cater for patients who have one acute symptomatic seizure and a risk of further seizures similar to 75%? We have many acute symptomatic seizure patients who in general have a risk of further seizures in the region of 30% but this group of patients may be heterogeneous. E.g. a patient with acute symptomatic seizure during a cancer-related episode of acute demyelinating encephalomyelitis (ADEM) may have a much higher risk of recurrence. The qualifying period for the 75% risk may also be given in the definition as a different follow-up periods will yield different interpretation (I noted in the document that the 75% refers to a lifetime risk).

2. In some parts of the world, the restriction on driving is very stringent and the only way in which patients can drive again is to remove them from their epilepsy diagnosis. In the Lossius paper the risk of recurrence after 5 years off medication is rare. In Bonnett's paper, which explored the issue of driving after stopping medication, the risk of recurrence already drops to 17% at 12 months for the next 12 months. I believe there may be room for discussion regarding the qualifying period for taking off the label of epilepsy between 5-10 years off medications.

Yours sincerely,
Howan Leung

27 September, 2013

Thank you for your extensive effort and for giving the opportunity to comment the proposed new definition of epileptic seizures and epilepsy.

I agree with the opinions expressed by numerous commentators, such as Burnham, Caraballo, van Emde Boas, Sartucci, Neville, Eeg-Olofsson, Lee, Duncan, Beghi, Wagner, Shinnar and Sillanpää among others. My thoughts seem to be in line with many other pediatric neurologists.

With kind regards,
Elina Liukkonen

27 September, 2013

We welcome the prospect of a revised clear and accurate definition, to be agreed and used in clinical practice; while acknowledging that because this is a developing field, no single definition will be perfect or long lasting. We would encourage the adoption of one definition of epilepsy to be used in clinical practice, research, by the legal profession and government agencies etc, rather than just in the medical environment (as proposed). To have parallel or differing definitions would create ambiguity which would not be beneficial to either the epilepsy field or persons with the condition. ...

Read complete comments from Pete Scott

27 September, 2013

Thank you for the opportunity to comment this document and the reminder.

I fully agree with Brian Neville's comment (August 30) in some of the most severe epilepsies in young children may fail to fulfill the proposed operational clinical definition of epilepsy. I suggest adding a remark to the end of this part of the document, eg. something like this:
"Although the operational definition of seizures is beyond the scope of this report, it is worth mentioning that severe epilepsy in young children may rarely manifest mainly as arrested cognitive development or loss of previously achieved skills."

Best regards,
Eija Gaily

Return to top

27 September, 2013

In our opinion this proposal is less controversial than the proposed etiological classification and the suggested new terminology.

Re: Definition of epilepsy and epileptic seizures
OK. However, we question the need for including the consequences of epilepsy into the definition. To name it epilepsy in patients with one unprovoked seizure plus an enduring seizure generating lesion is in agreement with current practice.

However, a 75 % risk of recurrence is not very operational. We suggest that it should be replaced by "high recurrence risk" (≥75 %).

Unprovoked seizures are difficult to define. All seizures are precipitated by particular circumstances. Unprovoked = spontaneous (?) implies that seizures are not provoked by distinct provocative factors outside physiological limits, i.e. outside situations that often occur in daily life. Such a definition make it redundant to include reflex epilepsies in the definition as reflex seizures are provoked by stimuli which is difficult to avoid in daily life.

To us it is unclear why the old term situation-related seizures are taken out. Such seizures are not synonymous with provoked or acute symptomatic seizures, e.g. seizures occurring within a week after a stroke. Situation-related seizures are most often associated with a particular life-style.

Re: Symptom or disease
We completely agree that epilepsy deserves the term disease, while the seizures are merely symptoms.

Re: Treatment
Treatment decisions should be individualized and among other things based on the desires of the patient. However, we can not expect all patients to be well informed about the risk-benefit of the different interventions.

Re: In remission or cured
We can agree upon the term "no longer present" instead of "cured", and that clinicians and the patient together have to decide whether to stop treatment or not.

Re: Probable epilepsy
We have all patients in whom the epilepsy diagnosis can not be established with certainty. We agree that "you probably have epilepsy" is far better than "you have probable epilepsy". However, how should we handle such cases regarding legislation, for example driving license?

Morten Ingvar Lossius

26 September, 2013

While we appreciate the efforts of the ILAE task force, there are several areas within the proposed changes in the definition of epilepsy that we find problematic as follows:

1. Is epilepsy a disease? Epilepsy is tendency to recurrent unprovoked seizures. Are recurrent unprovoked seizures a disease?


  • arrhythmia is tendency to recurrent hear irregularity. Is arrhythmia a disease?
  • syncope is tendency to recurrent loss of consciousness. Is syncope a disease?
  • angina is tendency to recurrent chest pain. Is angina a disease?

What these conditions have in common with epilepsy is that they are paroxysmal, often unprovoked and reflective of multiple underlying conditions. We would be better served by exploring underlying triggering mechanisms for the onset of symptoms in these conditions. For example, what is the role of physical, social and emotional stressors in provoking recurrent "unprovoked" arrhythmia, syncope, angina or seizures? By focusing on the symptom, rather than an underlying disease mechanism, we empower both the patient and doctor to learn more about what brings about the symptom. In the process we might find that so-called unprovoked seizures have as yet unrecognized provocative causes. Hence, seizure disorder is a better model than seizure disease.

Similarities to epilepsy: Arrhythmia, syncope and angina have multiple causation. Arrhythmia may be caused by heart disease (coronary artery disease), valvular heart disease; mitral valve prolapse (a condition, not a disease). Syncope may be caused by heart arrhythmia secondary to a form of vascular disease, low blood pressure, hypovolemia and a variety of diseases or a benign condition. Angina is recurrent chest pain with or without dizziness secondary to decreased coronary artery blood flow.

Recurrent seizures are a symptom as well. If there is altered brain structure as in brain tumor, vascular malformation, cortical dysplasia, stroke, West's syndrome, the underlying condition is the probable disease and the seizure a symptom of the disease.

2. Aside from the inaccuracies noted above, classifying epilepsy as a disease brings with it stigmatization that is an all too common occurrence despite admirable attempts by various epilepsy organizations to educate the public about what epilepsy is: the tendency to recurrent unprovoked seizures. Here, "seizure disorder" brings into view for the public the actuality of what epilepsy is, without the baggage of generations of stigma attached to the term "epilepsy is a disease".

3. Whatever the underlying risk we are dependent on the clinical experience of the treating physician to determine risk of recurrence. The 75% threshold is not measurable for the individual case, needs to reflect the literature in view of evidence based medicine and is subject to measurement error. What we are really dealing with is the phenomenon of "more probable than not"; i.e., is the risk of recurrent unprovoked seizures "more probable than not", in which event a >50% threshold would be as accurate as 75%. Does an arbitrary 75% become more of a certainty because of our timidity about being wrong.

4. "Epilepsy no longer present" after 10 years does not take into account non-medication measures that individuals might have learned to gain total control of seizures during the 10 year time frame they are seizure free. For example, individuals may have learned that not getting adequate sleep, or consuming large amounts of sugared beverages with resultant fluctuations in blood sugar, drinking even 1-2 glasses of wine, or getting into situations in which they feel trapped are triggers for seizures and may have gained total control of seizures by changing these behaviors. Now, off of medicine and told by the neurologist that "epilepsy no longer present", they might be encouraged to revert to old habits, particularly during a period of flux in their lives, thereby activating their "no longer present" condition.

Joel M. Reiter, M.D.
Donna J. Andrews, Ph.D.
Rosa Michaelis, M.D.

Return to top

26 September, 2013

Thanks for inviting comments. I am a neuropharmacological researcher. Antiepileptic drugs are benefit to seizure control but meanwhile have significant side effect.it do need an good definition for clinical diagnosis. I agree with this new definition of epilepsy and thanks for the hard task done.

Zhenghao Xu

26 September, 2013

Thank you for the opportunity to comment the work of the ILAE task force.

I disagree with the term disease for epilepsy, and I call it disorder.

The defintion is in general, very good, but I think that point 2 could be difficult; to identify the condition of high risk of recurrence, because, as the ILAE work say, recurrence risks are not know for the majority of indiviudal cases.

Congratulations for this work
Santiago Fernández Fernández

26 September, 2013

I have two comments.

The first is in the commission's definition of epilepsy as a "disease" rather than as a "syndrome". Although those favoring the latter have, I think, support the more rigorous definition, I agree with the commission's recommendation in that epilepsy, in terms of public understanding, acceptance, and funding, is more appropriately labeled a disease. The situation is analogous to the recurrent problems with public acceptance of the "theory" of evolution. Although "theory" is rigorous, opponents continue to seize upon the colloquial definition of "theory" to undermine scientific rigor in public debate. By defining epilepsy as a disease, we elevate it as a readily understood and serious entity on par with - as the commission points out - cancer, diabetes mellitus, and now, obesity. Certainly, if obesity represents a disease state, epilepsy with its attendant morbidities also merits the label in the world outside of the clinic.

The second is the "75%" probability. I appreciate the epidemiological underpinning presented by the authors, but it will be difficult to apply this label in the clinical setting and in the clinical research setting. It is a stark, black-white border where grayness usually prevails.

Mark Quigg

26 September, 2013

I would like to thank the taskforce for their good effort and to give me the opportunity to read this document. I agree with the new operational definition of epilepsy, although the term "no longer present" may be confusing and the term "higher predisposition to have seizure" needs to be better defined.

My best regards,
Silvia Miano

26 September, 2013

It is a worthy task that the ILAE has taken on to come up with an improved definition of epilepsy. Here is my definition:

Epilepsy is not a singular disease or condition, but rather a conjugation of many conditions all having recurrent seizures as their defining symptom. The epilepsies can be further defined as having the following set of criteria...

Referring to the disease as the epilepsies is the way to go. Everyone knows that cancer is plural, but in the case of epilepsy the public is unaware that seizure diseases can take many forms. This would be a positive change expanding the umbrella of support for the disease(s).

All the best,
Seth Wohlberg

Return to top

26 September, 2013

As a whole, the document provides well described operational, practical definitions. However I feel that there are still some other points which need clarification or operationalisation.

The concept of "unprovokedness" of a seizure remains tricky. An acute brain insult (like an intracranial hemorrhage) may provoke an acute seizure (within the first week) however, may predispose to further (late) seizures. Is this included in the definition of the provoked or unprovokes seizures? On the other hand, metabolic disturbances (hypoglycemia, hyponatremia, hypomagnesemia etc) may provoke seizures, and, when these conditions last for a longer time, may cause recurrent seizures; actually not unlike the situation after stroke (with acute and late seizures).
The definitions need to be clarified to include the difference; the current document does not emphasize these nuances.

The high recurrence risk. Not all conditions, e.g. for symptomatic brain diseases, have a known recurrent risk for seizures. The definition relies on clinical judgement (which is good), but this may lead to widely different diagnoses across hospitals, epilepsy clinics, countries etc.

The absence of a time frame within which two unprovoked seizures have to take place in order to diagnose epilepsy. A seizure in childhood and a seizure in a person's 70s is very unlikely to have the same origin/cause. I would regard these seperately and not diagnose epilepsy.

Thank you for the opportunity to comment on the document.

Rob Rouhl

25 September, 2013

I commend the effort of the task force and have been impressed with the thoughtful recommendation posted by the community! I echo the concern of an earlier comment on Table 2 that "at least two seizures" already implies epilepsy in our current definition. Perhaps the third point should read as "at least two reflex seizures". Furthermore, I would add, "elicited by the same stimulus". If a patient had a seizure induce by hot water exposure or fever as a child, then a visually induced seizure during adolescence, would we call this patient's condition reflex epilepsy? In order to diagnose reflex epilepsy, clinicians need an established relationship (more than one seizure) to a specific trigger.

C. Akos Szabo

25 September, 2013

NAEC is considering commenting as an organization on the ILAE Epilepsy Definition. I am curious that in several sections of the paper you discuss the need to rely on the physician"s discretion in determining risk of recurrence and accurate diagnosis. In the Imperfect Information section it even states, "optimal application of this definition often requires specialized diagnostic and interpretative skills..." but the article never mentions epileptologists or epilepsy centers as having the needed expertise to assess patient risk for recurrent seizures, diagnose syndromes and determine optimal treatment. I know this is an international document and that specialists may not be everywhere. Would it be appropriate to add information on epilepsy specialists and centers?

I am happy to take a stab at drafting, but wanted to make sure you felt it made sense to include. Would be happy to discuss – contact me at the number below or by email.

Ellen Riker

25 September, 2013

I would like to thank for the opportunity to comment the new definition of epilepsy.

I agree with the proposed modifications, but I think epilepsy should not be seen as a disease it is a condition.

Congratulation to the task force for this work.

Clarissa Vasconcelos de Oliveira

25 September, 2013

Thank you for the opportunity to review and comment the new definition of epilepsy. Congratulations to the task force for this work.

I think the term "disorder" may be more appropriate than "disease", because "disorder" indicates temporary or permanent derangement of electrical function of brain. Moreover, seizures and epilepsy are a symptom of a disease where the cause may or may not be known.

With regards to role or place of EEG in the new definition of "Epilepsy", I consider it's not entirely clear.

Concepcion Gomez

25 September, 2013

I disagree with labeling someone as having epilepsy after "one unprovoked seizure and a probability of further seizures similar to the general recurrence risk after two unprovoked seizures" for the following reasons:

1. Explicit risk factors meeting this requirement are not provided and an exhaustive list would challenging to develop and constantly subject to updating when any new data (in any language) came to light.

2. This definition requires the practicing physician to become a neuro-epidemiologic sleuth, digging through the literature to uncover risk factor estimates. This isn't practical for many a general neurologist... .and it is completely unreasonable to expect that non-specialist physicians with do so. Will they also then be expected to evaluate the quality of the study? What of physicians practicing in resource-limited settings without access to up-to-date literature?

3. There will likely be discrepancies in the available literature that split across the 75% threshold cited. What then? Are we back to asking the individual physician to make the call as to which study to believe based upon his/her review of the quality?

4. As you note, the diagnosis of epilepsy is not required to initiate treatment. Revising the medical definition to facilitate payment from health systems or insurers is the tail wagging the dog and sets a concerning precedent. ILAE should take the issue up directly with the payer or develop guidelines for payers delineating what warrants treatment.

5. One argument put forth is that this new definition will prompt a physician to consider treatment who might not otherwise do so in someone with one unprovoked seizure and a risk factor for recurrence of >75%. In reality, the physician who would not consider treatment in such a circumstance is probably not a neurologist and is highly unlikely to even be aware of this new definition if/when it comes into being.

6. The rush to labeling will mean that 25 out of 100 people WHO WILL NOT HAVE ANOTHER SEIZURE gets labeled with "epilepsy" and is iatrogenically subject to all the social stigma and legal morbidity associated with the diagnosis of epilepsy.

7. The definition leaves it up to the evaluating physician to determine what a risk factor is and whether that risk factor reaches the 75% threshold. Application of this definition will make the diagnosis extremely arbitrary (even more so than it is now).

8. Nothing precludes this definition from being used for research purposes. Imposing it upon clinicians is unreasonable and unwarranted.

Gretchen L. Birbeck

Return to top

25 September, 2013

I will like to request to author regarding any definition of epilepsy to include any high chance of seizure and epilepsy in certain circumstances large MCA territory infarct or a large SOL in brain encroaching to cortex so that doctor can prophylactically give any antieptics.

Other point i think everybody face the problem when seizure developed in a patient with stroke eigther within 15 days or beyond. here the definition of epilepsy as two unprovoked seizure is not applied when seizure developed within 15 days of stroke and here it known as acute symptomatic seizures. so i will request the author to clear this point also in new definition as stroke related seizure and epilepsy is the most common type of seizures and epilepsy in elderly patients.

Meghraj Patel

25 September, 2013

It is a brilliant effort by the ILAE task force to address such relevant and practical issues in dealing with patients with epilepsy. I agree with the proposed modifications except one issue i.e time frame.

As per the proposed new definition, any two unprovoked seizures occurring at wide time intervals (say 1 year & 50 years) shall also be considered as qualifying for diagnosis of epilepsy provided different etiologies are not ascertained for both these episodes.

At the same time new definition also proposes to label patients' who have been seizure free for 10 years off anti -seizure medications to be labeled as 'epilepsy no longer present'.

Both these may prove conflicting at times; for instance

Now consider a patient who had seizure at 1 year of age and as per etiology (say stroke for instance), he was ascertained to have a high risk (>75%) of seizure recurrence. He got a label of epilepsy and treatment with anti-seizure medications for 3 years (till 4 years of age). He remained seizure free for next 10 years off anti-seizure medication and at age 14 years he was labeled as epilepsy no longer present as per proposed new definition. He got cleared off all legal repercussions related to epilepsy after obtaining medical fitness certificate from treating physician and gets a driving license, etc.

Again, at 16 years, he has a seizure recurrence and no other etiology is ascertained except previous brain insult due to stroke. What would follow next?

As per proposed new definition; he would again be labeled as having epilepsy (though treating physician may not start treatment due to very low frequency of seizures and wide time interval between seizure recurrences). Then what would be the legal implications in such a case? Should the legal restrictions imposed on people with epilepsy as per national law of that country (for driving permit, sport activities, insurance & social benefits, etc) be re-enforced again in this patient? And if yes; how practical would this be?

I feel that both the time durations i.e maximum duration between two unprovoked seizures to qualify for diagnosis of epilepsy should be same to seizure free interval off anti-seizure medication to qualify for a label of epilepsy no longer present.

Roosy Aulakh

Return to top

25 September, 2013

1. I commend Dr. Fisher and the authors on a brilliant manuscript. This work adds a greater breadth to the 2005 conceptual definition of epilepsy by providing an updated working definition with a potential to influence patient management. I am pleased to offer comments that I hope are constructive. The "black box" definition is a great idea to emphasize the essence of the new definition. The authors may wish to keep it is as concise as possible;

a. For example, Epilepsy is a disorder of the brain that consists of at least two unprovoked seizures (or 2 reflex seizures) occurring more than 24 hours apart; or one unprovoked seizure and a high probability of further seizures (similar to the effect of at least two unprovoked seizures).

b. Reflex seizures might be eliminated as a 3rd separate line item in this way-to "simplify as much as possible..." and provide a greater impact.

2. The quantitative >75% metric appears intrinsically problematic at present because it is a known figure amidst a literature that is largely unknown. In practice, clinicians must decide on treatment when the likelihood reaches a 51% risk of recurrence (implying a reasonable degree of medical certainty)- at the point where recurrence is more likely than not. However, the case of a single seizure and a >75% risk is well defended. It is in theory a pragmatically reasonable limit that will help to stimulate thought before treatment. To soften it a bit, "approximately 75% or greater" may be better able to represent the concept. The literature citing a recurrence risk of 73% after a 2nd seizure and the 70% risk in example 2 with stroke in the manuscript would then be congruent.

3. I personally do not favor change to the term "disease" when describing epilepsy.
From an operational standpoint, seizures and epilepsy are a symptom of a disease where the cause may or may not be known. Despite many causes, seizures have a limited semiological repertoire. Changing terms would shift us from a causative focus to a symptom-based diagnosis. Representing epilepsy as a singular entity could have the effect of minimizing the need to understand it's complexities before treatment. I'm not sure if patients would perceive it as pejorative but it would be helpful to know this before a change is instituted.

From a conceptual standpoint, previous efforts to reduce stigma by referring to epilepsy as a "disorder" (ie Seizure Disorders centers) would reverse a past position. In this way return to an old concept rather than a new definition results. Such as change in terminology would also draw incomplete acceptance since change without predictable benefit may be divisive. It will be up to everyone in the epilepsy community to emphasize the need to continue the search for an underlying truly causative "disease" as time and technology advances, especially in the epilepsies with an "unknown" etiology and for those with drug-resistance epilepsy.

Thank you for the opportunity to comment on this important work.
William Tatum

25 September, 2013

I would like to thank and congratulation to the taskforce and and the team for their good effort and the job they performed.

Luxman Maharjan

25 September, 2013

Thank you for the opportunity to comment the new definition of epilepsy.

My main concern about this definition regards epidemiologic studies. For clinical and epidemiologic purposes diagnostic criteria must be simple and robust and let no place to doubts. With regards to this aspect I think that also the previous definition proposed in 2005 by Fischer et al. (Epilepsia. 2005;46(4):470-2.) was not at all free from concerns (for instance, I completely share the opinions expressed in: Beghi et al., Comment on epileptic seizures and epilepsy: definitions proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epilepsia. 2005;46(10):1698-9).

In my opinion, the more restrictive "old definition" of "epilepsy as "occurrence of at least two unprovoked seizures more than 24 hours apart" (Epilepsia 1993; 34:592-6) was straightforward and let less place for doubts than the new "operational clinical definition of epilepsy". Introducing the concept of "probability of further seizures similar to the general recurrence risk after two unprovoked seizures (approximately 75% or more)" may lead to huge difficulties in performing epidemiological studies as well as clinical trials, unless a more practical and immediately applicable definition is provided. I fear that introducing the "probability threshold of 75%" may introduce an excess of subjectivity in the diagnosis. For this reason, I think it will be mandatory to test the inter-rater reliability of these new diagnostic criteria to evaluate their real accuracy and clinical utility.

Francesco Brigo

25 September, 2013

Congratulations to the task force for this work. I agree with these new definitions that allow more flexibility to physicians and reflect the "real life".

Best regards
Sophie Dupont

24 September, 2013

I'd start a definition by calling it a "condition" rather than a "disease."

Karin Popkin

Return to top

24 September, 2013

I applaud the Commission initiative to listen "the floor". Is a good and democratic way to share ideas (is strange as is not usual...). Because the proposed definition "Epilepsy previously has been defined as at least two unprovoked seizures more than 24 hours apart... epilepsy also can be considered to be present after one unprovoked seizure in individuals who have other factors that are a ssociated with a very high likelihood of a persistently lowered seizure threshold and therefore a high recurrence risk.... ... of a third seizure in those with two unprovoked seizures, approximately 75%. The latter circumstance occurs with remote structural lesions, such as stroke, CNS infection, certain types of traumatic brain injury,...

OK! I agree! But, in traumatic brain injury you can have "immediate"or "acute"seizures, and also "non provoked" "early seizures" with more than 24-48 hrs interval up to 7 days, in the same individual (1). Later the individual can or cannot develop post-traumatic epilepsy – PTE – (recurrent unprovoked seizures later). Currently PTE definition do not encompass "early seizures" although they have been considered risk factor. With the new definition those having "early seizures" do they have Post Traumatic Epilepsy?

A Martins da Silva

(1) Martins da Silva A, Willmore J. (2012). Posttraumatic epilepsy. In: H. Stefan and W.H. Theodore, Editors. Epilepsy, Part II - Handbook of Clinical Neurology, Chapter 35, Vol. 108 (3rd series), Pp: 585-99. Elsevier B.V.

24 September, 2013

I think that probability of further seizures being similar to the general recurrence risk after two unprovoked seizures (approximately 75% or more) should take precedence over presence or absence of provoking factors in the definition. This would cover patients who have had seizures in the setting of provocative factors but where there is other evidence to support an enduring epileptogenic abnormality. For example:

A 15 year old male has had two seizures a week apart, each the morning after a high school party. He recalls his hands jumping erratically when he first woke up and a convulsive seizure occurred within 30 minutes of awakening on each occasion. On both occasions he estimates that he had slept for 3-4 hours; he usually sleeps 8 hours a night. He has an older brother diagnosed with juvenile myoclonic epilepsy and a mother who has been seizure free on sodium valproate started after convulsive seizures in adolescence. His routine scalp EEG shows active generalised polyspike-and-wave discharges but no photoparoxysmal response.

If I understand the proposed operational criteria correctly, the patient would not be considered to have epilepsy because his seizures were provoked by sleep deprivation but he has an enduring predisposition to seizures (i.e. fulfils the conceptual definition) and operationally he would be diagnosed as having JME by many.

Definition 2 could be changed to:
One or more seizures with or without provocation where the probability of further seizures is similar to the general recurrence risk after two unprovoked seizures (approximately 75% or more).

The statement "A seizure that is provoked by a transient factor acting on an otherwise normal brain to temporarily lower the seizure threshold does not count towards the diagnosis of epilepsy" tempts one into trying to defining 'provoked' more tightly to account for this situation but I suspect this would inevitably result in redundancy or circularity in the definition.

David C. Reutens

24 September, 2013

The proposal is in general, very good, but I think it would be more clear with some additional concepts.

The most difficult to understand is how to identify the condition of high risk of recurrence, about which there are many reports that mention the usual percentage of recurrence, but some of these situations could act differently in the single subject.

I think that the concept of provoked seizures should be explained adding the terminology of metabolic or structural factors that clearly disappear when properly treated and so do the seizures. Those factors should be mentioned as non sensorial factors or stimulus.

When the task force mention the reflex seizures as provoked seizures it should say provoked by sensorial stimulus.

I hope these comments will be considered.
Best regards.
Lilia Núñez-Orozco

Return to top

24 September, 2013

On Aug. 23, Dr. Kalarickal Oommen noted that Hughlings Jackson's definition of epilepsy had stood the test of time quite well for 150 years, which is only a slight exaggeration. In fact, Jackson gave two definitions. The first (1870) is more often quoted, but the second (1873) seems technically and conceptually more appropriate. Indeed, it was in that three year interval that he came to understand that a single theory of epileptogenesis applies to all kinds of clinical phenomena which we would now consider to be epileptic.

- definition in 1870 - "A convulsion is but a symptom, and implies only that there is an occasional, an excessive, and a disorderly discharge of nerve tissue on muscles."

- definition in 1873 - "Epilepsy - an occasional, sudden, excessive and rapid discharge of grey matter of some part of the brain."
Please understand that I am not campaigning for Jackson's definition to be 'official.' He had quite a good run, and such things can be expected to change over the centuries.

The current effort to redefine epilepsy interests me because I have been working on Hughlings Jackson as a historical subject for more than 50 years. I'm now 8.5 years into writing his biography, and probably I will need another 4-5 years to finish. It is tentatively titled: Clinical Neurology, Evolution, and Victorian Brain Science: the Life and Thought of John Hughlings Jackson. If any one is interested in discussing anything about him, please let me know.

Samuel H. Greenblatt

24 September, 2013

I do agree with the new operational definition of epilepsy.

Ikram Qaysia

24 September, 2013

Epilepsy is not a disease, it is a condition. If I have epilepsy I am not sick, I am living with epilepsy. I don't like this word "epileptic".

Hanna Novoszath

24 September, 2013

The Epilepsy Coalition of New York State represents community providers in New York State and holds professional relationships with the medical community, and in that capacity I congratulate the commission on this very important task.

The overwhelming suggestion, from patients and providers, is to define epilepsy as a "disorder," not a disease.

Thank you for the opportunity to join the conversation.

Janice Gay

24 September, 2013

I agree with the proposed definition.

Susan T. Arnold

Return to top

24 September, 2013

The ILAE document "An Operational Clinical Definition of Epilepsy" is very well written and will go a long way in improving the clinical diagnosis of epilepsy. I have a few suggestions:

1) Should non-compliance with anti-seizure medicines be a factor while deciding that epilepsy is "no longer present"? In India, often patients do not receive or take their AED medications properly. This results in breakthrough seizures and often a prolonged period of AED administration....followed by very infrequent seizures. Should a subsequent ten year period of being seizure-free still be mandatory before labeling this child as "epilepsy not present"?

2) Since reflex seizures is now being included as "epilepsy" the document should enlist with details all the known types of reflex epilepsy. This would increase awareness of this phenomenon and I believe that the list needs to mentioned. At present only "visual" type has been mentioned as an example.

3) Since the document is aimed to help the treating physician take a decision about starting AEDs, I suggest that the document mention when an EEG should be done , namely, the time duration "after a single unprovoked seizure or a single reflex seizure episode" when an EEG should be done. Is it 24 hr, 48 hr, 7 days, should the child be sleeping, duration of EEG, how many leads, etc.

I thank you for involving me in this feedback program.

With best regards, yours sincerely,
Sunil Karande

23 September, 2013

Thank you for the opportunity to comment the excellent and helpful work of the ILAE task force. Comment:

On account of the different epileptic seizure threshold of different cerebral areas, (cortical areas vs sub-cortical areas; cortical fronto-temporal areas vs other cortical areas), in patients with remote structural brain lesions, such as stroke, CNS infection, and certain types of traumatic brain injury, the recurrence risk of a second seizure in patients with one unprovoked seizure, it is likely may vary greatly according to the different cerebral area involved (Camilo O, Goldstein LB. Seizures and epilepsy after ischemic stroke. Stroke 2004; 35: 1769-75). In my opinion, it may be useful to consider this variable in the operational clinical definition of epilepsy.

With best wishes.
GianPietro Sechi

22 September, 2013

Congratulations to Dr Fischer and his group for this proposal and discussion.

I would like to raise a following issue. The new operational definition of epilepsy allows the diagnosis after a single seizure in many situations with high probability of the second fit.

In pediatric epileptology with progress in diagnosis we have more and more examples of genetically determined disorders with high probability of epilepsy exceeding the critical figure of 75%. A good example of such situation is tuberous sclerosis complex (TSC).

In total TSC community about 80-85% of patients have epilepsy (Curatolo et al. 2008). In the prospective study of 31 TSC infants followed since newbornhood 71% of them developed epilepsy till 24th month of age (Jozwiak et al 2011). In our experience with currently over 50 TSC infants followed since newborn period with vEEG carried out every four weeks, about 95% of those, who developed paroxysmal activity, demonstrated seizures within 2-3 weeks. A determination of EEG characteristics allowing identification of children with high risk of epilepsy is critical and currently such studies are ongoing within the EPISTOP, the large-scale collaborative project within the 7th Framework Programme of EC.

A new definition of epilepsy should also allow the diagnosis of epilepsy in such group of patients with very high risk (>75%) of clinical seizures and documented and recognized markers of ongoing epileptogenesis (electroencephalographic, molecular, neuroimaging), even in the absence of the first clinical seizure.

It is specific for pediatric epileptology, but early intervention allows prevention of long-lasting consequences of epilepsy, like behavior disorders and mental retardation (Bombardieri et al. 2010).

Sergiusz Jozwiak

Return to top

22 September, 2013

Thank you for allowing us the opportunity to comment. The new definition proposal is a significantly well thought out and well reasoned document. Thank you to the task force that represented a varied and diverse group of countries. I would want to recognize the hard work that went into this. The document prepared was very well organized and points made were well articulated and justified.

After reviewing the document and many comments on line, my thoughts are:
1. The new definition in its current form is very acceptable and allows flexibility to the practicing physician.

2. However we must recognize that the definition has limitations. Firstly, any attempt to define epilepsy is limited by and dependent on what we define as a seizure. In the current standard of practice, any event that is thought to be a seizure is judged to be so by the evaluating physician after he/she accumulates data from:

  • Untrained witnesses (family, friends, observers etc)
  • Patient narration
  • Somewhat trained health professionals (including physicians or nursing staff)
  • EEG (video or otherwise)
  • The response to antiseizure medications

As illustrated in the cases in the document, some of these are not reliable sources. And, the 'gold standard' video EEG may not always be available. In addition, the only factual test we have is an EEG, and while this is helpful, the very occurrence of electrographic seizures cannot be included in this definition, since as community of epileptologists, we have yet to agree what an electrographic seizure is defined as.

If we attempt to move forward to a new definition, we must first recognize that this definition cannot be absolute and at best, can only a guideline, since it's very foundation is based on opinion and not facts, clinical or otherwise, as well as 'successful accumulation of clues'. This is the very reason why many patients today who suffer actually suffer from nonepileptic seizures are defined as patients with epilepsy, until proven otherwise. In context, our working approach or 'definition' of epilepsy to some patients becomes: "by the sound of things, you may have epilepsy, until proven otherwise" rather than the other way round.

Thus, no 'definition' can be proposed as a 'defining' statement and should be treated as a guideline. And, thus we should call this effort as: "Guidelines for diagnosing epilepsy, rather than the definition of epilepsy".

3. Expanding the point above and supporting the unsaid fear of the task force, I don't feel that the definition is proposed to be an absolute statement (though it might have that implication, when it is used legally). Thus I would agree with the rest of my colleagues:

  • We should remove exact percentages like 75% chance of recurrence and substitute it by the word 'higher chance of recurrence than non-recurrence'. This essentially means more than 50% chance of seizure recurrence or occurrence. This will avoid the definition becoming obsolete if newer studies suggest different percentages, anything less than 75%.
  • Since our clinical tools to measure the presence of epilepsy are extremely limited, I think the definition should avoid attempts to define when epilepsy has 'resolved', 'remitted', 'cured', 'no longer present' etc. Instead, the definition may include that epilepsy is 'not a lifelong condition and may resolve over time'.
  • Consider an algorithmic approach to the definition of epilepsy.

4. The Task Force should include more representation from other countries. The social and cultural implications of the definition of epilepsy in China, Russia, countries from Africa, Subcontinent, Middle East etc may be very different from the rest of the world. The ILAE Task Force should be commended for making an effort to recognize this. Lowering the threshold of diagnosing someone with epilepsy, truly, can have significant social, epidemiological and cultural impact.

However having said that, again, the task force has done a brilliant job and the definition may be accepted in its current form.

Thank you for allowing the community to share our thoughts on this important topic.
Omar I Khan

22 September, 2013

I applaud the efforts of the group to redefine epilepsy by including situations that may confer similar risk for epilepsy as 2 unprovoked seizures. In this sense, I tend to agree with keeping the "approximately 75% or more"risk for subsequent unprovoked seizures in the definition, despite the well discussed limitation of not having sufficient data for many of the clinical situations we face in the clinics. I would be hesitant in advocating including in the definition all situations that are thought to be associated with a 50-70% risk of second seizure, as the risks and benefits of diagnosing - and therefore treating - in these situations would have to be judged on an individual basis.

The authors may consider including the term "at risk for epilepsy"for the group with 50-70% risk for epilepsy for the following reasons:
(a) this at risk population would be very important in future anti-epileptogenesis trials. A good definition of this at risk for epilepsy population would be as important as the definition of epilepsy for the design of these studies.
(b) would not confer all the negativity associated with diagnosing someone as individual with epilepsy, including the treatment side effects and associated costs.
(c) using the term "probable epilepsy"or "you have probably epilepsy"transmits a sense of diagnostic uncertainty. However the term "at risk"is factual (based on available evidence) and conveys a clearer message that the individual does not have epilepsy yet but carries a high risk for developing it.

Reflex epilepsy
"At least 2 seizures in a setting of reflex epilepsy": The way that this is phrased, it is assumed that the patient already has reflex epilepsy, which negates the need for redefining epilepsy in this patient. Perhaps also a time separation criterion similar to the unprovoked seizures be added here?
Suggestion: "At least two reflex seizures occurring more than 24 hours apart."

"Epilepsy no longer present"
(a) The proposed definition does not address the situation of reflex epilepsies no longer present. If an individual no longer develops reflex seizures to the known triggers, off medication, shouldn't he/she be considered as qualifying for inclusion in the "epilepsy no longer present"group. Conversely, if an individual with reflex epilepsy is reflex seizure free for 10 years because he/she is careful not to be exposed to the trigger, it does not mean that his/her epilepsy is no longer present.
(b) For the sake of brevity, I would favor a shorter term, i.e. epilepsy-free (in accordance with "seizure-free"term).
(c) I would propose to add a comment on a "seizure freedom"definition. For some high risk conditions it would be possible to lead to seizure freedom on meds. Adding this term would offer a hope for some patients in this group, who are currently not likely to be included in the "epilepsy no longer present". For instance, an individual with genetic epilepsy who receives a drug/treatment that corrects the impact of the mutation could be seizure-free on treatment.
(d) Perhaps it may be considered to not just limit the definition to individuals "off meds"but rather "off treatment", given the nonpharmacological treatments that are increasingly being used.

Aristea Galanopoulou

Return to top

22 September, 2013

My opinion regarding the operational clinical definition of epilepsy is as follows:

1. Disorder vs. Disease
I personally consider that epilepsy itself implies both disorder and disease based on the current knowledge. Pediatric epileptologists are looking after many children with idiopathic focal and generalized epilepsies most of whom enter in remission or cure until adolescence. I feel uncomfortable to apply the term "childhood seizure diseases" for these patients until the underlying causes are clearly defined (such as mutation of responsible ion channels). On the other hand, those who continue seizures such as symptomatic focal epilepsy caused by focal cortical dysplasia, hippocampal sclerosis or chronic encephalitis appear to be diseases. Thus, at present "epilepsy is considered to be a disease or disorder of the brain defined by — "is better than the proposed one sticking to "a disease".

2. Recurrence risk
I would feel more comfortable to eliminate the term (approximately 75% or more), because the percent data of recurrence risks was insufficient. Instead, the term (the presence of active epileptiform EEG abnormality, MRI lesion suggesting cortical dysplasia, etc.) should better to be described. This operational definition is more practical and recognizable.

3. "Remission" vs. "No longer present"
Theoretically, as epilepsy itself is the enduring predisposition generating recurrent seizures, the term "no longer present" may be a much more correct expression than "remission". However, considering from the fact that this proposal is an operational clinical definition, we should better to choose more practical and understandable term "remission" and even more "cure". Pediatric epileptologists recognize that the term "cure" is more appropriate than "remission" for children with benign rolandic epilepsy and other idiopathic focal epilepsies. I prefer the more simple word "remission and cure" although I recognize that the prognosis of patients with certain adulthood epilepsies is difficult to predict despite even many years of seizure absence.

4. Length of seizure-free for at least 10 years off-medication
The length of remaining seizure-free period acceptable for "remission" should be carefully determined because it can affect the daily life of adult patients; i.e. legal process such as driving licenses. There has been insufficient evidence declaring at least 10 years off-medication. The Task force referred from only two papers, which described the seizure-free period as 5 and 10 years. It should be better changed as "those who have remained seizure-free for at least 5 to 10 years off-medication at the current evidence.
As for age-dependent epilepsy syndrome, "now past the applicable age" is excellent.

Hirokazu Oguni

20 September, 2013

Congratulations on this excellent document and the energy that has been placed on a very important topic/issue, and thank you for the opportunity to review and comment. I am a 'preclinical' epilepsy/TBI research scientist so please consider my comments in that context.

I applaud the attempt to revise and 'modernize its definition, and concur with 99% of your vision for a refined definition. My only concern deals with the topic of 'disorder vs disease', and I'm not sure that it can easily (or universally) be resolved. For example, as a preclinical TBI researcher post-traumatic epilepsy (PTE) resulting from a brain injury is better defined as a functional disorder (regardless of the duration or persistence of the disorder, e.g. months-years). There is clearly an underlying pathology that relates directly to the brain injury, but not all epilepsies can be described by this underlying pathology. In contrast, a disease is defined not by function but by pathology. Aside from this, there is no doubt 'disease' carries with it stronger clinical connotations vs. disorder and indeed may benefit the patient in certain ways. However, as has been pointed out by others here it could also have a very negative impact as regards daily quality of life issues (legal or otherwise). It's the objective to make a definition more understandable to the lay public, or to the medical and legal community at large?

Thank you for this outstanding work.
Frank Tortella

20 September, 2013

We very much appreciated the effort of the commission and we agree with most part of the content of the manuscript.

However, the 75% recurrence risk criterion may be difficult to apply in the clinical practice.

Since it is well known that a considerable percentage of patients experiencing the first unprovoked seizure may be diagnosed as having a well defined epileptic syndrome (King et al., 1998; Jallon et al. 2001; Olafsson et al. 2005), we propose to change the second point of the operational clinical definition in table 2 as follows: "One unprovoked seizure in the context of a well-defined epileptic syndrome (or condition) (i.e.: JME, symptomatic epilepsy in patients with a static or progressive CNS disease and so on)".

Umberto Aguglia and Edoardo Ferlazzo

Return to top

18 September, 2013

I am very appreciative and honored for the opportunity to comment on this ILAE Task Force report. The report is generally very useful adjusting many aspects and definitions in epileptology. However, I have few comments to state.

1. High recurrence risk
"This operational definition makes no attempt to enumerate the conditions that would increase risk for a second unprovoked seizure above the threshold cited above. Doing so is the task of the physician caring for the patient."
Comment: The 75% risk for another seizure is vague, too high and lacks evidence. Leaving this estimation for personal experiences will lead to under diagnosis of illness, when it is very likely to be diagnosed and treated. As mentioned in "Example 5", that patient will not be diagnosed nor treated; based on the "75%" threshold, although he has lateralizing MRI and EEG findings that complies with his seizure semiology. We can state that the recurrence rate for cortical dysplasias is not precisely clear, yet adding data from investigatory tools shall take the risk even beyond this overstated threshold.

2. Unprovoked seizures separated in time:
"The Task Force retained the current thinking that unprovoked seizures clustering in a 24 hours period be considered to be a single unprovoked seizure for purposes of predicting recurrence risk."
Comment: We should keep in mind that if a case with an initial unprovoked seizures cluster did experience recurrence, it is more likely to be also a cluster. Hence, the decision should be better tailored according to patient's life-style risk.

3. Epilepsy no longer present:
"The Task Force chose to define epilepsy as being no longer present for individuals who had an age-dependent epilepsy syndrome but are now past the applicable age or those who have remained seizure-free for at least 10 years off anti-seizure medicines, provided that there are no known risk factors associated with a high probability (≥75%) of future seizures. Allowing epilepsy to be declared "no longer present" may lift the stigma from some who should no longer be considered to have epilepsy."
Comment: "Epilepsy no longer present" is a very important point that should be strongly encouraged and highlighted for both medical and public communities. However, I think the 10 years interval is too long. Provided that there are no known risk factors associated with a high probability (≥75%) of future seizures, as stated by this report, that interval can be way less.

Khaled O. Abdulghani

18 September, 2013

Thanks for inviting comments. I would like to suggest following points in respect of new concept of definitions and terminology.

It is good effort that in revised terminology more emphasis given on seizure disorder than epilepsy, on basis of new development in field of neuroanatomical, neurobiology, neuroimaging, cognition, but impartment issues are associated with newer concepts and terminology that there are no clear definition /boundaries specified in this tern, though this term is older and taken from historical background, therefore may not applied universally. In many countries general people even clinician draw its meaning from its gross symptoms, as these symptom associated with ingrain stigma that this disorder no longer lasting. Therefore it is also impartment of our association that ILAE take responsibility to save the person from this issue.

The epilepsy deals many issues like psychological, social, personal and occupational, neurobiological and other newer development. Therefore instead of seizure disorder, persons with epilepsy/ proposed new term can be more suitable.

New terminology like Focal, generalized and unknown causes of seizure not gives the definite boundaries for diagnosis, therefore the chance of misdiagnosis and managements will likely to be more than the previous classification.

J S Yadav

18 September, 2013

Congratulations to the task force for this work. I think that the definition is satisfactory and needs no additional comments.
Thank you for the opportunity to comment on such an important issue.

Best regards,
Asiyan Kilit Yilmaz

18 September, 2013

Thank you for asking for my comments on the new definition of "Epilepsy." Overall, I agree with this new definition. However, I have some comments to make. Please v iew my comments considering that I am not an epileptologist but a psychiatrist.

Here are my comments:

Regarding the issue whether epilepsy is a "Disease" or "Disorder," I think the word "Disorder" may be more appropriate. Sometimes, epilepsy occurs as part of a known disease. Moreover, it has the characteristics of a "syndrome."

The term "Epilepsy no longer present" simply means "cure" or not. To me, this term is slig htly vague and, unfortunately, it slightly weakens the definition.

Epilepsy is a clinical manifestation of certain abnormal electrical activities in the brai n, and it can be confirmed by detecting abnormal EEG findings. Therefore, I believe, EEG i s a very useful tool for proper diagnosis of epilepsy. However, the role or place of EEG in the new definition of "Epilepsy" is not clear. Is it omitted from this new definition? If so, then it is unfortunate.

Masahito Fushimi

Return to top

18 September, 2013

I am happy to see the great efforts being taken by the ILAE and Terminology Commission. I am a Pediatric Epileptologist working in this field since last 8 years, also I have used the different ILAE definitions and classification schemes in my studies ( presented at 7th and 8th AOEC) hence I strongly feel that a consensus should be reached regarding these terms.

Comments on Operational definition of Epilepsy

Part 1: Definition

I would first like to discuss Hauser's study in detail as our assumption for 75% risk needs to be stratified according to etiology.
In his study, the overall rate of recurrence after 1st unprovoked seizure was 30%, while there was further increased risk between 2nd and 3rd seizures ( i.e. 75%).
Out of total 204 patients, 63 developed 2nd seizure and 41 developed 3rd seizure. The data is not equally distributed as patients with remote symptomatic group were only 59 ( out of which 26 developed 2nd seizure, while 21 developed 3rd seizure), while those in idiopathic category were 145 ( out of which 37 developed 2nd seizure and 20 developed 3rd seizure).

Hence the risk stratification should include this consideration and both these categories cannot be clubbed together to predict seizure recurrence.
'Among the 37 patients classified as having idio-pathic or cryptogenic epilepsy who had a second seizure, the risk of a third seizure was 64 percent at five years. Among the 26 subjects classified as having remote symptomatic epilepsy, the risk of a recurrence was 87 percent at five years. The univariate rate ratio for recurrence among these latter patients, as compared with those with idiopathic epilepsy, was 1.9 (95 percent confidence interval, 1.0 to 3.4). The re-currences in people with symptomatic epilepsy tended to occur sooner after the second seizure than was the case for those with idiopathic epilepsy.'

Thus, The cumulative recurrence risk (on Kaplan-meir curve) for 3rd seizure after two unprovoked seizures was approx. 60% for idiopathic vs 80% for remote symptomatic which is significant.( figure 2 in Hauser's study) 'None of the following were associated with the risk of recurrence: seizure type, findings on EEG, neurologic findings, history of a previous acute symptomatic seizure, history of Todd paralysis (hemiplegia or monoplegia lasting from a few minutes to several days after an epileptic seizure), family history of epilepsy, age at the time of the first unprovoked seizure, sex, history of multiple seizures or status epilepticus at the first episode of unprovoked seizure, prescription of antiepileptic medications at the time of the first unprovoked seizure, and length of time between the first and second seizures (6 months vs. 6 months). '

Although many of the above factors have been found to be significantly associated with recurrence after AED withdrawal, can we assume on the basis of just one study that none of the above factors are significant for recurrence after 1st seizure?

In developing countries like India, the major chunk of epilepsy is due to perinatal insult or CNS infections i.e. remote symptomatic etiology. How can clinicians arrive at the figure of 75% after first unprovoked seizure? Some guidelines need to be given to those who are treating the patients with epilepsy but are not well versed with it generally.

In cases with first unprovoked seizure, we roughly stratify the risk of recurrence according to the following factors- 1. Focal seizures ( 10%) 2. Abnormal EEG ( 20%) 3. Abnormal neuro-imaging (20%) 4. Neurodeficit like cerebral palsy (20%) 5. Prolonged NICU/ Hospital stay during acute insult (20%)

So the risk for recurrence may vary from 10% to 90% depending on the number of risk factors present. But the decision to treat may still depend on the literacy of parents, accessibility to medical care etc.

It is important to understand that in developing countries, defining first seizure with low chances of recurrence will be detrimental as most of the care is still provided by the non-specialists, sometimes not even pediatricians (in the case of children). Also, the stigma generated is more and defining them as epileptic will cause a lot of confusion as it will be taken as a permit to treat for epilepsy.

Hence defining the risk at 75% without giving the necessary factors / template will serve to add to the confusion.

Part 2: Other terms:

Disorder v/s disease- I prefer the term disorder which indicates temporary or permanent derangement of electrical function of brain.

Epilepsy no longer present- I will prefer the term 'presently inactive' or 'no longer active' as it appropriately conveys the meaning. If we say no longer present- then does it 'reappear' when they get relapse?

Also, I would like to reduce the duration of seizure free interval in children to 5 years to say it is no longer active.

Probable epilepsy- The term has its place where we are not sure about the diagnosis of seizures.

Case examples- I would like to ask how will we define in cases where a child has first complex febrile seizure with an underlying h/o perinatal asphyxia and developmental delay.

I hope my comments are useful.

Kavita Srivastava

Return to top

17 September, 2013

Thank you for your giving me an opportunity to express my opinion on the proposed new definition of epileptic seizures and epilepsy. My comments are as follows:

1. The proposal is using throughout the text the concept of "seizure". I prefer "epileptic seizure", because there are so many non-epileptic seizures/fits

2. One unprovoked (epileptic) seizure and a high risk (>75%) of further seizures is a problem. In a sense, from the viewpoint of physician, I understand the broadening of the concept of epilepsy to encompass subject with one epileptic seizure who is at substantial risk of having further seizures. However, from the viewpoint of patient, I cannot see any benefits of labeling a person, on statistical grounds, as having epilepsy. Agreeing with Peter Wolf, I have never met a person who'd enjoy having told that he or she has epilepsy and feel joy to belong to the epileptic community. On the contrary, they better want to forget their epilepsy at questioning. What are the criteria of having a more than 75% risk for further seizures? Whatever the criteria, tens of percent of people would be stigmatized and lose occupational choices, driver's license, insurances, social benefits etc. Why not wait and see? Instead of applying statistical probabilities to individual risks, the physician should tell the person that the risk for further seizures is increased, what he or she can personally perhaps do to prevent seizures, and tell him or her that if, however, further seizures occur, then it's time to re-think the diagnosis and treatment. Compliance/concordance about treatment is certainly better, if recurrent seizures occur, instead of one single seizure.

3. If we have a person who has had one unprovoked (epileptic) seizure and is considered to be in need of drug therapy, the treatment concordance is often on trial, particularly if untoward side effects occur. I don't think that, after one single seizure and perhaps drug adverse effects, they would prefer to ingest drugs.

4. Epilepsy no longer present. I prefer "remission", because it tells the reality. "Epilepsy no longer present"? Who can guarantee it, if there has been a seizure freedom let's say 5 or 10 years? The proposal makes use of relatively short "long-term" follow-up studies to draw conclusions about the risk of recurrence. Recurrences may appear very delayed. In our very long-term prospective population study (Sillanpää and Schmidt 2006), at the end of 37-year follow-up, we showed that in every third patient (48/144) relapse was seen after 5-year remission. On 10-year seizure freedom off medications, 5% (7/144) still had got a recurrence (Unpublished data). I think we are on much safer side when we say that epilepsy is in remission than that epilepsy is no longer present.

5. Is epilepsy disease? A certain inconsistence in the Proposal text about epilepsy as a "disease", "condition", or "epileptic syndrome" reflects the fact that epilepsy is far from self-evidently being a disease. Whether medically a disease or not, it would be hard to marked epilepsy as a disease to patients, who have for years been trumpeted that epilepsy is a symptom, disorder or syndrome, but not a disease. Many patients with epilepsy use to say that they have epilepsy, but they are otherwise doing well. In practice, they are not ill, but they are sick. So, disorder, not disease.

Best regards,
Matti Sillanpää

16 September, 2013

I have 3 comments on this important paper.

1. A seizure that is provoked by a transient factor acting on an otherwise normal brain to temporarily lower the seizure threshold does not count towards a diagnosis of epilepsy.

The term "otherwise normal brain" is problematic. For example, what happens if a child with a congenital hemiparesis develops diabetes and has a seizure with a severe episode of hypoglycemia. The implication of the current wording is that people with pre-existing brain abnormalities cannot have a provoked seizure. I would suggest that the sentence be changed to "A seizure that is provoked by a clearly identifiable factor acting to temporarily lower the seizure threshold does not count towards a diagnosis of epilepsy."

2. As a persistent reader and reviewer of the febrile seizure literature (see for example, Camfield PR, Camfield CS, Scheffer IE, Marini C. Febrile seizures and Genetic Epilepsy with Febrile Seizures Plus (GEFS+) in Epilepsy Syndromes in Infancy, Childhood and Adolescence, eds Bureau et al 2012) I have a number of comments about complex febrile seizures. The definition of two of the risk factors listed is not in line with most of the literature. Recurrence within an illness is incorrect is usually recurrence within 24 hours. The duration of a febrile seizure that defines it as complex is 15 minutes, not 10. The section on multiple risk factors is dominated heavily by one paper by Annegers. It is a good paper but it is important to realize that in that study there were only 1 or 2 patients who actually had all 3 risk factors. The 49% risk of epilepsy is based on statistical modelling, not on observed patient outcomes and was not accompanied by confidence intervals that must be wide enough to allow a large truck to pass. The prestige of the NEJM and the authors should not be sufficient for this analysis to be accepted carte blanche. While these "niceties" do not alter the conclusions of this section, it is important to be correct!

3. "Juvenile myoclonic epilepsy is known to be subject to lifelong elevated risk for seizures" Geithner et al., 2012). Structural brain lesions, such as malformations of cortical development (Rowland et al., 2012), may elevate risk for seizures long-term. 30-40% of people with JME have very long terminal remissions (>10 years) – that is the conclusion of the case series Geithner paper and our own population-based study (Camfield, Neurology 2009). Likewise, many children with structural brain lesions have very long terminal remissions (Dhamija R, Moseley BD, Cascino GD, Wirrell EC.
Epilepsia. 2011). Why should these patients be excluded from the "no longer present" category?

Peter Camfield

16 September, 2013

The proposed epilepsy definitions will probably be helpful since they apply to clinical situations that are fairly common. I still prefer the term "epilepsy in remission" for cases on or off medication that have remained seizure free for prolonged periods and this includes postoperative cases but would excluded syndromes that stop having seizures at a certain stage of maturity. For the childhood disorders known to have limited duration of seizures and where data shows virtually no risk of recurrence, I would suggest that the term "resolved epilepsy" or "transient epilepsy" be considered.

Paul C. Van Ness

16 September, 2013

Thank you for inviting me on comment for Re: ILAE / Definitions of Epilepsy. I read many comments and also good suggestion by renounce person in the field epilepsy. They are commented and good suggestion.

In my view epilepsy term is suitable, because literate and illiterate person, general physician, pediatrician, and other medical practitioner aware this terminology and they know the symptoms of this disorder and suggested treatment for disorder. It's necessary to change and update the symptoms, definition according to use. Previous definition has many UN useful and unclear terminologies and their use is for the view of neurologists, other medical professional and pathologist. So the new definition is better then before and it is know short and clear for education, medical practice, and also understanding.

Dr Kaur Samiksha

Return to top

15 September, 2013

The term provoked seizures should be added to the definition.
How can an ancestor pass on epilepsy when they may not themselves meet the definition for epilepsy?
Should the term no longer present be replaced by resolved epilepsy?
Should emphasize that the revised definition is intended for clinical diagnosis and not all research studies.

Mark D. Bej

15 September, 2013

Congratulations for the very hard work.

Anyway some considerations are due:
The possibility to consider a patient having epilepsy even after only one seizure is what everyone is used to do considering particular aetiologies. However this assumption seems very hazardous, engendering a number of patients (particularly in the vascular syndromes of the elders) treated with AEDs after a first fit.

Defining "No longer present" something that have been cured and not "cured" sounds eccentric.
If the definition is requested for legal interest (driver licence, insurance ...) why do not accord the same possibility to those patients seizure-free since 10 years with a pharmacological treatment?
When we can consider a patient "past the applicable age", one day, one month, one year later?

Best regards
Laura Tassi

15 September, 2013

I thank the taskforce for this significant and important work.

I would like to share that Dr. Wagner and I have discussed the importance of realizing epilepsy is more than seizures frequently as we continue our research in the area of improving outcomes regarding quality of life for youth with epilepsy. We need to continue the discussion about the "spectrum of epilepsy" and its impact on the person with epilepsy and their family whether seizures are present or not. For example, just this week I received a call from a parent in need of documentation for the school as to what epilepsy is because the school wants to withdraw services as the child no longer has seizures post-epilepsy surgery now for 1 1/2 years. This child still has ADHD and other learning issues that are part of the spectrum of epilepsy in spite of being seizure free. A broader definition of epilepsy will increase understanding as to how epilepsy, not just seizures, impacts the person with epilepsy.

Gigi Smith

15 September, 2013

I would like to congratulate the Task Force for doing this wonderful job and thank them for asking the opinions of the epilepsy clinicians and researchers from around the world on some of the basic definitions in epilepsy.

I totally agree with the Task Force that the epilepsy should be considered as a disease and not a disorder for being associated with the long-term (and potentially debilitating) neurological, cognitive, psychological and social consequences.

Two unprovoked seizures:
I do not agree the recommendations of the Task Force as well as the existing concept of the "provoked" vs. "unprovoked" seizures. If 100 patients withdraw alcohol acutely, not all will have seizures, and all those who get withdrawal seizure, may not develop into epilepsy. Similarly, technically speaking, reflex epilepsies are "provoked" seizures but we know that actually, they are not. I think it is more than just a provoking factor or I would say there is some inherent predisposition in these people to develop seizures on withdrawal of some provoking factor or with the reflex stimulus.

I would therefore favor that the concept of "provoked" or "unprovoked" should be removed from the definition of the epilepsy.

High recurrence risk:
I do not agree with the Task Force's recommendation of the concept of 75% cut off risk for the recurrence of seizures after two unprovoked seizures to diagnose epilepsy. The reason is that the 75% cut off is quite vague and can vary widely from an epileptologist working in a metropolitan city like New Delhi or London to a health worker working in the remote village in Uganda or Peru. The magnitude of the risk factors also vary diversely between the regions of the world e.g. infectious disease are quite common cause of epilepsy in South Asia, South America and Africa, which may not be a leading causative factor for epilepsy in Europe and North America.

It would be more uniform and systematic if the Task Force comes up with the some mechanism or grades/scale of the risk factors, which is handy for any health professional working in the field of epilepsy to assess the risk of recurrence and thereby help in diagnosing the epilepsy.

Implications for treatment:
I agree with the author's notion that the diagnosis and the treatment decision are two different issues. If we diagnose a person to have epilepsy after applying the proposed conceptual definition of epilepsy and calculating the risk to be more than 75% (by a uniform grading system), then the issue must be discussed with the patient and the relatives and the treatment should be started based on desires of the patient, risk vs. benefit ratio and the available resources.

Unprovoked seizures separated in time:
I think it is difficult to say as to how much time gap should be present between the two seizures to label it as epilepsy. I think that once a person gets a seizure (whether provoked or unprovoked), he remains at a high risk of getting a second seizure attack.
If one is assessing the risk of the recurrence of a second attack, the concept of time gap does not hold good. For example, if a person comes with a single seizure episode and his risk of getting a second seizure attack is more than 75%, then irrespective of whether he gets a second attack within 24 hours or after that does not matter in diagnosing him as having epilepsy and the subsequent treatment.

Epilepsy no longer present:
The term "no long present" is very dilute and confusing. I think the term "remission" should not be replaced.

The definition of epilepsy, which I think would be appropriate and would be applicable to all parts of the world, is:
Epilepsy is a disease of the brain defined as having "one seizure attack and the risk of recurrence of the second attack is > 75%" (assessed by a uniform grading scale).

Vikas Dhiman

Return to top

14 September, 2013

Congratulations to the ILAE task force for producing this excellent and useful paper. What a mighty job! Thank you very much for offerining me the opportunity to comment. Great idea to open the paper to collegues for discussion.

Here are my comments:

As a basic researcher I find "disorder" more appropriate than disease for epilepsy and I would favour the term antiepileptic drug versus anti-seizure drug

I also think it would be useful including a definition of "epileptic seizure" as in the 2005 report

I welcome the reference to "reflex epilepsy" in the definition.

And, for the future, an animal research definition of epilepsy would be of help.

Best regards,
Carlos de Cabo de la Vega

14 September, 2013

Firstly we would like to congratulate the Task Force for the commendable job they have done. It was a long awaited step in the direction of clarification of epilepsy definition. Although, surely it would work good in patient care, but a few queries we would like to raise here. However, people have already been discussing many things out of these but few points of concern are as follows:

1. This definition is going to give clear defined diagnosis to patients at the basic level of diagnosing epilepsy.

2. This will be of direct help to patients, as the definition can directly be used for patient diagnosis in OPDs (out patient departments) of hospitals.

3. The most important point as reflected by the amended term "no longer present" (Table 1) in the epilepsy definition is surely going to improve the social well-being of patients, and would decrease the associated stigma to some extent.

4. Author himself has stated that the term "no longer present" does not guarantee that epilepsy will not return, therefore other than this, the terms "Long-term remission" or "Prolonged remission" could be of choice, as mentioned by Juan Gomez Alonso. As it is well known that the term remission in itself explains that there exists probability of recurrence in future.

Definition should not be limited to clinical perspective only, but should rather also be available for other areas such as research. As we are researchers so, we would always be more concerned for, whether the definition can be of any direct help in categorizing/ classifying patients for research purpose.

5. Many people have been suggesting regularly, regarding 75% risk prediction criterion for second seizure (2nd item of Table 2). We would also support the fact that list of entities or examples defining the content for "75% recurrence risk" would be needed for general neurologist, epileptologists as well as researchers. List of entities defining "75% recurrence risk" would also be important in order to avoid heterogeneity while enrolling patients for research purpose.

As based on the new definition, patient with single unprovoked seizure and 75% recurrence risk for second seizure shall be kept in "patient with epilepsy" group. But such a patient if was misdiagnosed for second seizure prediction because of unavailability of well defined "75% recurrence risk" conditions or on the basis of different perspective of the clinician towards this, will increase undue heterogeneity in patient pool.

It is also possible that if ones enrolled, such misdiagnosed patient may never get the awaited second seizure, thus increasing the heterogeneity in patient pool. For such a patient researcher will always be doubtful about his epilepsy status and thus inclusion of such patient for the final phenotype genotype correlation will remain a question.

6. In Table 2 item 3rd author has asked for the need of "at least two seizures in a setting of reflex epilepsy". From our point of view waiting for a second seizure in case of reflex epilepsies will always be risking patient lives. The reflexes could range from vary rare to very common i.e. seen in daily routine life and thus can be unavoidable at times. Therefore waiting for a patient to get exposed to such reflexes just in order to have second seizure can be fatal. Therefore in our opinion patient must at least be educated regarding the reflex he had, which actually provoked the seizure.

With regards
Ruchi baghel and Ritushree kukreti

Return to top

14 September, 2013

Definitions have consequences, and parts of the proposed operational definition of epilepsy have the potential to do harm.

"Epilepsy is a disease of the brain defined by any of the following conditions: ...
2. One unprovoked seizure and a probability of further seizures similar to the general recurrence risk after two unprovoked seizures (approximately 75% or more)."

Epilepsy requires an enduring predisposition to seizures, with a certain level of persistent risk (arbitrarily ≥75%).
However, after a first unprovoked seizure and including for those patients with a higher risk of seizure recurrence (e.g. previous stroke or head injury), the likelihood of seizure recurrence does not persist at the same level. Whilst the risk of seizure recurrence is high initially, this risk rapidly decreases with an increasing interval of seizure freedom. Tony Marson and others have already made this critical point.

Using case example 2, a patient with a previous stroke and first seizure, the initial likelihood of future seizure recurrence may be high (perhaps sufficient to be "epilepsy" by the new operational definition), but if this same patient remains seizure free for 1-2 months then their conditional risk of seizure recurrence falls substantially (i.e. "not epilepsy" by the new definition). This patient will only fleetingly fulfill the new operational definition of epilepsy.

If this patient has the misfortune of being seen early, a premature diagnosis of epilepsy may be made. The diagnosis it seems cannot then be unmade without 10 years of seizure freedom off medication.
If a delay in patient assessment of a month or two has occurred and the patient has remained seizure free, then the diagnosis of epilepsy is not made.

"making the clinicians more comfortable in initiating treatment after some unprovoked seizures"
Whilst acknowledging the difference between the diagnosis and treatment of epilepsy, the proposed operational definition supports premature diagnosis and treatment by "making the clinicians more comfortable in initiating treatment after some unprovoked seizures", even based solely on the anecdotal and highly variable experience of individual physicians when evidence is lacking.
Individual patient circumstances may justify treatment after a single unprovoked seizure, but this requires a personalised approach to patient care rather than the premature diagnosis of epilepsy.

These elements of the proposed operational definition of epilepsy will not bring greater clarity to the diagnostic process. Patients will not thank us for making an unnecessary diagnosis of epilepsy that is difficult if not impossible to reverse.

Yours sincerely,
John Dunne and Nicholas Lawn

13 September, 2013

Comments from Sherifa Ahmed Hamed

13 September, 2013

The idea of updating the diagnostic concepts of epilepsy is laudable and comes at a good time where technological tools can better spread information worldwide to the community.

In a general way I must agree with the new definitions reported by the Task Force of the ILAE even though I would like to highlight some main concerns about the presented manuscript:

  • From my point of view, epilepsy should be considered as a disease (not a disorder) since we can identify some criteria such as the etiology, the symptoms and anatomical alterations besides the association with additional long-term characteristics (chronic symptoms) which are usually observed until the patient's death (e.g. sclerosis, epigenetic alterations);
  • The terms "provoked" and "unprovoked" should be better discussed. In my opinion, both terms are obviously in somehow "provoked" by something. Even though seizures may be evoked by a primary insult (e.g. trauma, tumor, infection), they can also be provoked by "subeffective" factors (e.g. epigenetic alterations) which may alter neuronal activity/organization leading to seizures (as a secundary factor but not as an "unprovoked" feature). Additionally, the concept that immediate transient stimuli observed during recurrent reflex seizures are not "unprovoked" is not clear to me.

I would like to congratulate the Task Force of the International League Against Epilepsy for its initiative and excellent work.

Best regards,
Alexandre A. Hoeller

13 September, 2013

I thank the ILAE Task Forces for the opportunity to take into account our views in relation to the new proposed definition of epilepsy. My comments in this regard are the following:

  1. I believe that the term disorder is best suited trying to reconcile the basic and clinical aspects of epilepsy.
  2. Taking into account the clinical and practical aspects, in health centers, physicians handled the term disease in people with epilepsy in order they have a better understanding about their pathology.
  3. With respect to the three possible conditions to assume that a person has epilepsy, I join your proposal due your clear explanation and from my point of view it is valid.

Best regards.
Laura Guadalupe Medina-Ceja

11 September, 2013

Congratulations for this great effort by all the authors to cover in your new definition the things which were not covered before and to include new concepts derived from the clinical practice which has medico legal implications as well.

Dr Majdi Kara

11 September, 2013

I can certainly understand the need to define epilepsy that is "no longer present" due to vocational, social, and other implications. But, is it correct to say that epilepsy is no longer present or just that seizures are no longer present? Many individuals who have seizure remission continue to have neuropsychological deficits, some more subtle but still present, or psychiatric symptoms (anxiety, depression). As contemporary conceptualizations of epilepsy consider it to be a spectrum disorder (Jensen, 2011) and to include these other symptoms and deficits as falling under the umbrella of epilepsy, how can we then say that epilepsy is "no longer present?"

Jensen, FE. 2011. Epilepsia, 52, Suppl 1, 1-6.

Janelle Wagner

Return to top

Read comments received prior to 11 September, 2013

Home | Contact Us | Privacy & Security | Login | Sitemap
Creative Commons License
Text on this website is available under a
Creative Commons Attribution-ShareAlike 4.0 International License
except all videos and images, which remain copyrighted by the International League Against Epilepsy.