ILAE International League Against Epilepsy
 Search search
 Contact Us

Ilae Facebook Ilae Twitter

Definition of Epilepsy


The following comments were received on or prior to 10 September, 2013.

who will post them here on a weekly basis for the entire community to consider.


10 September, 2013

For thousands of years, Epilepsy had been describe for different cultures, without being able to be interpreted properly the complexity of its manifestations and therefore its management, being called by Hippocrates "The Sacred Disease".

This has been proven in practice and there is still no consensus on the classification and definition of Epilepsy. This is consequence of the complexity of it, to the multiple etiologies and symptoms that characterize it.

In my opinion, we are now able to conceptualize Epilepsy, in a more comprehensive, when Jackson described him as a intermittent disorder, which for its time was an excellent description. Anyway, I do not think we should overshadow the issue and come to an agreement.

I believe that the review of the Task Force that ILAE commissioned is very adequate, but with due respect, I must make some comments. Epilepsy has traditionally been considered a brain disorder rather than a disease and we must consider it as such. It is also true that we must take into account that in clinical practice patients can present with unprovoked seizures and the tendency to recurrence of crisis, which recall the main characteristic of the disease. This may in some patients be a high risk of recurrence of seizures. All this has implications for the decision to impose or not treatment with antiepileptic drugs.

I am not agree to include in the operational clinical definition the seizures in a setting of reflex epilepsy, as there is more than one type of provoked crisis and a topic of the disease.

I propose to consider the exclusion of item 3 related to recurrent reflex seizures, to summarize the definition of Epilepsy more generally, encompassing its features and update the glossary of epilepsy terms, which would include different clinical considerations of this disease.

In my humble opinion, Epilepsy should be considered as chronic disease, resulting in dysfunctional brain disorder, characterized by recurrent seizures and related to multiple etiologies.

Juan E. Bender del Busto

10 September, 2013

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

I agree with the proposed definition.

Best regards,
Nina Ventura

10 September, 2013

Congratulations to the ILAE task force for this excellent and helpful document. From a basic scientist point of view "disorder" seems more appropriate than disease and anti-seizure drug is unusual, antiepileptic drug is more common.

Best regards
Gabrielle Rudolf

10 September, 2013

I much appreciate the operational clinical definition of epilepsy by ILAE Task Force and I find the opportunity to comment very useful. Here are some comments:

- The term "Disease" implies the concept of longstanding predisposition to have seizures. However the term "Disorder" seems to be closer to the paroxysmal characteristics and to the fact that a structural pathological brain condition is not always associated. Furthermore I personally feel that the transition to the term "Disease" can lead to social discrimination and aggravation of the stigma.

- The "High recurrence risk" is a bit still unclear in my opinion. I understand the effort to help with the quite frequent circumstances when people have a single unprovoked seizure associated with, for instance, symptomatic lesions generating an enduring predisposition for unprovoked seizures. However, from my limited experience as clinician, it appears really difficult to quantify the 75% chance for another lifetime seizure. I believe that further epidemiological studies are essential, aiming to estimate the risk for a second unprovoked seizure associated with a number of different conditions.

- According to the new operational definition, epilepsy is considered to be "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 and without any known risk factors associated with a high probability (>75%) of future seizures. Considering "remission" according to MeSH definition as a diminution or abatement of a disease over time, spontaneous or inducted by treatment, I would prefer it rather than "no longer present". In my opinion, "remission" is also much more easily understood and recognised by public.

- "Unprovoked seizures separated in time". I believe that an outer time limit for occurrence of the second unprovoked seizure should be estimated or, at least, the circumstances which clearly exclude the condition of oligoepilepsy (such as different causes for unprovoked seizures at different ages) should be better established.

Thank you for your really helpful work!

Best regards
Leandro Provinciali

10 September, 2013

I appreciate the possibility to comment the proposal. It seems that there still is a fundamental problem, the definition of an epileptic seizure. The old definition, as that written in the 2005 report, states: "An epileptic seizure is a transient occurrence of signs and/or symptoms due to abnormal excessive or synchronous neuronal activity in the brain."

A definition is a statement that explains the meaning of a term. It should to set out the essence of something, specify the necessary and sufficient conditions for a thing being a member of a specific set. According to the old definition, a transient occurrence is an epileptic seizure if there is abnormal neuronal activity in the brain and the occurrence is caused by this activity. In clinical practice there rarely is an opportunity to observe neuronal activity during a seizure. Therefore, according to the old definition, a transient occurrence can rarely be classified as an epileptic seizure.

The other included necessary condition is that the transient occurrence is due to abnormal neuronal activity. Because of this condition, the theory that epileptic seizures are caused by abnormal neuronal activity, is trivial circular reasoning. No enlightment is offered about the pathophysiology of these seizures.

Epileptic seizure is a name for a heterogenous entity. Any meaningful and comprehensive definition would be long and complex and consequently futile in clinical practice. Instead of such definition, ILAE should write a practical operational guideline, preferably an algorithm, for the differential diagnosis of all (not only epileptic) seizures.

".. the problem .. is not to give seizures a name, for any name would do, but to find their cause and if possible, to predict the future and prescribe treatment." Stuart Renfrew 1967

A.H. Bardy


  • Anonymous. Definition. Wikipedia, 27.8.2013
  • Renfrew S. An introduction to diagnostic neurology. Livingstone, Edinburgh and London 1967
  • Bardy AH. Problems in classification of epileptic seizures and syndromes. Medical Hypotheses 1988; 27:141-144

10 September, 2013

On behalf of the Norwegian Institute of Public Health we thank the task force for their excellent work in making a revised operational definition of epilepsy and for the possibility to comment on this important work. A clear and operational definition is important both in epidemiological research and in clinical work. We have some comments to the proposed definition:

- The 75% recurrence risk criterion may be difficult to apply in clinical practice. The ability to determine recurrence risk will depend on the knowledge of the neurologist or pediatrician treating the patient, and it will also be related to the knowledge about recurrence risk existing at the time of diagnosis. This knowledge will change over time, which may generate spurious differences between epidemiological studies conducted at different points in time.

- The "epilepsy no longer present for individuals who had an age-dependent epilepsy syndrome but are now past the applicable age" statement implicates that the epilepsy has to be classified into an epileptic syndrome. This is not always possible in epidemiological research due to lack of detailed information about the epilepsy.

- Ten years seizure-free off medication before epilepsy is considered no longer present might be too long in some patient groups, for example after epilepsy surgery with total resection of an localized benign tumor.

Kari Modalsli Aaberg

Return to top

9 September, 2013

Thank you for asking me for my opinion on the new Operational Clinical Definition of Epilepsy.

I am a peadiatric neurologist and not an epileptologist. So, please consider my opinion in this context.

Here is my Opinion:

The term "disorder" plays a significant role for the physicians but i don`t think for the public. As soon as an epilepsy is diagnosed patients/parents take it very serious independent of its underlying cause. It is not the epilesy by itself which is an enduring predisposing factor but the underlying disease. Therefore it seems for me as a physician confusing to consider epilepsy all of a sudden as a disease and no longer as a symptome of various underlying diseases.

Regarding Table 2 Item 2:
It is a subjective decision of a physician depending on the underlying disease and on his/her experience to start an anti-seizure medication or not after a single seizure occurence. It is however very difficult to draw a precise red line of 75% or more. (as you know drawing red lines are dangerous- just for fun).

Diagnosing epilepsy and initiating treatment are for me not very apart from each other. If I decide to tell a parent that his child has epilepsy, then I have to do some thing against it. If I am not convinced of the diagnosis, then I won`t tell some one that he has epilepsy.

As mentioned from the Task Force, there are so many disadvantages for some one to be diagnosed as having epilepsy. If it has no consequence of treating the epilepsy, then why complicating the life of that person by saying you have epilepsy?

Finally the phrase "no longer present" suggests that the risk of recurrence is not more than that of the normal population. In my opinion this can lead some one into a false sense of security. What about looking for other alternative expressions like "seizure recurrence unprobable"?

With best regards,
Yohannes Menedo

9 September, 2013

The task force realized a significant document. While different innovations of the proposed definition can be shared other controverse points should be clarified.

-The definition of epilepsy as a "disease" and not as "disorder" is not convincing. On my opinion, the term "disorder" is a mirror of the heterogeneous presentation of epilepsy. The term "disease" is better if it is applied to the single presentations. The examples of cancer and diabetes are not pregnant. Nobody thinks that leukaemia (a blood cancer) and lung cancer are the same disease. In the same context, Ohtahara syndrome and Absence Epilepsy are not the same disease. The term "cancer" refers to a "common pathogenesis" to different diseases. So, in this context cancer can be considered as a "disorder" of cell's growth. Diabetes can be also considered as a disorder of glucose and insulin metabolism because its variable presentations and its complications (retinopathy, neuropathy etc) are related to common pathogenesis (increased glucose levels and decreased insulin synthesis). The sentence "disorder implies a functional disturbance, not necessarily lasting; whereas, a disease may (but not always) convey a more lasting derangement of normal function" seems redundant (the duration of "diseases" and "disorders" cannot be objective criteria). The choice to use the term "disease" because is poorly understood by the public and minimizes the serious nature of epilepsy" cannot be shared. I think that it is more precise to outline the multifaceted aspects of epilepsy with the term "disorder" because it imply a better definition of the different clinical evolutions and outcomes of the different forms of epilepsy (i.e. BECTS and Dravet syndrome) .

-I agree to define epilepsy a first unprovoked seizure in a patient with a recurrence risk comparable to subjects who had experienced two unprovoked seizures (for example, in the clinical practice a child with a known syndrome including epilepsy such as Angelman syndrome is usually considered to have epilepsy since the first seizure). I think that some attempts to define possible the risk factors that could permit a diagnosis of epilepsy since the first seizure could be made. I’ ma conscious that this would require more discussions among the Task Force but I think that some well established criteria would be more useful as diagnostic tool than a mere definition of a recurrence risk higher than 75% (it is very difficult in the clinical practice to estimate in the single patient if the recurrence risk is higher than a precise threshold risk number after a first unprovoked seizure). I would suggest as candidate criteria: 1) the presence of diagnostic criteria for known epileptic encephalopathies (we diagnose West syndrome even if we observe a single episode of epileptic spasms if developmental delay and hypsarrhytmia are present) or genetic syndromes including epilepsy (i.e. Angelman syndrome); 2) the presence of a symptomatic lesion that increase the risk for further epileptic seizures;3) a single unprovoked and prolonged status epilepticus (?). However, it is true that further epidemiological data are required to better define those criteria.

- I have some doubts on the position of the Task Force on complex febrile seizures. In the report the recurrence risk of further unprovoked seizures after the first episode is indicated as lower than 75% but the references of the indicated epidemiological data are related to studies from 1987-1996. More updated epidemiological studies are necessary to confirm this conclusion.

Different authors consider febrile seizure as benign epileptic manifestations in accordance with the comparable pathogenesis and the possibility of recurrences. I think that this aspect should be considered, at least for recurring complex febrile seizures. I think that some febrile seizures could be treated by the Task Force like reflex epilepsies (diseases due to provoked seizures but characterized by an enduring predisposition to seizures). About the part on reflex epilepsies, when predisposition to seizures can be considered as normal?

-I'm fully agree with the proposed criteria to define epilepsy as "no longer present".

Mario Mastrangelo

Return to top

9 September, 2013

The task force did an incredible job to establish this operational clinical definition of epilepsy; Thank you so much!

In general, the attempt of providing a general definition of epilepsy including the contemporary state of knowledge is important for indivual patients as well as for scientific issues.

There are a few thoughts to three already extensively debated topics:

1."remission" vs. "no longer present"
Personally, I prefer the term "remission" because it is an accepted medical term in other fields of medicine and it allows to keep open the very small probability that someone having had epilepsy and being seizure-free for several years will relapse despite good outcome predictors. Also, proving of "non-presence (absence?)" might be very difficult in medicine. Thus, if the Committee persists on "no longer .." I would much more like the wording of Mario de Curtis saying "no longer (clinically) active" or, as some contributors proposed, to remain on "remission" and "cure", the latter describing seizure freedom after 5 (according to the oncology field) or 10 years.

2. ≥75% recurrence risk
The ≥75% margin for seizure recurrence risk does not seem to be based on large evidence. In view of substantial uncertainty regarding seizure recurrence risks in various clinical conditions and to allow for more clinical judgement in individual patients, I would feel more comfortable with a seizure recurrence margin of 50 to 75%.

3. disorder vs. disease
This issue seems very complex to me, because it is ambiguous and cannot be easily resolved: If the ILAE would like to endorse the notion that epilepsy is no longer a disorder but a disease, it hopes that dealing with a disease would facilitate political and financial support for care, research and public attention of epilepsy. However, many patients then will feel more stigmatized when having a disease instead of a disorder. Conversely, epilepsy remaining "only" a disorder would let a lot of patients feel much better ("I am fine, it's only a disorder, I am not ill ..."), but, at the same time, the notion of epilepsy being "only" a disorder might negatively impact efforts to make it clear to the public that epilepsy can be a very serious condition that needs substantial support in the fields of medical and psychosocial care, and of research.

4. One practical question: regarding repetitive acute symptomatic seizures (e.g. resulting from severe hyponatremia) over a few days and remitting after two weeks: has this patient epilepsy?

Thank you again,
Stephan Rüegg

9 September, 2013

This is a commendable effort. My appreciation for the excellent task done.

I have the following comments:

1. It would be a useful guide for the general physician if some common conditions, which lead to risk of recurrence of 75% or more, are included. In developing countries due to a limited number of trained neurologists and large number of people with epilepsy, care is provided by physicians.

2. I prefer the term 'disorder' since it is more acceptable to the patient and family rather than 'disease' in view of the implication in arranged marriages, driving license etc. which have already been stated.

3. The term 'no longer present' may be useful for legal purpose but patients and their families seem to prefer the term 'controlled'. Epilepsy may no longer be present but what about the consequences such as behavior disorder, cognitive impairment etc. Would a person of epilepsy who is enjoying social benefits, be deprived if seizures are no longer present but the consequences persist.

4. It will be useful if the minimum seizure-free period is defined for tapering and subsequent cessation of antiepileptic drugs.


9 September, 2013

I have stated my view on Epilepsy in my recent book, Alzheimer's Disease: What Is It After All? in which I point out that epilepsy can lead to dementia through frequent seizure attacks and that aura in many forms are signs leading eventually to Dementia Epileptica. If you want to know more than the relationship between epilepsy and dementia, please order a copy of my book and you will find out the details. For you infomation, I an forwarding to you a flyer to that effect.
Fred C. C. Peng

Return to top

8 September, 2013

Below please find my comments on the proposed changes in the ILAE definition of epilepsy:

1) Epilepsy is to be defined as a "disease" rather than a "disorder". I think this will cause confusion in the minds of the public, who tend to think of "diseases" as infectious. I don't favor this change.

What I would favor is replacing the word "epilepsy" with the term "seizure disorder". I think that would greatly reduce stigma.

2) The traditional definition of at least two unprovoked seizures 24 hours apart would be maintained, but supplemented with the definition of "one unprovoked seizure and the 75% probability of further seizures". I think the new definition of epilepsy would be understood by epileptologists, but would cause a lot of confusion among the general practitioners and pediatricians who unfortunately do a lot of the initial diagnosing and treatment of epilepsy. I think it could lead to a lot of people being diagnosed as "epileptic" and put on anticonvulsants when they don't actually have epilepsy. I don't favor this change.

3) Epilepsy would be diagnosed after at least two seizures in a setting of reflex epilepsy. This seem sensible. I favor this change.

4) Epilepsy would be "no longer present" for people who have been seizure free off medications for 10 years. This also seems sensible. I favor this change.

I am,
Sincerely yours,
W. McIntyre Burnham

6 September, 2013

Thank you for your efforts in dealing with this difficult task. I do have few comments:

The definition of 'Epilepsy' was always a 'problem'! And the epileptic patients are really in a dilemma seeing their doctors NOT ABLE to agree with each other for a certain 'definition' that lasts for more than a while!

1. Disease or Disorder?
The word 'Disorder' may be preferred for the following reasons:
a- Seizures or even 'epilepsy' is sometimes a part of a known disease!
b- Sometimes it is a syndrome that is age-dependent which might remit or no longer exist or say 'cure' one day!
c- It is hidden or 'not apparent' unlike 'Diabetes' or 'Cancer' for example. The patient of 'Epilepsy' might be one of your 'Task Force Members'! One may claim that 'Diabetes' is not apparent also, yes may be not seen BUT it is a 'running' disease that keeps destroying its patients every hour and minute!, WHILE 'Epileptic patients' might be gotten 'seizure free' for months or years without destroying the patients all the time. So the patient may be 'seizure-free BUT not 'diabetes-free!
d- A 'Disorder' is a 'promising word' whereas a 'Disease' is a 'depressing one'!

2- 'Epilepsy No Longer Present': this new term simply means that 'Epilepsy is not present now or is 'Absent'! And the patient had no epilepsy for the time being, well why not called 'CURE' which may be preferred for the following reasons:
a- It may give a real and genuine 'hope' for patients to hear about cure and about possibility of having one day a 'normal life'
b- If the patient got seizures out of age for 'age-dependent-syndromes' or after the 10 years seizure freedom 'off-treatment', this may be a new type of 'Epilepsy' that any person of the non-epileptic population may have!

3- Regarding reflex epilepsies or those triggered by 'light' or else, I completely agree with the new concept which is a very useful addition to the previous concepts and which was a 'withdraw' in the previous definition

4- 'Two seizures' or '1st seizure +high risk':
The old definition of 'two unprovoked seizures' that are more than 24 hours apart of each other was a very 'immune' definition! It usually disregards the 1st seizure for both the diagnosis and probably the treatment. This disregard is really a 'Regard' for the non-expected and usually non-well described 1st event whether by patients or by witnesses or both i.e. (the usually cloudy picture of the 1st event). It is always an 'Event' BUT not necessarily always a 'Seizure'!

Note that I am not speaking about the 'very few' situations where the 1st seizure is well documented or say photographed or seen by an 'Epileptologist'! Not even by a 'General Neurologist'!
The 1st seizure+ high risk concept is going towards over-diagnosing 'epilepsy' and over social stigmatizing besides over-side effecting etc. with probable 'non real existence' of 'Epilepsy'! The risk of over 75% would be not easy to predict and may open the door widely for 'false diagnoses'
I do suggest the existence of the old 'two unprovoked' for a better confirmation of 'Epilepsy Diagnosis' especially when we realize that the ILAE is concerned with people all over the world and not merely 'the epileptologists'. People especially non-neurologic population might be very unaware of the alphabetical concepts of 'Epilepsy'.
I wonder what is the value of 'proper diagnosis' for 10 new epileptics by 'tertiary care centers' through the new concept of '1st seizure +high risk' against improper or wrong diagnoses of 'one million' people as 'epileptics' in developing countries or non-tertiary centers through the same concept!

5- Last but not the least is the role of EEG:
Where is the place of EEG in the new concept of 'Epilepsy diagnosis'?
Is it omitted? Is EEG no longer useful to lighten a seemingly dark situation?
Or is it only confined to the diagnosis of 'Electro-Clinical Syndromes'
I believe that without an abnormal EEG there will be 'no convulsions', 'no spells', 'no electro-clinical syndromes', 'no status' etc., simply because all these semiologies are but a clinical presentation of this abnormal electrical activity! A difficulty to illicit this abnormal EEG is not the fault of EEG. It is ours! And I think that a better role for EEG should be awarded today or in future for the proper diagnosis of 'Epilepsy'.

Best regards,
Ghaieb Bashar Aljandeel

6 September, 2013

Like all others, I wish to thank the task force members for the performance of presenting definition issues and for putting together such a quality in the material provided for definition revision.

I wish to take the opportunity to open a discussion with the panel by attracting their attention towards clinical trials. Irrespective of their proposed indication, when a potential risk of seizure induction (or epilepsy) was detected at the non-clinical level for molecules entering clinical development, based on the definition proposed in the document, what category of patients should be excluded from clinical trials?

You will find my comments and suggestions in this document, but I still try to list them below in case you and the panel think some can be posted on the web page

Abstract P3.
I would suggest having some hints about how the 75% probability of recurrent seizure are calculated as early as in the abstract.

Table 2, P5.
In point 2, I would suggest to clarify that there is a medical context that could be responsible for probability of further seizures.

"2. One unprovoked seizure and a medical context with a probability of further seizures similar to the general recurrence risk after two unprovoked seizures (approximately 75% or more)."

Page 5, text.
For the paragraph about clarification of the "Two unprovoked seizures", could it be relevant to add "without identifiable anatomophysiological causes"?

Page 7, implications for treatment.
The proposition that a treatment may be distinct from a diagnosis, given the labeling of AEDs, may sound odd. This may require changes in all products' labels to avoid off-label prescriptions. Indeed, if an anti-epileptic drug is given to treat and/or prevent seizures, how can the underlying condition not be epilepsy? Should it be called "seizure-prone medical condition"?

Page 9, 3rd paragraph.
Risk factors. My question would be to know what would be the recommendation for diseases such as Alzheimer's in which the risk of seizures is believed to be higher than in matched controls? Is the definition of epilepsy applicable in case of seizures in AD patients?

Page 10. Consequences of the operational definitions.
The statement around treatment that might be justified in some patients for whom a definitive diagnosis of epilepsy has not been made may open discussions. This is a difficult concept. How, ethically, can an anti-epileptic drug with side effects be given to treat a patient for a disease the definition of which is not met in that patient? Although this is the case for a prophylactic treatment, what would, here, be the prophylaxis when trying to avoid a disease that the patient is not supposed to suffer from, as per definition?

Once again, thanks a lot for giving the opportunity to review and discuss this very important piece of work.

Best regards.
Hervé Bester

Return to top

6 September, 2013

I'm grateful for the new proposal of the task force, and consider it useful for clinical purposes.

Lidia Cáceres

5 September, 2013

The proposal for epilepsy definition has 4 major points as follows:

1-It is a disease. It is important step from the scientific community to the society in general. It makes clear the relevance of epilepsy and why it should be treated, whereas, when we say that epilepsy is a condition, people argued and question why they even need to treat?! So, this new approach of saying that epilepsy is a disease should increase seriousness and decrease the treatment gap. This is a plus. Some may argue that it would increase associated stigma, but let say that the word "condition" has not alleviate the stigma so far.

2- epilepsy is no longer present. My view is that we should adopt the same instance as the item one that epilepsy is a disease. The reason is that when we deal with a disease that has no cure, on one hand the physicians tends to adopt nihilist stance and do the minimum, no investment on this, on other had, the patients would say, well, why need to go all the way as there is no cure. But, if we signal that there is a cure, even that is in a small percentage (which is true), it fuels our inner feeling of hope, and where there is hope there investment. I would encourage the adoption of word cure. Analogies of success is seen in other diseases, take for instance cancer, which some institution write cancer with a strike through the word to make clear what they meant.

3-one seizure is epilepsy if there is underlying lesion that raises the probability of second seizure. Yes, it is fine.

4-Two reflex epilepsy seizures is epilepsy because the underlying proneness for seizure is there. This may bring some discussion about this proneness, which might encompass febrile seizures, seizure in alcoholism, drug addition, because all have a proneness. Let's review this, as it may increase lot to questionable treatments.

Li Li Min

5 September, 2013

My comments are as follows:

1. I appreciate the switch of epilepsy from a disorder to a disease. It is correct both scientifically and also socially. The new definition is hassle free and clinician friendly.

2. I apprehend over diagnosis and treatment of epilepsy with this new definition. Though it is clarified that treatment decision is different from diagnosis, still there is a very thin line between diagnosis and treatment. Probably, it is unavoidable.

3. The duration of 1st seizure may have a role in defining epilepsy. We found that patients with prolonged seizure more than 5 minutes usually develop subsequent unprovoked seizures. This we found more in relation to cerebral malaria.

With regards, yrs sincerely,
Manoj Mohapatra

5 September, 2013

Thank you for this great and helpful work. I have the following comments:

- Epilepsy definition as well as other pathologies (i.e. Multiple Sclerosis) needs to have clear-cut clinical, electrical and anatomical (if possible genetic) criteria in a given period of time and to be less colloquial.

- To precise the role of EEG in Epilepsy definition and follow-up, in general and in particular cases (i.e. diffuse spikes in EEG recording without clinical or epileptic seizure recorded).

Best regards
Suela Dylgjeri

5 September, 2013

I propose you my comment - I'm sorry, I lose differents words and also my english is not great

"It's very important to make an Individual diagnosis because my epilepsy has been diagnosed since the divorce from my parents
The explication on the page thirtheen - is exactly my situation - after differents clinical examinations - I have medications for 15 years but also seizures and differents risk of the medications
Also I decide to STOP the medications and also I haven't just ONE or TWO seizure at ONE month - That was not possible with the medications and I have twenty seizures with three different medications and also one COMA with the medication

We haven't many example without medication -also I propose I give you my different evolution of the next 10 years without medication

I work also AUTISM since 1987 - it's the similar PROBLEM and the DEFINITION for EPILEPSY it's very important to look that it's not a DEFINITION of mental disease because I work with mental disease and many epileptic people are confused with them

The classification of OMS is very important because EPILEPSY is different at the different countries in the world

Thanks for all your work
FLEURY Corinne

Return to top

4 September, 2013

Thanks for giving me a chance to go through the new definition and inviting my comments. I am glad that many of the ambiguities in the definition of epilepsy have been discussed.

  • The social stigma of epilepsy is a very important factor at least in developing countries like India and needs to be addressed seriously for young females who are not yet married.
  • The duration for "no longer epilepsy" could have been shortened to seven years. You can take any cut off point but the risk of recurrence even after that will always be there whether it is 10 years or 15 years. If we want to be practical and reduce the stigma and other limitations imposed on epilepsy patients we have to draw a line and I suggest "seven years" for this purpose.

S K Jha

4 September, 2013

First of all, I want to congratulate the authors for their outstanding and important work. It is a really useful work the task force has done and was absolutely necessary.

I fully agree on the classical definition of epilepsy as the occurrence of at least two unprovoked seizures underlying that they must occur 24 hours apart; I am wondering about the usefulness of choosing a bound for the longest interval because there are no definite data supporting a cut point. On the other side, my personal opinion is that the 10-year period of time in order to consider an epilepsy "no longer present" is very long; it is well known that there are some types of epilepsy (e.g. Benign childhood epilepsy with centrotemporal spikes, Panayiotopoulos syndrome, childhood absence epilepsy) that show complete remission in few years, in the majority of patients. I suggest to reduce the 10-year period and to suggest to consider that some factors (e.g. types of epilepsy, etiology, neurological deficits, family history) have a more important role in defining the remission.

Alberto Verrotti

4 September, 2013

I'm not an epileptologist but a neuroradiologist with interest in epilepsy.
I agree on their definition of both epilepsy & seizure.

Emiko Morimoto

4 September, 2013

In my opinion, based on what I have learned from multiple resources, including people themselves, and what I've been through due to having epilepsy myself, stigma is not the biggest problem pertaining to epilepsy; it's the 'lack of communication'. For example, with the different medical facilities I've been to, I took the time to visit the neurological sections to ask the medical personnel where they had information pertaining to seizures and epilepsy. Everyone's response was that the seizures and epilepsy handouts, leaflets, pamphlets, etc. were stored in a closet, an office desk, a filing cabinet, or behind a counter. The information was available in the medical facilities, it was just not meant to be easily seen for public viewing. To me, this is part of the 'lack of communication'.

As for the previous comment I made about whether to call epilepsy a 'disorder' or a 'disease' (August 7, 2013), maybe a third choice could be used instead. How about defining epilepsy as a 'neurological condition' based upon an era of recurring seizures. Creating a definition short and simple like this can be easy to remember, and it can be the PRIMARY definition of epilepsy. If a patient has more detailed signs or symptoms then make their diagnosis for a specific type of epilepsy (...symptomatic epilepsy, idiopathic epilepsy, temporal lobe epilepsy, catamenial epilepsy, reflex epilepsy, etc).

Bruce Moreau

3 September, 2013

This excellent ILAE report about a revised definition of epilepsy is fundamental for epileptologists to use the terms accordingly in their daily practice.

As to implications for treatment, a clinical approach during acute symptomatic seizures is crucial to recognize the etiology and the possibility to start specific treatment to avoid future epilepsy. On the other hand, according to our opinion it is also crucial to recognize the epileptic syndrome to evaluate whether to treat or not to treat (for example: benign infantile seizures, Panayiotopoulos syndrome, etc). Considering reflex epilepsy, by definition the seizures are stimulus sensitive, however,  some cases may also have spontaneous seizures. Thus, these patients may also be considered to have reflex epilepsy. Finally, we would like to emphasize that the epileptic syndrome, type of seizure, and etiology should be always considered when we analyze all these operational clinical definitions of epilepsy.

Best regards
Roberto Caraballo

Return to top

3 September, 2013

Congratulations to the Commission for this document with an improved definition of epilepsy. The opportunity to comment is greatly appreciated.

I believe that the use of the term "disease" instead of "disorder" would have a greater impact on the perception of epilepsy in general population.

I prefer the term "epilepsy in remission" rather than no longer present.

The 'at least 75% risk' to have a second seizure may be difficult to determine. This definition shall not lead to adverse consequences for the patient's health and care.

Céline Dubé

3 September, 2013

Thank you for the opportunity to comment on such an important topic and congratulations to the Commission for the hard work.

I agree with most of the definitions and I am sure this will help unifying the clinicians and researchers work. Please find below my comments:

  • I agree with the concept that "unprovoked" seizure does not mean that a seizure had not a provocative factor but rather that this was unknown.
  • I think it is very important to consider as epilepsy the "reflex seizures"as the tendency to respond to such stimuli is often life-long and the presence of some stimuli is sometimes impossible to control
  • I find the term 'at least 75% risk' difficult to apply not only for general neurologist but also for epileptologist as epidemiological studies can be controversial. I suggest to stress more the concept that these percentages are not strict
  • I think that the term "no longer present" is confusing on a medical ground. I'd still use the term "remission" or "cured" depending on cases (i.e. age related remission; cured by drugs or surgery etc)
  • An outer time limit for the second unprovoked seizures is not specified. However as after 10 years without seizure and off medication, by definition, the condition would be considered "not longer epilepsy", I think that a time span of roughly about 20 years would be appropriate to consider two seizures as two different events. Most of the numbers and percentage are indicative anyway.
  • I think it would be useful to introduce the term "possibly" and "probably" especially for those conditions that are uncertain or in case the information are not sufficient to classify it.

Kind regards
Antonietta Coppola

2 September, 2013

I completely agree that "disorder" is not understandable by the public and in fact epilepsy is considered as a disease. Since I am not a MD, I do not comment on the clinical and neurological aspects of the document, but overall it seems a clarifying proposal .

I congratulate the task force for the nice elaboration.

Thank you for allowing me to comment.
António Dourado

Return to top

2 September, 2013

First of all I want to congratulate the commission with the task accomplished; having been there "during an earlier life" I know what they have gone through to get to this point.

Secondly I applaud the idea of having an operational definition next to — not instead of — the conceptual one.

The latter made many people unhappy, notably the clinicians who have to deal with epilepsy in day-to-day practice and the epidemiologist who felt that their temple was being destroyed.

Both will, in majority, be happier with this new operational definition although, from the comments already provided by many esteemed colleagues, it is clear that this new definition, like the previous one, is not covering all bases.

This, in fact, does not matter much: epilepsy (or the epilepsies) is/are a matter of nature and nature does not care much about definitions: people do and definitions change as people and their needs change over time: for most (not all) practical purposes the operational definition will be helpful for some time to come and then it will be changed and improved again.

Some comments nevertheless, not original since they have been voiced by others as well but to give my support to these voices and add mine.

Disease vs. Disorder
I strongly favour "disorder" or "condition" rather than "disease". Even when acknowledging the "enduring condition" of the conceptual definition we must admit that we still do not know the full biological background of that condition nor whether this is (unlikely) a single cause, (more likely) a multifactorial one or that (quite probable) multiple causes can result in this enduring condition.

The comment that "disorder" or "condition" is more difficult to explain to patients is irrelevant and incorrect: it is the task of the doctor to explain and if he or she does that well neither term is more difficult than the other. On the other hand, even if some diseases are multifactorial in principle the system is "one disease – one cause"; contrary to what is implied in the text most doctors would not consider "cancer" a unified disease and even patients are aware that Cancer is a condition, rather than a disease and that they need to have a specification regarding type, treatment and prognosis.

Moreover, most patients, including at least one who has made that comment on the website, prefer to have a condition, rather than a disease, in line with what we have been teaching and preaching for decades.

Numbers & oddities
I agree with those that for practical purposes 50% probably is more realistic than 75%, that ESES/LKS etc. still needs to have a place and that 10 yrs may be theoretically correct but difficult to work with in real life. These are relatively minor issues, however, that can be dealt with in the final version.

Remission vs. No longer present
More importantly and like some other colleagues I want, in the end, to point out the intrinsic contradiction in this concept for a operational definition: if we accept the enduring condition as the conceptual truth underlying epilepsy and if even the authors agree that following even a single unprovoked seizure and thus certainly following a period of active epilepsy the risk for occurrence (or recurrence) of seizures is always greater (even if often far lower that 75 or even 50%) than in a person who never had a seizure, we must acknowledge that epilepsy may for all practical purposes have become dormant and need not to be treated and can even be considered "gone" for practical purposes such as driving, but probably never is either cured or "no longer present".

"Remission" is the correct term and if one feels that that is helpful one can add: "no longer active" but "no longer present" is both incorrect and scientifically unproven and highly unlikely.

Apart from these comments, again my compliments. Whereas classifications and definitions will continue to come and go, clinical, scientific and psychosocial epileptology will continue to be an ever changing and dynamic, fascinating, challenging and rewarding field. Thanks for yet another important contribution and for offering me the opportunity to give my comments.

Walter van Emde Boas

Return to top

2 September, 2013

This document is the result of a tremendous synergism to update the definition of epilepsy.

As a non-clinician I found the new definition to be generally straightforward. This being said, it remains somehow unclear how such a definition could be applied. One concern is the 75% seizure threshold used in cases of an identifiable pathological event (e.g., stroke, trauma) triggering the first seizure: such a threshold may be susceptible to variations (regional, dissimilar infrastructure/facilities, different operators/neurologists/nurses). It is unclear whether the severity of precipitating pathological events (stroke, TBI ...) will be taken into account. Confusion exists between the old and the new terminology (disorder and disease). Quoting the abstract: "Epilepsy was defined conceptually in 2005 as a disorder of the brain characterized ... The task force proposes that epilepsy be considered to be a disease of the brain defined...". As a personal opinion I found the term disorder to be less stigmatizing than the term disease. The word condition also appears throughout the text, possibly complicating the picture. Finally, an improved set of definitions could aid basic researchers to tailor existing models to appropriate experimental end-points.

Nicola Marchi

2 September, 2013

Thank you for your emphasis on my opion on the definition of epilepsy!

Epilepsy is a kind of disorder manifests as cerebral disfunction attack due to neuron abnormally discharge in synchronization.

Hao Yong

2 September, 2013

After listening to numerous people - professional and lay, I have only one comment.

Do not change what is not broken: changing the name of a proposition (alone) does not change the truth-value of that proposition. If any change in name does not result in increased truth about that proposition, to whose benefit is it to change what is entrenched in previous research / literature / teaching material? Change for the sake of change will only confuse all concerned.

Daniel G

2 September, 2013

Dear members of the Task Force,

There is no doubt that we need a operational definition on the diagnosis epilepsy. In the daily practice we identify patients with a first seizure and abnormalities in MRI and/or EEG and it should be allowed to diagnose the first manifestation of an epilepsy by this first seizure in this circumstances. In the presented paper the argumentation seems a little bit confusing:

"It is important to note that a single seizure plus a lesion or a single seizure plus epileptiform EEG spikes does not automatically satisfy criteria for this operational definition of epilepsy. Data must be available to support an approximate 75% risk or more for another lifetime seizure."

How can we calculate the recurrence risk in daily live?

Kind regards

Hartmut Baier

Return to top

2 September, 2013

What a brave constructive enterprise!

With great interest I have read the formulation of an operational definition of epilepsy for purposes of clinical diagnosis, as well as the commentaries that addressed clearly many valuable issues.

I like the idea of thinking into risk of seizure recurrence and remember our productive discussions in Brussels in order to loosen the rigid rules on European driving license regulations. Late Thuur Sonnen was instrumental in this concept and it has had very positive outcomes.

Like Vermeil, Huberfield, Wang, Chen and Beghi, I propose to stop in this new definition to make a distinction between provoked and unprovoked seizures. Similarly there is no need to make reflex epilepsy a different category, especially since you have introduce "risks of recurrence" and "no longer present". You mention yourself that "unprovoked" is an imprecise term, which I of course fully agree with. In our current world – regardless whether so called developed or not- potentially provocative internal and external stimuli are much more abundant and often concealed; it therefore has become much harder to recognize whether a seizure was directly or indirectly provoked or not. Why would it be different if the second seizure is provoked? Although re-occurrence of a febrile seizure cannot be considered sufficient for the diagnosis of epilepsy, in most other cases (reflex seizures) it is. Maybe some exceptions could be entered in a much simplified definition. Only acute symptomatic could be kept as such.

Another issue is the use of ictal EEG findings in general such as a first GTCS provoked by disco lights and confirmatory myoclonic jerks during a PPR (EEG recording). Is that sufficient for the diagnosis of epilepsy or do we wait for another to occur? And if there is also a family history of epilepsy?

Most of the literature regarding seizure recurrence risks in non-lesional epilepsy is more than 10 years old. Of course older literature has value, but in our current world recurrence risks might have changed. Maybe there is indeed more current literature?

For better coverage of all expertise in epilepsy in defining an operational definition of epilepsy this initiative of asking comments is very good. In this respect, it is advisable to have a series of "overlapping" meetings with different (child!) epileptologists, general neurologists and pediatric colleagues as well as patients and parents. It could be a kind of estafette.

I hope my comments and ideas are helpful to you and if I can be of any help, let me know,

Dorothée Kasteleijn- Nolst Trenité

1 September, 2013

Thanks to the Task force for their work, here are some of my comments:

‘Disorder’ rather than disease: Although used interchangebly in many situatios, ‘disorder’ reminds more a functional abnormality, like epilepsy, with or without justified infrastructural involvement ( like ‘sleep disorder’, ‘anxiety disorder’); yet, ‘disease’ commonly refers to conditions with structural impairment and frequently non-paroxysmal (except relapsing-remitting conditions like MS) fashion as ‘Alzheimer’s disease or vascular disease’.

I would recommend ‘is over’ instead of ‘no longer present’ referring to age-related, self-limited conditions as ‘childhood absence epilepsy’ or ‘Panayiotopoulos syndrome’ . It is a term that does not totally leave out the weak possibility of recurrence of the same or another epileptic syndrome (as, photosensitive IOLE with evolution into JME; or, as complex febrile seizures and later developing MTLE syndrome). 'Remission’ be better preserved for epilepsies with less predictable outcome after a rational seizure-free, treatment-free time with or without EEG/imaging abnormalities.

For the 2. definition of epilepsy in table 2 my preference would be something as: One unprovoked seizure and a probability of further seizures based on the individual characteristics of the patient, the epileptic syndrome, intensity (or, severity, or, significance ) of the waking-sleep EEG abnormalities (as, ESES)... if we think that the majority of first seizures start at young ages.

Thank you for giving me this opportunity.

Aysin Dervent

1 September, 2013

I warmly congratulate for the excellent and difficult work of ILAE task force, and thanks for the opportunity to comment the proposal changes.

Here are my comments about definitions of epilepsy:

Probably, the duration of one seizure need to be better clarify. I agree with Dr. Wong comment about the criticity of the quantitative
definition of 75% risk of recurrence seizures. At this point a list of potential features that can rise the risk of recurrence should appear.

Anyway, thanks for your important work.

Sincerely yours,
Eduardo Cumbo

1 September, 2013

congratulations for this great work

Regarding the term disease, we prefer the term disorder, because it is more comprehensive and better respond to the situation of people with epilepsy can have seizures without yet be considered sick.

Regarding the definition of epilepsy, it is made after two seizures. What happens to some epileptic syndromes of the child as Landau Kleffner syndrome and encephalopathy with continuous spikes and wave  during sleep: Should he consider epilepsy even in the absence of seizure syndromes? must be added the special conditions in the definition?

The risk of recurrence of 75% is difficult to determine by general neurologists and the risk should be more detailed
thank you for giving us the opportunity to participate in this discussion

Chahnez Triki

Return to top

1 September, 2013

Firstly I would like thank all of you for getting me involved in commenting this important paper. I congratulate the epileptic Task Force for their work and effort and the resulting document. The proposed operational clinical definition of epilepsy sounds very useful, properly and timely addressing the long-standing matter as to how diagnose subjects with seizures; however I have some comments:

- I share other colleagues' perplexity about taking 75% recurrence risk of seizures as a cut off for a diagnosis of epilepsy in practice;
- I believe that "epilepsy" should remain a disorder rather than a disease, at least in most cases.
- I outline the utility of criteria for each level of certainty for epilepsy, similar to that used for other neurological disorders/diseases (Definite – Probable – Possible – Unlikely – Definitely not a seizure, et cetera) and propose to consider adding tiers of certainty.

Anyway, overall a very good and convincing job: Congratulations.

Ferdinando Sartucci

1 September, 2013

Comments as follows: I think the individualized clinician assessment of recurrence risk at >75% is quite problematic. Putting a number on the cut-off for recurrence risk without evidence or data is confusing. Since this is number is not evidence-based except for the recurrence risk of 73% after two unprovoked seizures as per Hauser et al, I would favor just using qualitative words, like "epilepsy is defined as one unprovoked seizure and a high probability of further seizures, of at least 50-75% over the next four years." This decision is made using the judgment of the clinician, based on history, EEG and MRI findings. This would still enable earlier, well-justified treatment intervention, but makes the case more strongly that this is a judgment call, and is not wedded to any statistic for which there is no solid evidence in many situations. There is also the consideration in the US that persons who carry a diagnosis of epilepsy have great difficulty obtaining life insurance, therefore, there is a downside to earlier, clarified diagnosis of epilepsy.

Cynthia L. Harden

1 September, 2013

Thanks for the new comments about Epilepsy Definition.

Here are my comments.

Determining 75% risk for having the second seizure it is not a Base Medicine Definition right now for the different kinds of Epileptic Syndromes and possible constellations. That could be and arbitrary number from my point of view but could serve as an instrumental definition only for research works.

When we use the term "NO LONGER PRESENT" it may say in many languages NOT PRESENT AT ALL, that it is not the case, because we all know that there are many patients with clear cortical hyper-excitability without anti-seizure medication and more than 10 years with any seizures but with EEG always active electrically. Those patients may have a seizure in any stressed context as electrolytic disarrangement or fever of any cause etc.

I would like to see the Electro clinical Definition of epilepsy including the EEG parameter that could help to resolve many of those patient with only one seizure and without any IRM abnormality but that could have some clear Epileptic Syndrome o constellation features.

If we use the Epileptic Seizure as one part of the definition of Epilepsy that means that it is necessary to demonstrate the hyper-synchronicity and the excessive cortical activity to talk about Epilepsy as a disease, so, what I think, is mandatory include the EEG criterion some where to give more statistical and practical strength to the definition.

Thank for your other comments.

Rodrigo Andres Solarte Mila

31 August, 2013

I would agree with Dr Vicente's statement that remission and cure are better than no loger present. Otherwise, I think a great job has been done.


Return to top

31 August, 2013

Thank you Committee for your diligence and fine work. I am a child neurologist and have but two comments:

1. Benign Epilepsy of Childhood with Centro-Temporal Spikes has been viewed by child neurologists and families as a disorder outgrown with age. To lump this condition with the other epilepsies and re-label it as a "disease" is unwise. It is not a disease, and is likely to cause parents and children undue distress if re-labled. Indeed, to label all epilepsies as diseases may be equally unwise as the lay public may interpret the word "disease" to imply contagion as was commented upon by a patient in an earlier posting. There are already many "conditions", like epilepsy, that are considered serious by the lay public (consider cerebral palsy and autism). The diagnosis of "epilepsy" is scary, and serious enough, for most patients and families in my 30 years of practice and teaching.

2. In the Section titled High Recurrence Risk, last paragraph on complex febrile seizures, third line, there is a colon after the word "focality". To avoid confusion this should be replaced by a semi-colon.

Thomas T. Reiley

30 August, 2013

The task force has recommended adoption of a sensible definition, one based upon epidemiologic studies.  A fundamental problem, that of spontaneous relapse and remission, makes diagnostic certainty impossible in many cases (even when 2 seizures have occurred) so the adoption of a criterion for 75% probability of seizure recurrence represents a practical approach. I support adoption of this proposed definition. I prefer the term "epilepsy in remission" rather than considering an individual as no longer having epilepsy when a patient has been in a state of prolonged remission and off medication, however.  While the probability of further relapse is low, the probability is relatively low for individuals still taking medication after an extended seizure-free interval. Hence, the distinction between such individuals is arbitrary.  Therefore, this third category may have some value, since it could apply to all individuals who have achieved a 10 year remission, whether taking medication or not.

Michael R. Sperling

30 August, 2013

I think that the new definition is adequate and very clear and is very important differentiation between disease vs disorder.
Respect to recurrence risk of >75% for patients after ten years free of seizures or history of age-dependent epileptic syndromes, I think should be more differentiated according to each patient's history.

Thank you for allowing me to comment!!

Enrique J. Ortega

30 August, 2013


I have read the revision of the definition of epilepsy. It would seem that by adding photosensitivity, you have given more people who have more than one type of seizure the ability to fit into the definition of epilepsy. Have you considered how the patient who has both epileptic and non-epileptic, as well as photosensitive seizures reacts to being told that, "no you don't have epilepsy all of the time?" In some patients epileptic and non-epileptic (those that do not show up on the EEG) seizures feel very similar and incapacitating. I would have liked to have a definition of epilepsy that could be broken down further than is presently done. We need a definition for genetic epilepsy that incorporates secondarily generalization, as well as other factors.

Nancy C Schumacher

30 August, 2013

As epileptologists, it is without doubt essential to define, as clearly as possible, the entity that is at the center of our clinical universe. So I applaud the renewed effort.

The definition of epilepsy has always ridden piggy-back on the definition of an "epileptic seizure" -- although that is somewhat circular. I have two comments:

(1) The first comment pertains to the definition of an "epileptic seizure" from the 2005 report: "An epileptic seizure is a transient occurrence of signs and/or symptoms due to abnormal excessive or synchronous neuronal activity in the brain." This definition is crafted with a clinical seizure in mind, not the phenomenon occurring in the brain itself (which has other dimensions including an electrographic one, in addition to the subjective and objective clinical manifestations). Nevertheless, a focal subclinical electrographic seizure (with no apparent symptom or sign) does not fit under this definition.

(2) My second comment pertains to the second item in the updated definition of epilepsy: "One unprovoked seizure and a probability of further seizures similar to the general recurrence risk after two unprovoked seizures (approximately 75% or more)."

Why not state this more simply as: "One unprovoked seizure and a 75% or greater probability of further seizures'

However, as several others have pointed out, how does one measure this probability of further seizures? How much probability of another seizure does any one of a spectrum of possible MRI findings represent? And what probability figures do we associate with any of a several possible EEG findings? Same goes for elements of clinical history that may have a bearing. It is not just a matter how a general neurologist will figure this out. I don't think the science exists to make this probability estimation possible -- except in the form of expert clinical opinion.

One could consider defining risk in terms of more concrete imaging/EEG/clinical findings. But this would make a very cumbersome definition. Alternatively one could simply say "One unprovoked seizure and clinical or diagnostic findings that favor a high probability of further seizures", or something to that effect. This doesn't sound very exact. But we can't have "exact" if there is not enough science to base it on.

Manoj Raghavan

Return to top

30 August, 2013

The main problem with this definition is non-convulsive status epilepticus. If we look at Landau-Kleffner syndrome, it is thought that 20-30% may not have seizures, at least early on, and that the likely cause is a high rate of subclinical seizures particularly in sleep. Infantile spasms are a further example but they are accompanied by seizures although loss of skills can antedate their onset.

Then there are situations with subclinical status in which the manifestations are cognitive and behavioural and even if seizures may occur the more prolonged episodes may not be accompanied by seizures.

What really matters is that some of the most severe forms of epilepsy appear to occur with a high rate of subclinical seizure activity whether or not clinical epilepsy occurs.

Thus this activity must be recognised and acknowledged as the cause of some of the most severe forms of epilepsy. It reads more like a document intended for adults.

Brian Neville

30 August, 2013

Congratulations to the ILAE task force and Robert Fisher to bring forward the definition topic.

I completely agree with what is written in Table 2 besides the word ‘disease’. Epilepsy is a symptom of some underlying brain dysfunction, a disease. Thus, ‘disorder’ is a better and less provocative word and will not be poorly understood.
I’m also concerned about the figure "≥75%". There are so many factors influencing the course of epilepsy, but this figure may be practical in the discussion with patients.

High recurrence risk
There is a discrepancy between the sentence "...epileptologists consider epilepsy to be present after a single unprovoked seizure, because of very high recurrence risk", (e.g a child with focal cortical dysplasia), and further down ...”a single seizure plus a lesion... does not automatically satisfy criteria for this operational definition of epilepsy”. Focal cortical dysplasia is a lesion! This must be clarified.

CFSs are risk factors for future epilepsy. In estimating a risk factor also inheritance and occurrence of an epileptiform EEG must be considered.

Reflex epilepsy
In spite of earlier classifications I have always considered repeated photostimulated seizures as epilepsy. The association with "an enduring abnormal predisposition to have such seizures” is important.

Epilepsy no longer present
This is a better and more clear term than "remission”, which in reality means reduction, diminution, decrease.

Last paragraph: "Some risk factors...
I have difficulties in understanding the sentence "..., but do not necessarily raise the risk high enough to justify such a diagnosis...” Again the ≥75% figure is diffuse in assessing recurrence risk. A recurrence risk for seizures is not precisely known.

Imperfect information
The importance of taking a very careful personal and family history should be highlighted in this paragraph.

I can see that the task force uses the term "anti-seizure drug” instead of "anti-epileptic drug” (AED). I think this is logic, but has it been accepted?

Uppsala, Sweden, 2013-08-30
Orvar Eeg-Olofsson, MD, PhD

30 August, 2013

I agree with the definition of "epilepsy" even after one only seizure, when there are strong suspicions of recurrence of further seizures. From a clinical point of view, I believe that it should be also taken into consideration the context of seizure occurrence, especially regarding the occupational risks. I have always considered with utmost attention (and prompt therapy) a first seizure of generalized type, i.e. without warnings, in a subject working in dangerous settings like scaffoldings, trucks or professional car driving, dangerous electrical or mechanical tools like circular power saws et similia. We had all had the experience of the ineffectiveness of information when loss of employment is feared.

Mario Manfredi

Return to top

30 August, 2013

I commend the task force for doing a great job with the new operational definition of epilepsy. I have three issues to comment on:

The first is the introduction of >75% probability of future seizures which would be difficult to determine in clinical practice. The task force noted that the new definition is more complicated that the old one especially in clinical setting. I look forward to seeing a revision or appropriate description of criteria for determining 75% probability.

Secondly, I agree with the use of the term "epilepsy no longer present" rather than "remission" or "cure". I think it is a better and more appropriate description.

Thirdly, the impression that an epileptic syndrome referrable to a particular age is considered not no longer present after that age is passed may imply that such individuals are "cured". This is not too clear or could such patients be classified under the syndrome of another age group?

Adesola Ogunniyi

30 August, 2013

I did respond earlier. I have major concerns re legal status of "epilepsy no longer present" and declarations on driving licence applications, particularly in the wake of R v Gillette.

Roy G. Beran

30 August, 2013

An excellent and concise review of the issue. Two comments:

It is not clear if the plan is to replace Table 1 with Table 2 or supplement Table 1 with Table 2. I believe it is crucial to emphasize the point made in Table 1 that:

Epilepsy is a disorder of the brain characterized by an enduring 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

This is often insufficiently emphasized and results in the misunderstanding, by patients, schools, support systems, etc., that control of the seizures is NOT the end of the problems associated with epilepsy.

In table 2, I would CAPITALIZE and BOLD the 'OR'

Epilepsy is considered to be 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

Thanks to the committee members for their efforts.

David E. Mandelbaum

30 August, 2013

Here are some of my comments on the proposed operational definition:

"An epileptic seizure is a transient occurrence of signs and/or symptoms due to abnormal excessive or synchronous neuronal activity in the brain." suggest remove the term synchronous and instead "An epileptic seizure is a transient occurrence of signs and/or symptoms due to abnormal excessive neuronal activity in the brain."

This definition of having to meet 75% chance of recurrence in life term may be confusing for non epileptologists. It would be better if certain clear cut situations can be given, wherein the risk of recurrence could be considered as more than 75% (cortical dysplasia, remote symptomatic seizures following head injury, strokes, CNS infections). This would be of importance in medico legal situations, wherein the percentage of risk of recurrence can be disputed. It will be better if the definition is scientific, practical and address practice issues especially legal!

Sridharan Ramaratnam

30 August, 2013

Excellent job and very clear manuscript from my point of view (the one of a basic scientist). However, the expression "epilepsy is no longer present” is a bit unsatisfactory to me. It’s ambiguous because would imply to some extent that the "disease” is still there and that seizures are a symptom no longer present (what can be the case but bring us to a different debate). You can have flu every winter but you don’t say in summer that the flu is no longer present; you just say I’m not sick this summer. I would prefer "is no longer epileptic” that does not exclude at all the possibility for getting epileptic again or eventually use the term "remission" more commonly used for other diseases.

Thank you very much again for this important contribution
Alfonso Represa

Return to top

29 August, 2013

I must congratulate you all for efforts and time spent to redefine Epilepsy.

I have few points to comment on:

  1. In developing countries like India, understanding of epilepsy is poor.
  2. There are some other problems having similar symptoms like epilepsy are confused with epilepsy.
  3. Without EEG, it can not be confirmed as epilepsy.

I personally feel that EEG should be considered in definition.

Mohd. Farooq Shaikh

29 August, 2013

Thank you for this!

I have lived with epilepsy my whole life and am still, so I do consider it a disease, not a disorder. The doctors told me that I was cured when I was about 12 years old, well naturally I wasn't. They put me in a bit of a pickle and now I am in worse condition where I have siezeures on a daily bases just about all day long and medicines don't help and the VNS didn't either. So, in the case of remission or no longer present as Deborah Lee suggests, I believe it should possibly be in remission because it never really goes away. there always will be jerks in between as well. In between the remission part that is. We as patients are the ones that know the best. We are the ones fighting the disease, right?

I also don't think we should try to candy coat this by saying its not a disease, like ms or fibromyalgia. I feel the same as they do I never get rest. My body is exhausted from all the seizures I have. So, I think it's safe to say it's a disease. there are different types of epilepsy YES, but they all take a toll and we all fight just as hard. Sometimes the nueros can't figure it out - the eptologists sometimes can't figure it out.

So a disease YES
and remission YES
and lets remember that for a wide range of patients siezeures or episodes don't show up on EEG's

Thank you
Tina E. Muhs

Second comment, 30 August:

I was born with epilepsy and it took 6 yrs. to find that out major testing so I find this to be a disease NOT a disorder!!

All through grade school because of poor testing, I was told I NO LONGER HAD EPILEPSY, and a discontinuation of all medication when in fact it was in REMISSION!!! only to be brought back in my senior year as with vengence!! because of no therapy or medication. I had grand mals and no doctor believed me they diagnosed me with myopothy!!! So, therefore it was in REMISSION once again.
You can't tell someone that they don't have Epilepsy anymore, you can tell them they don't have it anymore. Someone that is diagnosed with Epilepsy has epilepsy forever, and there is no such thing as it going away, it can go in to REMISSION where as all medication must be kept going in a lower dose as needed or higher if needed as the eptologist says, thus keeping the patient into a real mess of epileptic problems as I have been put in and deal with.

Also, in your defintion , I didn't see anything written as far as the noises and sounds disrupting and triggering the siezures, which is a different kind of epilepsy!! There isn't just one kind of epilepsy a text book epilepsy if you will like you learn in school. I have 3 diffrent kinds and I am the only one like this per doctors. Which there are problaby many more. So, your theory doesn't add up except for your text book learnings and conclusions only and I say ONLY. I believe that we have been let down and this new way is trying but only doing the same because for Epileptics outside your textbook diagnosis!!

Tina E. Muhs

29 August, 2013

I have some comments about the definition.

- For me the main problem in the new definition is the 75% risk to consider the diagnosis after a single seizure. I read the article and the section supporting this recommendation is very weak. I think there is some evidence to estimate which patient can have a seizure recurrence, maybe is not the best evidence, but it is not discussed appropriately in the article. I think EEG and MRI many times guide the decision to start treatment after a single seizure but also help in the decision for the diagnosis of epilepsy. I am surprised that the article mentions that the EEG is not important in the diagnosis. I think the article has to be more specific in that section with the criteria to define 75% risk.

- I think the definition of no longer epilepsy is weak. The articles describes different potential times to consider "no longer epilepsy" which makes confuse the paragraph for the reader. Also no specific criteria are described apart from the time. I feel that we do not have strong information to potentially create a definition in this subject. I am not sure that the section has to be included in the article.

I am Ok with the new criteria of two seizures in a setting of reflex epilepsy

Dr. José Téllez-Zenteno

Return to top

29 August, 2013

If the goal of proposing an alternative diagnostic pathway (Table 2, item 2) is to provide a practical definition for the high risk for recurrence after a single unprovoked seizure, the intent and content of the ILAE Task Force paper are internally inconsistent, and are likely to create confusion, rather than clarity, in clinical settings.

Although intended as a guideline rather than a sharp cutoff, the paper overemphasizes the >75% threshold. The "probability of further seizures similar to the general recurrence risk after two unprovoked seizures" would be expected to range from 59%-87%, reflecting the 95% CI of the recurrence risk in the Hauser, 1998 reference.

An operational clinical definition has operational implications. According to the paper, "choosing a specific threshold risk might be excessively precise" but "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." While it is an acceptable academic position, the clinical utility of this approach is questionable. Absence of proof is not proof of absence. In reality, the statement "if a treating physician determines ... an enduring predisposition for unprovoked seizures ... then that person too should be considered to have epilepsy" reflects a more pragmatic approach. The proposed alternative definition using the >75% threshold neither reflects nor informs clinical practice, as demonstrated in Case Example 5, where "many clinicians would reasonably treat this man with anti-seizure medications" though "epilepsy cannot yet be said to be present according to either definition."

It is also unclear why two seizures are needed for diagnosis in a setting of reflex epilepsy (Table 2, item 3), if "an enduring abnormal predisposition to have such seizures" can be documented with an abnormal photo-paroxysmal response on EEG after a single photic-induced seizure. This requirement contradicts the statement that the "diagnosis of a specific epilepsy syndrome associated with a persistent threshold alteration can be made after the occurrence of a single seizure."

O'Neill D'Cruz

29 August, 2013

Thank you for the opportunity to comment on the definition of Epilepsy. My comments pertain to the implications section.

There are implications for epilepsy research in animals where there may be patterns of EEG activity that appear to be seizures but are actually normal EEG. Here the word "normal" and "seizure" are critical and need to be carefully used, just like the word "epilepsy."

One reason is that many rodents have rhythmic discharges recorded during EEG with epidural or implanted electrodes. Some of these discharges occur with behavioral arrest, and may be confused with Racine stage 1 seizures. Some may be accompanied by head nodding, mastications and other minor movements that are also associated with Racine stage 2 seizures. The discharges, when recorded with cortical or thalamic electrodes, have a spike-wave morphology so they are assumed to be thalamocortical oscillations and the behavior is similar to an absence seizure. Is this a seizure? Do these animals have epilepsy?

When recordings are made from hippocampus, rhythmic spiking occurs at approximately the same frequency, which may be mistaken for a hippocampal seizure or limbic epilepsy if that animal has no simultaneous recordings from other brain regions.

For these reasons and others, defining epilepsy in animals is arguably as difficult as it is in humans.

Helen E. Scharfman

29 August, 2013

Congratulation to the epilepsy Task Force. I agree on their definition of both epilepsy & seizure.

Best regards,
Dr Tayseer Zein

Return to top

28 August, 2013

I congratulate the epilepsy Task Force for their work and effort, however I have some comments.

1. It is indeed possible to be diagnosed with epilepsy after having one single seizure  if the patient has a predisposition to further seizures, as it was mentioned in the 2005 ILAE definition and the current proposal. However it is quite difficult to grade a 75% risk for a physician. First, epidemiological studies can be controversial and results variable. 2. It is easier for an epileptologist to be aware of these studies, but what about a general neurologist? If we assume this statement the correct diagnosis of epilepsy will be restricted to epileptologist. 3. Why is used a 75% risk? Probably in terms of  treatment onset patients are treated with a less risk (specially if they have had a tonic-clonic seizure). I find it is important to connect the definition of epilepsy with treatment onset.

2. The term "no longer present" is not used in other diseases, so it will be more difficult to be understood by other colleagues less close to epilepsy. I think "remission" and "cure" are suitable terms. For example if a patient with BECTS  has his last seizure when he was 7 years old and he is currently 20 years old he is cured. If a patient had a childhood absence epilepsy and was withdrawn from medication just six months ago we can assume he is in remission.

3. The scenario of epilepsy surgery iss not fully described. When should we assume curation in a patients that has been operated on? I consider that to wait for ten years without medication in those cases to consider epilepsy "is not longer present" is too much (Ie. a patient  with a gangliocitoma completely resected). This is very important in terms of driving license in some countries.

Kind regards

Vicente Villanueva

28 August, 2013

Congratulations to the Commission for this document with an improved definition of epilepsy. The opportunity to comment is greatly appreciated.

-The ‘at least 75% risk’ to have a second seizure may be difficult to be applied on large scale, e.g., by primary physicians care. In addition, there are still limited epidemiological studies assessing the exact risk for each potential situation.

-The definition of epilepsy ‘no longer present’ does not have the same impact of ‘remission’ which is instead firmly established in medical and even forensic practice.

-Case examples appear sometimes to be out of touch with the real world.  And what about if case #5 would be 18 year old and he would ask to take the driving license?

-Based on Table 2, point 3 a girl with a tonic-clonic seizure in disco and a clear photoparoxysmal response at the EEG should await for the second seizure before receiving the formal diagnosis of epilepsy. Conversely, a girl with a tonic-clonic seizure after a party in disco and a sleep-deprivation EEG with generalized abnormalities could receive the ‘formal’ diagnosis of epilepsy even after the first episode.

Anyway, overall a very good job! Thank you for your great effort
Pasquale & Salvatore Striano, Italy

28 August, 2013

Thank you for allowing me the opportunity to comment on the paper "An Operational Definition of Epilepsy”

Disorder versus disease:
I believe that "epilepsy” should remain a disorder rather than a disease. The word disease has more serious social impact in some respect than the word disorder. Also disorder allows for more heterogeneity than the word disease which is more applicable to epilepsy. In addition, using disorder is helpful in distinguishing etiology: for example epilepsy is the disorder which may arise from the disease SLE. Finally, the Task Force thought to use the word disease because of the enduring predisposition to seize; however, what of those childhood epilepsies that do not imply an enduring predisposition to seize? As a pediatric neurologist, I find the word disorder much more appropriate and familiar such as Autism Spectrum Disorder which in some respects similar to epilepsy: may represent a lasting derangement of normal function, significant heterogeneity, and opportunities for remission.

Recurrence risk of 75%:
While I think this a good idea, I think the manuscript should provide more concrete guidelines on what may generate a recurrence risk of 75% or greater, especially it is to be used by more general neurologists or even non-neurologists. While it may be second nature for an epilepsy expert to be able to make this determination, it may be difficult for a less experienced physician to do so. Although each patient will be different, some concrete examples of what may constitute conditions where a recurrence risk of 75% or greater might be expected would be useful: ie one febrile seizure- no, one febrile seizure plus positive SCN1A testing-yes. Also, it might be useful to have a position on how this might relate to treatment. Since I can imagine the greatest use for these guidelines is for the non-epilepsy specialists, it may be of benefit to have more helpful positions: how does making this diagnosis under the new definitions affect treatment? Since the definition now allows for the diagnosis of epilepsy to be made without a second unprovoked seizure, that implies treatment should be started earlier.

Remission versus "no longer present”:
I do agree that there are certain childhood epilepsies that can be clearly "outgrown” implying epilepsy is no longer present. And after 10-years seizure-free off medication, that epilepsy can be determined to be no longer present. Ironically, I do prefer this terminology because the use of the word "remission” to me implies a disease (such as in remission from leukemia) and I would prefer to stay away from the concept that epilepsy is a disease.

Again, thank you for allowing me to comment.

Deborah Lee

Return to top

27 August, 2013

Here are my comments to “An operational clinical definition of epilepsy”:

Epilepsy, epilepsies or epileptic syndromes - singular versus plural?
The Task Force discusses this issue under the heading “Disease“. I just want to switch the focus. Do recurrent epileptic seizures related to a brain tumor belong to an identical disease or group of diseases as Childhood absence epilepsy? I would prefer the plural in order to draw the attention to the diversified situation. The plural seems to be more important than the terminological discussion about disorder, disease, condition and so on. What justifies assuming a single disorder, disease, condition … in all patients suffering recurrent seizures?

High recurrence risk
The Task Force has to be congratulated for the effort to rationalize the mechanical “Two unprovoked seizures” rule. This rule – although simple – has the potential to simplify medical decisions in an undue way. A clinician has to consider all variables for an increased risk for seizure recurrence. The Task Force should specify these variables like type of structural lesions, EEG patterns, neurological deficits and seizure types. For example: Can an absence seizure or other specific seizure types occur outside epilepsy? Another point might be added. Several patients came to medical attention after a single unprovoked GTCS. No one of the above alluded variables is present. However, close history taking, especially with patient´s relatives reveals the pre-existence of “minor” epileptic seizures like absence seizures or auras, and therefore documenting the existence of epilepsy. It would be helpful especially for not specialized clinicians to list all these factors and combinations of symptoms and signs suspicious for the existence of an epileptic disease in a patient not fulfilling the “Two unprovoked seizures rule”. Concrete data indicating a > 75% risk are needed. In this aspect, I fully agree with comments of Raffaele Manni, Shuli Liang and many others.

Several comments discuss the negative legal and social consequences of widening the definition. However, the reluctance to use a correct medical diagnosis mirrors the prejudice we want to overcome.

Gerhard Bauer

27 August, 2013

The task force have clearly been wrestling with some challenging scenarios. There are two parts of the definition that I would urge the task force to reconsider.

1. ‘one provoked seizure and a probability of further seizures ... approximately >75%.

  • There are a number of problems with this definition both conceptually and operationally.
  • Risk is ‘continuous’ (ranging from 0 to 100%) and also time dependent. Any attempt to dichotomise, and not take time into account, is problematic.
  • The 75% figure is arbitrary - although informed by a population average from previous research. So, some people with newly diagnosed epilepsy have a recurrence risk of <75% in the next few years. I think that we would still say that they have epilepsy.
  • From a dichotomy point of view, I doubt that it makes sense to give a diagnosis of epilepsy to someone with a recurrence risk of 80% (time period to be defined) but not to someone with a 70% risk.
  • The time period over which any threshold applies must be defined.
  • Estimating recurrence risks for individuals is challenging as populations are heterogeneous and there are a number of prognostic factors. Our work on prognostic models in these populations show that models can be built but also that some variability remains unexplained, and that any estimate comes with a confidence limit.  As a result, the proposed definition is unlikely to be applied reliably or uniformly in clinical practice. We would need to ensure that there were no perverse incentives for giving a diagnosis of epilepsy (eg different tarrifs for single seizure and epilepsy).
  • There is a risk or confusing a definition of epilepsy with likelihood that the patient will benefit from starting treatment. I note that there is comment in the manuscript to this effect, but I still fear that changing the definition will result in a significant proportion of first seizure patients being started on antiepileptic drug treatment, despite overall evidence of benefit for many.
  • Ultimately, I’d suggest that the most important issue is the patients risk of recurrence (and remission), and whether we call something epilepsy or not is of course semantic.
  • My preference would be to abandon definition 2 for now and move to ensure that we are able to identify patients’ seizure recurrence risk (in say next 2 years) and their likely benefit from AED treatment.
  • So long as patients are appropriately characterised, investigated and managed, I can see not benefit from changing the definition from the perspective of either clinical practice or research.

2. Epilepsy no longer present.

  • We need to be clear about what it ‘present’ or not. Proving absence or a negative is always difficult.
  • Seizure occurrence risk over a unit of time will never be zero, although what is of interest presumably is the risk in the individual carrying a diagnosis of epilepsy relative to a non epilepsy population.
  • Definitions based on sound epidemiological data will not generally be possible. The risk is that we end up with a multitude of conditional definitions that will increase in complexity over time.
  • My preference is to continue with the notion of remission (x years) either on or off treatment.
  • The word remission does not preclude the possibility of relapse and does not necessitate an attempt to try and prove absence.
  • Finally, I’d suggest that it is practically impossible for someone to have a 10 year remission off treatment and then to have a >75% probability of seizure recurrence, even though the time frame over which that recurrence might occur is undefined.

All the best

Tony Marson

Return to top

27 August, 2013

This is a clear and concise up date of the 2005 definition of Epilepsy.

I have only one comment regarding the suggested change from "disorder" to disease.

We all agree that Epilepsy is a common denominator to a life time brain disorder with a variety of etiologies causing the brain to "seize". As such' the term "disorder" seems to me appropriate. The argument given that the term Disorder is "poorly understood by the public and minimizes the serious nature of epilepsy" is far from convincing. Attention Deficit Disorder and Autistic Spectrum Disorder are examples of conditions which affects a significant portion of the general population during their life time, with serious implications on the life quality or the affected. I do not believe that replacing disorder by disease regarding those 2 conditions, as well as epilepsy, is going to improve the understanding of those disorders by the public or change the seriousness of being affected by them.

N. Gadoth

26 August, 2013

Further to my comment of 20th August, I once again speak to you as a man who has had epilepsy for a lifetime. First, undiagnosed, with all the impossible misunderstandings that meant. Second, mis-diagnosed, with the resulting impact on sense of identity. Third, and finally, an accurate diagnosis, with the accompanying struggle to find my true identity.

I am a man who has the serious neurological disorder of frontal lobe epilepsy. Complex-partial with secondary generalised seizures. It is not contagious, although there are some who still believe it is. They think I have a disease, and avoid me so as not to catch it.

Now some of the most eminent epileptologists on the planet are telling me that those ignorant and bigoted people are correct. Does this mean it's time for another identity crisis?

In UK disability rights circles we have a saying,"Nothing about me without me". I seem to be the only lay person who has epilepsy to venture into this discussion. Did I waste my time, and yours?

Richard Johnson

26 August, 2013

Good morning:

Thank you very much for the opportunity to comment the difficult and hard work of the ILAE task force.

1. Epilepsy is a disorder of the brain found in many different diseases. Epilepsy consists in and enduring predisposition to seize and can be observed in mild and reversible diseases and also in progressive and fatal diseases. In my opinion, a patient cannot be defined as "epileptic" and needs a more precise diagnosis of the disease that affects him or her.

2."High recurrence risk of 75% of more". This term is very unprecise and confusing. Moreover although is impossible to enumerate all the conditions in which the recurrence risk after a single seizure is very high, the most common should appear in the paper. I will ask also for definitions of abnormal EEG and abnormal neuroimage.

Best regards,
Mercè Falip

26 August, 2013

Thanks for all the offered and congratulations to the Task Force, expecially the Table 2 determine the definition of epilepsy which support clinical situation and for NCSE, ESES, and CSWS, need to be add in the different way.

Hassan Jassim Al Hail

Return to top

26 August, 2013

This is an excellent effort to revise the concept of epilepsy. Here are my comments:

1) The need to define epilepsy should be calibrated according to the objective. As I understand it, the main goal of the position paper is to give a practical support to neurologists to discriminate epidemiologically and prognostically different conditions. If this is the case, the definition should be simple. My impression is that the present triad of defining criteria is unbalanced. While the first and most of the second principles are widely accepted, the threshold of 75% risk and the third criterion are both based on information that is not broadly available and are specific knowledge of epileptologists. This may weaken the value of the new practical/operational definition and, on the other side, may not be accepted by all specialists. For example, as discussed in the task force draft paper, if a reflex epilepsy is recognized (most likely by an epileptologist, hardly by a non specialist), epilepsy is diagnosed and there should be no need to record at least two seizures to define it. In this case, a definition tuned on expert knowledge may be questioned by epileptologists.

2) Disorder vs disease: The problem with defining epilepsy is complicated by the fact that the disorder is described by the symptom. No seizure, no epilepsy. The term epilepsy by itself is an abuse. All seizures, acutely provoked or due to an underlying disease, are just seizures secondary to a transient or permanent alteration of excitability due to a "disease". So, almost by definition epilepsy cannot be considered a disease. As any other general term utilized in medicine (neuropathy, cancer, dementia, etc.) the term does not define a disease but a group of diseases – a disorder.

3) As for epilepsy remission, "no longer present" negates the presence of a condition (disease, in this case) that might be still active, but latent. I wonder if the term "present" should not be substitute by "active" No longer active suggests that the etiology (genetic, lesional etc) is still there, quiescent, and that it is very unlikely (but not impossible) that it will be reactivated even under special conditions.

Marco de Curtis

26 August, 2013

I think that ILAE guidelines should be absolutely clear, so they would be uniformly interpreted by doctors around the world, and this would make future studies easier to compare.

In my view, the abstract section could simply say "the diagnosis of epilepsy should be made after two unprovoked seizures occurring more than 24 hours apart or after a first unprovoked seizure when there is a >75% risk of recurrence". The data that justify this sentence could well appear in later sections of the paper. On the other hand, epilepsy with reflex seizures would not deserve special mention since its diagnosis would follow the same rules as any other type of epilepsy.

Regarding "epilepsy no longer present", I would prefer other expressions, such as "prolonged remission" or "long-term remission", given its traditional use. As a general rule, the Task Force chose a well-defined limit (10 years without seizures nor medications) to make this diagnosis. If a patient has been asymptomatic for a decade being off therapy, there is a solid reason to think that his brain has lost its previous predisposition to generate seizures, even if a condition with a theoretical >75% risk of further seizures is present. In the case of epilepsies with age-dependant expression the diagnosis could become more complicated. The case example number 4 was that of a 25-year-old man with benign epilepsy with centro-temporal spikes who suffered his last seizure 12 years before. This patient was presumably without treatment in the last 10 years, so we could considered his epilepsy in remission according to the general rule. What we do not know is if we could have done the same diagnosis when the age of this patient was 17 or 20 years. It seems that remission in patients with age-dependant epilepsy would deserve a more detailed guideline.

The Task Force initiative of offering their article for consideration to other ILAE members before publication should be acknowledged.

Juan Gomez-Alonso

26 August, 2013

Thank you very much indeed for your work and the opportunity to comment.

1. Disease: not sure about the switch in terminology. On one side, we are talking about epilepsies, which encompass a vast range of etiologies and conditions that are hard to lump in a unique "disease". On the other side, new data from literature have highlighted the burden of comorbidities and the risk of premature death accompanying seizures, therefore the term "disorder" could actually minimize the serious nature of epilepsy.

2. Two unprovoked seizures and high recurrence risk: very reasonable to include single seizures in a high risk context. Nevertheless, suggesting the physician to judge whether there is "a risk similar to the general recurrence risk after two unprovoked seizures (approximately 75% of more)" could really be unpractical for an operational definition. I understand neurologists could still rely on the classic definition if they do not feel confident on prognosis, but there is a risk this criterion will be applied by few people. It might be an idea to suggest one unprovoked seizure plus epileptiform abnormalities on a baseline EEG performed more than 24 hours later (similarly to the role of MRI in the new criteria of dissemination in time for multiple sclerosis) or a clear cut epileptogenic lesion on neuroimaging or a diagnosis of a specific high risk syndrome; of course these conditions would not raise the risk of recurrence above 75% per cent, but this cut off comes from the classic "two unprovoked seizures" idea and is someway arbitrary (it is not possible to calculate the risk of "one and a half unprovoked seizures").

3. Reflex seizures and epilepsy no longer present: I feel these criteria have a remarkable practical value.

Kind regards
Emanuele Bartolini

26 August, 2013

It's a great opportunity to ILAE members to take part in I would say historical discussion on epilepsy definition. Thanks a lot!.
Yes and again yes! Epilepsy is not a disorder, but a disease! And not only because "the disorder implies a functional disturbance and the disease may convey more lasting derangement of normal function". From my point of view, Epilepsy is a nosological entity, a separate disease with her, only her peculiar clinical features, course (including the so-called natural course of the disease) and consequences.

Many years ago in the former Soviet Union, if the patient, for example, had post-stroke seizures, they were considered as an epileptic syndrome or a disorder on the ground of the underlying disease. Now it is called post-stroke epilepsy, that is a separate disease with its specific course. And because certain diseases can be treated and cured, it may also be applicable to epilepsy (e.g. BECTS).

Some comments argues that the term 'disease' will lead to stigmatization of patients. But in this case we are talking about the conceptual approaches to the definition of nosological entity, not the psycho-social consequences of such a term. In the diagnosis of the GP hardly sounds term "epileptic disease" or "epileptic disorder", rather "epilepsy" or a specific epileptic syndrome – and all! Therefore, patients are unlikely to have a claim to a new definition of epilepsy

With best regards,
Lidiya Mar'yenko

Return to top

26 August, 2013

Here are my comments about definition of epilepsy:

Duration of one seizure is used to decide its diagnose; need to clarify the duration; most of physicians think that this duration less than 5 minnutes;

In my observations:

Some types of epilepsy; ex. one patient with laughing seizure may last 15-20 mintues; or a temproal epilepsy case maybe up to 30 minutes, with behaviour disorders, without tonic-clonic seizure. Otherwise, one ictal duration is 2 mininutes, but its post-ictal cognitive disorder lasts 15-30 minutes or more;

Dr. Le Hieu

26 August, 2013

About the new definition of epilepsy, I exactly agree with Dr. Michael Wong comment (20 August) about the criticity of the arbitrary quantitative definition of 75% risk of recurrence of seizures.

I think that all neurologists are involved in diagnosis and treatment of epilepsy and it is difficult to require a "their own assessments of the literature on risks of recurrence for all possible situations".

Thanks to ILAE collegues for their excellent and important work  and for the opportunity to comment the proposed changes to the definition of epilepsy.

Fabrizio Antonio de Falco

26 August, 2013

The operationalization of defintions, so they are usable in practice is most useful exercise and I sugget to err towards advocating simplicity and practicability over complexity and precision, that may be hard to apply in practice.

In terms of risk of recurrence. >75% is going to be difficult to apply in practice, so suggest >50% instead (ie more likely than not).
The duration of time during which the risk applies needs consideration: one year, 5 years, lifetime?

I favour remission over "no longer present" . Remission has good public understanding and validity from oncology. Similarly, relapse is a well-recognized term.

For the generality of people, disorder is preferred to disease

John Duncan

25 August, 2013

I congratulate the authors to improve the Epilpesy definition I think the "risk of 75% or more to have a second seizure" could be better to clarify adding a list with these risk factors in the new definition Also to add in the definition the etiology of provoked seizures to better understand what are the unprovoked seizures.

Thanks for your work,
Gemma Aznar Laín

Return to top

25 August, 2013

Well done job, this definition considers most of the uncertainties. Recent Glasgow data suggests that "in abeyance" would be most appropriate than "remission".

JMK Murthy

24 August, 2013

I applaud the Task force for taking on this difficult job.

I have only  a few comments

a) By any definition, epilepsy is manifestation of a disturbance of normal neurological function and is therefore a disease. Whether this poses an individual a high risk of recurrence or not depends on a range of factors and should be discussed in appropriate detail with the individual patient. Each legal framaework ahs a duty to consider, not whether the individual has a disease or not, but whether the risk of recurrence as far as can be estimated poses requires any change in employment of driving eligibility. 

b) Differentiation of  provocation  and aetiology surely depends on temporal association:
 - Aetiology – factor that leads to a long-term increase in seizure recurrence risk
- Provocation – Factor that leads to a temporary increase in seizure risk.

c) In determining the role of AED treatment, the task force should separate the need for treatment from the diagnostic label. As a start, could I propose the following two lists (as a start, and not necessarily complete!)
i) Indications for long-term AED therapy:
1)    1- Epilepsy: enduring risk of unprovoked seizures
2)    2 - Single seizure with factors leading to increased risk of seizure recurrence
3)    3 - Presence of Biomarkers (eg EEG, genetic changes) associated with significantly high risk of seizure occurrence
4)    4 - History of seizures induced by unavoidable, unpredictable, or minimal  provocations  
ii) Indications for withdrawal or withholding AEDs:
1) Lowered risk of seizure recurrence in patient with previous diagnosis of age-specific syndrome
2) Duration of seizure-freedom statistically associated with appropriately low seizure recurrence risk
3) Seizures sufficiently infrequent, unintrusive, or mild that the patient chooses to refrain from regular daily treatment

d) Lastly, the term 'cure' should reflect any situation where the risk of seizure recurrence has dropped to levels approximating background population risk even without provision of AED treatment. Examples would include
- Prolonged seizure freedom post surgery
- Prolonged seizure freedom after remission of age-specific syndrome

John Paul Leach

24 August, 2013

The work done by Fisher et al is impressive.

The definition of epilepsy is a clinical diagnosis; therefore, EEG characteristics are not part of the diagnosis. However, performing an
EEG when possible may reduce over-diagnosis of epilepsy in cases where clinical phenomena mimic epilepsy.

In addition, how can we include under the definition of epilepsy, epileptic types such as:

  • NCSE: nonconvulsive status epilepticus, which in children is thought to be under-recognized;
  • ESES: electrical status epilepticus during slow sleep;
  • and CSWS: continuous spikes and waves during slow sleep.

Sincerely yours,
Varda Gross-Tsur

24 August, 2013

To define simply epilepsy is a difficult exercise, the clinical situations being so different from the neonate to the elderly. This practical revised definition is necessary for basis clinical diagnosis, and epidemiologic, social, or health economics aspects.

I have three comments:

1-As a neuropaediatrician, I would keep the previous following sentence: "Epilepsy is a disorder of the brain characterized by a predisposition to generate epileptic seizures, and by the varying neurobiological, cognitive, psychological, and social consequences of this condition" (table 1, 2005)(removing the word "enduring", and adding the word "varying"). It appears quite unsatisfying to characterize the epilepsy only by the occurrence of seizures, as most epileptic syndromes of children may show interictal "epileptic" activity, which may be prominent in some rare syndromes without overt seizures, such as Landau-Kleffner syndrome.

2-The item 2 of the new definition is important, however, is it necessary to mention the rate of 75%? It seems quite high for some epilepsy syndromes (such as BECTS). I would rather propose: "One unprovoked seizure associated to an underlying condition (structural, genetic, or a specific EEG activity), with a probability of further seizures similar to the general recurrence risk after two un provoked seizures".

3- The notion of time limit, or extinction of an epilepsy, which is not so rare in children and adults, is also important to point. However, I would prefer the terms of "cure" (after a successful epilepsy surgery), and the term of "remission", which are more medical than "no longer present".

Thanks to the group for this huge and necessary work!
A. de Saint-Martin

Return to top

24 August, 2013

I have a feeling that physicians will respond with greater enthusiasm than patients. Both the old and new definitions are plagued by the inherent discrepancy between the threshold at which we become certain that an "enduring condition" exists and the practical need to treat to lower risk. People who have a single, unprovoked seizure and no other risk factor have an increased risk for future seizures that is attenuated while they take anticonvulsants. On rare occasions, it may make sense for such a person to take an anticonvulsant just as it may in rare situations make sense for people with ongoing seizures to forgo anti-seizure medicines (children with Rolandic epilepsy, pregnant women with absence epilepsy, etc.). Perhaps it isn't necessary to draw the line at a place that we feel is a good "treatment" threshold. Instead we should treat based on risk for recurrence but "diagnose" based on a higher level of confidence.

This new definition ensures that a quarter of all people with newly diagnosed "epilepsy" based on the operational definition really don't have an enduring condition because without treatment they would be seizure-free for the rest of their lives. I agree with Dr. Beghi that using possible or probable epilepsy until a higher threshold is reached would be better.

Paul A. Garcia

23 August, 2013

Jackson's definition is about 150 years old now and is still good for a conceptual definition for epilepsy.  It is simple, scientific and to the point.  Since it missed couple of essential features, here is what did to modify it, when I became an epileptologist in 1984.  "Epilepsy is the recurrence of a seizure which is the subjective or objective behavioral manifestation of an aberrant and excessive electrical discharge in the brain."  As a conceptual definition, for me, this has been clinically useful and practical for teaching students, professionals as well as laymen.

As to the operational definition, we can accept the one that is proposed (except for the part about reflex epilepsy, the relevance of which I cannot comprehend), consider it a work in progress and tweak it as we go through the future with its promise of new knowledge and resources that will sure come our way.

Kalarickal Oommen

23 August, 2013

Congratulations, great job.

-I found the new definition perhaps too longer and with to much items to be considered "practical" for daily-life clinical practice

- As the authors appropriately claim, terms like "provoked" or "unprovoked" are inappropriate and can give rise to misinterpretations of the seizures etiology. Would you consider the change for "at least two seizures of unclear etiology"?

- High recurrence risk – as consireded by the authors, symptomatic lesions generating an enduring predisposition for unprovoked seizures with a similar risk  to those who have had two unprovoked seizures, being so varied, are hard to be enumerated  in a practical definition. So, the second part of the item 2 of the definition, besides being hard to understand "at the first view", does not help, at all, to consider the diagnostic of epilepsy nor the decision of treating or not treating a particular patient.

Futhermore, a "practical " definition is also to be used by general practiciens, at least in precocious evaluation, before refering  the patient to the epileptologist?

I simply suggest "One unprovoked seizure but potential for recurrence"

- Being epileptic brings a lot of social stigmas. One should always try to say "people with epilepsy" and not "patients with epilepsy". Being so, why not to leave "disorder" in place insteady of changing for "disease"

José Pimentel

Return to top

23 August, 2013

We write in continuation of our emails of August 7 and August 18, 2013 in regards to the paper 'An Operational Clinical Definition of Epilepsy'.

We read the eye opening comments of Richard Johnson who is a person having epilepsy. As a person having epilepsy, he is one who can best describe the pain of having faced the stigma due to this disorder. He has very correctly said, "Stigma is the biggest problem we who have epilepsy must deal with, second only to ignorance. He has further emphasized, "It is important that people who have epilepsy are comfortable with the terms used to describe us." He has asked a very pertinent question, "Who exactly was consulted?"

In hind-sight, it would have been appropriate to include in the task force, at least one representative from persons having epilepsy in addition to the representatives from the IBE.

Similarly, the comments of Dr Bruno Z Monnerat from Zambia, "I guess a fellow physician from rural Zambia could have a hard time calculating the odds for a new seizure after the first one... and probably has no clue what reflex epilepsy is all about. It's not his fault. Working definitions must be simple and clear, even if they are imprecise. This is to improve access to care, especially in underserviced areas."

As we have said earlier, the ground realities and socio-cultural milieu in the developing and underserviced regions of the world are sometimes very harsh and totally different from those of the developed world. In order to put forward the view points and problems of the developing and underserviced regions, it is imperative that such a Task Force must have adequate representatives from these parts of the world where a majority of those having epilepsy live today!!

We hope these comments are useful.

Satish Jain
Menka Jain

22 August, 2013

First of all thank you for getting me involved in commenting on this important paper.

My remarks are as follows:

The proposed operational clinical definition of epilepsy sounds very useful, properly and timely addressing the long-standing matter as to how diagnose subjects with one unprovoked seizure.

However I share other colleagues' perplexity about taking 75% recurrence risk of seizures as a cut off for aa diagnosis of epilepsy to be made after one unprovoked seizure. In fact this figure is supported by a robust but limited piece of evidence and its extension to clinical conditions other than the one in which it proved to actually reflect the extent of the seizure recurrence risk, sounds arbitrary. Furthermore in most "one unprovoked seizure" clinical conditions the risk of future seizures recurrence has not so sharply defined yet in scientific literature.

Whatever figure (if any) was chosen, I agree with Dr Petereit that a table listing the clinical conditions in which the risk of recurrence is greater than the chosen cut off would help and would be appreciated by clinicians. More, an overtime up-date about seizures recurrence risk figures according to the emerging pieces of evidence in the literature in various clinical conditions would be welcome.

I agree that the clinical condition of "seizure freedom for more than 10 years off AEDs treatment" should be emphasized from a clinical point of view and that a clinician should be enabled, at this point, to state that Epilepsy is no longer present (would the word "active" fit better?) However I share the concerns by Dr Beran and I agree with him that also a legal opinion should be asked about this matter.

Raffaele Manni

22 August, 2013

Appears to me to be a reasonable working definition.

Yours sincerely
Laurence Bindoff

Return to top

22 August, 2013

I agree with the document and I appreciate the work.

However, the phrase, high probability of further seizures (approximately 75% or more), is a key factor in definition of epilepsy in the paper. unforturnately, the list of risk factors associated with a high probability (≥75%) of future seizures cann't be found. Therefore, the task force should definite which are risk factors associated with a high probability (≥75%).

Shuli Liang

21 August, 2013

I think it important to note that any definition has intrinsic limits.

No definition can be given, encompassing all possibilities, satisfying all needs or covering every specific case. The search for the "ultimate" definition would thus be endless.

In spite of that, I think that the present concepts are useful. Many people will however wonder if a certain undelying condition carries a recurrence risk of 75% or more (point#2). This may be unknown in many instances.

Roberto Cantello

21 August, 2013

As a research scientist I will not comment on "An operational clinical definition on epilepsy" by Dr. Fisher. However, I have one comment concerning the initial treatment of patients following the first "unprovoked" generalized convulsive seizure.
In our studies with experimental animals (rodents) we have demonstrated that a single generalized convulsive seizure lasting no more that 40 sec may be in origin of the amygaloallocortical neurovascular injury leading to long term atrophy observed after relatively long period of time (up to 10 weeks). The seizure-induces atrophy is associated with cognitive impairment.

This observations suggest that in patients as well, the initial convulsive seizure may trigger processes of a neurovascular injury which may remain "silent" for long period of time, and consequently undetected. It is well known that in patients, structural hippocampal and parahippocampal damage which underlies temporal lobe epilepsy is often correlated with cognitive dysfunction. Considering the clinical observations and our experimental findings, we have suggested that, in addition to anticonvulsive treatment, an appropriate clinical evaluation and therapy for seizure-associated cerebrovascular accident should be considered in patients after the initial generalized convulsive seizure.

Thus, understanding epilepsy requires comprehensive clinical and basic science approach as well. An Operational Clinical Definition of Epilepsy presented by Dr. Fisher is an important contribution in his direction. I suggest that similar ILAE task force for basic science in epilepsy should be established considering great number of experimental models and considerable research funds allocated to basic research in epilepsy often irrelevant for clinical application.

Kind regards
Sima Mraovitch

21 August, 2013

Thank you for the work which is very helpful. I have only one comment:

The most unclear point is the recurrence risk of seizures of more than 75% in case of one unprovoked seizure. It is important to stress, as the experts did, that a MRI lesion and/or epileptic discharges on EEG do not automatically satisfy the criteria of more than 75% recurrence risk. For the daily work, it would be helpful to mention here more concrete aspects which may lead to an epilepsy diagnosis even when we do not have strong evidence-based criteria. For example, if the patient describes a clear temporal semiology, temporal spikes on EEG will fit well to a temporal epilepsy. On the other hand, a visual aura combined to e.g. frontal tumor will not lead to the diagnosis of epilepsy.

Yvonne Weber

Return to top

21 August, 2013

Congratulations on a succinct, evidence based and highly readable document. I especially approve of our Re-adoption of the term "disease."  Although the Epilepsy Foundation of America some years ago, took the stance that "Epilepsy is NOT a disease," in order to limit prejudice, it was never clear that societal prejudice was positively impacted by this and led to the nebulous term "seizure disorder" which has no ICDM or evidence-based reality to it at all.  I flinch every time a patient says "I don't have epilepsy, I have a seizure disorder." Moreover, it was Bob DeLorenzo's position that biting the bullet and saying "This is a disease," might garner more public funding of research initiatives.

At any rate we are on the right track with sound epidemiological data and profound thinkers in our field, I am heartened.

Barbara Swartz

20 August, 2013

I guess a fellow physician from rural Zambia could have a hard time calculating the odds for a new seizure after the first one... and probably has no clue what reflex epilepsy is all about. It's not his fault.

Working definitions must be simple and clear, even if they are imprecise. This is to improve access to care, especially in underserviced areas.

My suggestions are:

a) Epilepsy is a lifelong brain disease.
b) For the diagnosis of epilepsy at least two unprovoked, nonfebrile epileptic seizures must occur, more than 24h apart.
b) After 2 years seizure-free, epilepsy is considered controlled.
c) After 10 years drug and seizure-free, epilepsy is considered inactive.

Thanks for your attention and congratulations to the Task Force.
Bruno Z Monnerat

20 August, 2013

This is excellent and logical work. I would like to submit for consideration the term in abeyance for some of the adult patients who are ten years seizure free rather than no longer present.

Thank you to the task force for all of the hard work.

Sheri Cotterman-Hart

20 August, 2013

I have read the article and I think it is an excellent and much-needed elaboration.

The previous definition of 2005 was not sufficiently exhaustive. In this new definition the attention is drawn to a number of issues important for clinicians, including the moment of identifying and initiating a treatment, what is the risk of relapse, remissions, recovery, and the new terms "probable or possible epilepsy, as well as "no longer present".

In my opinion, in the definition "no longer present" – the period of at least ten years without seizures and AEDs is too long for some patients. From a clinician's point of view, an interesting information in this study would be a brief remark regarding the end of treatment (after two, three or five years?).

Undoubtedly, this new definition is more precise and helpful to doctors and to patients suffering from epilepsy.

Best regards,
Barbara Blaszczyk

Return to top

20 August, 2013

With respect to the paper presenting operational definition of epilepsy, firstly, I congratulate the authors for their excellent and important work.

An good operational clinical definition of Epilepsy is mandatory, for scientific and also for the diary clinical pratice of epileptology. As clearly as is it, more usefull it will be.

In my opinion, the second condition, "one unprovoked seizure and a probability of further seizures similar to the general reccurrence risk after two unprovoked seizures (approximately 75% or more)" not looks operational.

As a suggestion, instead to say "risk of 75% or more", a list of clinical and laboratory features that can rise the risk of recurrence should appear.

Regarding the text body, the way of explain using examples, there is nothing to add.

Once more, thanks a lot for this opportunity,
Carina Gonçalves Pedroso Uchida

20 August, 2013

1. I agree with other reviewers. Epilepsy is a disease, rather than a disorder

2. I think the use of the definition described in this study should not be limited for clinical doctors, but also to satisfy the practical use in all other areas, such as in scientific researches, in medical management, etc. Otherwise, if the definition described by clinical doctors is different from that described by researchers, there will be communication barrier between different kinds of people, which would cause potential problems. We didn't agree that the author who proposed the new definition considered it to be only properly used in clinics, but not in research studies. Instead, we think it can also be used in conditional scientific studies in Table 2.

3. Provoked and unprovoked seizures are not easy to be understood, sometimes there is no clear boundary between them.

4. The decrease of Epileptic Seizure threshold is the main pathological mechanism to cause epileptic seizure, compared to persons without this kind of condition. As long as the threshold is decreased, the possibility to induce seizure will be increased and this enduring pathological status intends to cause recurrent seizures. As the condition described above, we can consider the diagnosis of epilepsy under this situation,Hence, we suggest using consistently decreased seizure threshold induced unprovoked seizure to replace the item 1.2.3 in Table 2. For practical purpose, we can define some wildly known factors that can cause decreased epileptic seizure threshold, such as family history, two unprovoked seizures; two reflex seizures, many known repeatedly occurred epileptic syndrome, epilepsy after encephalitis recovery, a single seizure occurring at least a month after a stroke (Hesdorffer et al. 2010) or a child with a single seizure and a focal cortical dysplasia, etc. We need to encourage clinical doctors to public papers on Pubmed, to reveal epileptic markers existing in the constantly decreased seizure threshold status, which can also be practical and applicable in the clinics.

5. What would be that potential risk factor, which is associated with a probability of future seizure higher than 75%?

Best Regards,
Xuefeng Wang

20 August, 2013

I commend the task force and generally agree with the proposed new definitions of epilepsy. I also agree with many of the previous comments and will just reinforce some points about the 75% criteria.

I'm fine with using a 75% recurrence risk following a single seizure as meeting a new definition of epilepsy, given that this is a similar risk for recurrence following two unprovoked seizures. I agree that application of this quantitative criterion should be supported by data, when available. On the one hand, the paper states that "Data must be available to support an approximate 75% risk or more of another lifetime seizure", but at the same time the paper "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." To the contrary, I agree with a previous suggestion that inclusion of a published list or table of conditions or disorders for which there are established data indicating a 75% risk of recurrence (or whatever criteria or threshold the taskforce ultimately agrees upon) following a single seizure would be very useful to help guide clinicians. Rather than expecting all clinicians to make their own assessments of the literature on risks of recurrence for all possible situations, instead it would be helpful and appropriate for the task force to analyze the literature systematically and establish a relatively standardized list of conditions that generally fits the proposed criteria. If the taskforce is unable to come up with a standardized list, then it's probably unreasonable to expect individual clinicians to do so and the new definition becomes arbitrary or less meaningful. Of course, as with all clinical guidelines, any standardized list should be accompanied by the necessary disclaimers. As always the case in clinical practice, it would then be up to the clinician to apply these data or recommendations to their patients in an individualized fashion, as the risk may vary depending on a patient's specific circumstances (e.g., severity of underlying condition/disorder, effect of other comorbidities, genetic predisposition, social/environmental factors).

Michael Wong

20 August, 2013

The task force has done an excellent, thoughtful job in developing this operational definition. The reasoning behind the changes are clearly presented and demonstrate that serious consideration was given to both the need for change and the potential ramifications of those changes. I have no substantive comments regarding the document but note that while the conceptual definition includes "and by the neurobiological, cognitive, psychological, and social consequences of this condition" the operational definition does not. This is not addressed in the document, which may have been intentional, but in some cases these additional consequences of epilepsy can persist beyond "epilepsy no longer being present". Some consideration should be given to discussion of the potential effects of disconnecting these persisting consequences of epilepsy from the diagnosis of epilepsy.

John Messenheimer

Return to top

20 August, 2013

First, well done for producing this excellent piece of work.

However, as a person who has epilepsy, I am strongly against the use of the word "disease". In my experience this word, disease is normally associated with infection. We know that epilepsy is not contageous, something which we already have to emphasise when dealing with ignorance and stigma.

Stigma is the biggest problem we who have E must deal with. Second only to ignorance.

The public happily uses the word "disorder", so what is wrong with simply calling E a brain disorder?

It is important that people who have epilepsy are comfortable with the terms used to describe us. Who exactly was consulted?

Richard Johnson

20 August, 2013

These operational definitions of epilepsy are a real improvement and fit most circumstances.

A few comments:

"At least two unprovoked seizures occurring more than 24 hours apart".
Unprovoked may be a tricky consideration. Endogenous or exogenous factors may increase seizure probability in up to 2/3 of the patients with epilepsy. Some patient may even experience seizures only in favoring circumstances. Further, the effect of such triggering factors may be quantitative. This was the basis of the epileptic threshold measurement by pentylenetetrazole for example, patients with epilepsies requiring lower doses to initiate a seizure.

Why not state "At least two seizures occurring more than 24 hours apart, unprovoked by an acute and reversible cerebral insult or aggression. The seizure probability may however be increased in patients with epilepsy by diverse exogenous and endogenous favoring factors".

Disorder or disease
Personally I do like the term disorder
- it emphasize the functional nature of epileptic activities
- a similar seizure (in term of electrical or clinical signature) may occur either in the course of an epilepsy-disease or in reaction to an acute and transient brain aggression.

Risk of recurrence.
Should the >75% risk of recurrence be considered both for the diagnosis (I feel this is OK) and for the risk of relapse (I feel that a much lower risk should be considered)?

Gilles Huberfeld

20 August, 2013

Many thanks to ILAE for the opportunity to comment on the proposed changes to the definition of epilepsy.

My comment refers to the use of the word 'disease' in the definition of epilepsy. I agree with the comments of others – defining epilepsy in this way is potentially detrimental to people with epilepsy. 'Disease' implies a condition that can be 'caught' – defining epilepsy in this way can only add to the general public's confusion and lack of understanding of the condition. As previously noted, epilepsy is a condition long associate with experience of stigma, defining epilepsy as a 'disease' may serve only to further compound such experiences.

Dr Adele Ring

Return to top

19 August, 2013

The clinical definition has improved. I certainly agree with the term disease instead of disorder.

What will we do in case the recurrence risk is not defined or with the epilepsy syndromes without seizures or with a very low recurrence risk as in Landau-Kleffner or BCECTS?

Katrien Jansen

19 August, 2013

My opinion about the new definition of epilepsy is as follows:

According to Conceptual Definition of Seizure and Epilepsy-2005 Report, the definition of epilepsy requires the occurrence of at least one epileptic seizure. On the other hand, the "two unprovoked seizure" definition has been often adopted not only in epidemiological studies but also daily clinical practice. The new Operational Clinical Definition of Epilepsy can help us how to deal with the problem about one unprovoked seizure. However, in my opinion there are some other situations where we diagnose a patient with only one unprovoked seizure as having epilepsy. Even if a patient experiences only one seizure, we can diagnose him or her as having rolandic epilepsy, Panayiotopoulos syndrome, electrical status epilepticus during slow sleep (ESES) and Landau-Kleffner syndrome (LKS) by clinical symptomatology and characteristic EEG features. A significant number of patients with benign partial epilepsy have only one unprovoked seizure during lifetime and many of them do not need antiepileptic drug medication. Nevertheless they belong to specific epilepsy syndromes. A patient with ESES and a patient with LKS may have only one unprovoked seizure but they show characteristic clinical symptoms such as mental deterioration and aphasia. They can also be diagnosed as having specific types of epilepsy, and definitely need antiepileptic drug treatment.

Best regards,
Yoko Ohtsuka

19 August, 2013

Dear Collegues,

I agree with the proposed integrated definition.

Sincerely yours,
Roberto Cotrufo

19 August, 2013

I appreciate the operational definition of epilepsy and, particularly, the first attempt to provide a definition for epilepsy no longer present. I would like, however, to emphasize a few positive and negative points regarding the two definitions.

Definition of epilepsy

  1. I fully agree on the traditional definition of epilepsy as the occurrence of at least two unprovoked seizures provided that they occur 24 hours apart; I have mixed feelings on choosing a bound for the longest interval (five years is a possibility); however, I may agree on the need to leave it open because there are no straightforward data supporting this cut point;
  2. I don’t see why should we consider reflex epilepsies as a separate entity considering the need to have two seizures for the diagnosis (which is in line with the definition of epilepsy); a note could be sufficient to indicate this entity as epilepsy and not a sequence of acute symptomatic seizures;
  3. I have concerns on the use of the 75% relapse risk for defining as epilepsy a clinical condition characterized by an unprovoked seizure in patients with a fairly high risk of recurrence.; as this definition will be used by physicians mostly involved in primary care, it may be difficult to provide this estimate at the physician’s desk (where a "Yes" or “No” is the answer expected by the patient); I would prefer to use a qualitative definition here; perhaps, the term “probable epilepsy” could work well for all cases with a first unprovoked seizure and an underlying clinical condition potentially increasing the risk of relapse.

Definition of epilepsy no longer present

  1. The 10-year period of seizure freedom off-drugs works well because it is reasonable and is in line with the current European regulations for commercial driving; the European directive could help addressing (and perhaps solving) the legal implications of this definition;
  2. With reference to age-related epilepsy syndromes, I understand the point of those who pretend to identify a time period, but as this interval at present would not be supported by evidence-based findings, I would leave it to the caring physician’s judgment.

Ettore Beghi

Return to top

18 August, 2013

My congratulations and gratitude to the authors re the proposed new definition of epilepsy. I'm not a neurologist but a geneticist involved in epilepsy research, so look to this from a bit different angle. Still, I do concur with by far most statements of the authors.
Comments are:

- a geneticist would not use the term disease to describe epilepsy. Epilepsy is a disorder that can be caused by a disease, but also by other causes. Lay persons do know the term disorder very well, and in fact do know the difference between disorder and disease too. Using the term disease might therefore be confusing. I would strongly favour to return to the term disorder.

- I miss the definition of the term seizure, and also the certainty with which an event can be tagged as being a seizure. It seems essential as the term is used in the definition and as the definition used for seizure is not uniform

- I'm uncertain whether it is wise to state that it should be two 'unprovoked' seizures. It will depend on are ability to recognize events that act as a provocation to a seizure. No doubt this will increase with time, inhibiting diagnosing epilepsy in an individual. One might state such an individual would then have reflex epilepsy. But then I wonder what the use is of making a distinction between 1 and 3 in Table 3, and why it is not simply 'At least two seizures occurring more than 24 hours apart'. Reflex epilepsy can then be discussed in the comments.

- I understand very well why you refrain from a list of all entities with a risk of 75% of above for a life-time seizure: it would be extremely difficult to be exhaustive. But it would help clinicians if a small series of (relatively) common entities that go along with such a risk would be mentioned in the text, emphasizing these serve as examples and it remains the task of clinicians to decide about this in other entities

- the period of time before one may state that the epilepsy is no longer present, is extremely important, not in the last place because of all practical consequences for individuals with epilepsy. I have insufficient knowledge of literature on this point but the data you provide in the manuscript are not convincing to me that 10 years is the best period to choose. More literature data to sustain the choice of 10 years seems needed

Again, my thanks for all work you have done to arrive at the present definition.
Raoul CM Hennekam

18 August, 2013

I certainly enjoyed the erudite article by Dr.Fisher and his colleagues. I do however,have a few comments:

1-Even in the old classification, "seizures with specific modes of precipitation" were considered to be epileptic(reflex epilepsies).

2-There should be more discussion of risk factors after the first unprovoked seizure( focal or generalized neurological findings,abnormal neuro-imaging findings, family history of epilepsy,abnormal EEG,nocturnal occurrence,age of onset).

3-The new study should place greater emphasis on the significance of EEG in determining the presence or absence of epilepsy.

4-In statistics, the occurrence of an event is "probable", when the chances of occurrence are greater than 50%. I do not understand the arbitrary choice of 75%.


18 August, 2013

Have reviewed the proposed revision. Clearly, much thought has gone into it. From a conceptual standpoint the revision makes sense. The 75% number also is appropriate as it matches the risk after two unprovoked seizures. (the data are similar in children– Shinnar et al 2001 after 2 seizures). The " no longer present" is a welcome addition, though a 5 year off medications may be more appropriate.

That said there are a number of problems I see with the practical application of the new proposed definition.

  1. There are almost no groups who qualify after one unprovoked seizure so, while theoretically appealing, not clear that will change much practically.
    1. In adults (Hauser) risk with abnormal EEG is about 50% and with remote symptomatic still less than 75%
    2. In children, risk with an epileptiform EEG about 60% (Shinnar et al 1996, Berg and Shinnar 1991) and with remote symptomatic seizure about the same. It is curious that the extensive pediatric first seizure data is almost completely ignored in this document.
    3. The only subgroup identified with >75% is the very small (<5% of the kids) group of children with remote symptomatic etiology, prior febrile seizures and a first unprovoked seizure. While recurrence risk in them was >90% it is a very small group (Shinnar et al 1991, 1996).
  2. Vast majority of acute symptomatic seizures (which are the other group aside from first unprovoked where the new classification may make a change) also have substantially less than a 75% risk of epilepsy. This includes post-trauma and stroke.
  3. As in the setting of first unprovoked, most recurrences are early on, to some degree this is academic as we will know if they will have a second seizure fairly soon.
  4. Am confused by the detailed discussion of complex febrile seizures that implies that, while not epilepsy, they are getting there. The risk of subsequent epilepsy, if we exclude febrile status epilepticus (using the 30 minute definition) is 10-20% which is not even approaching the 75% mark. For febrile status, it is about 40%. The FEBSTAT study is suggesting that we may be able to identify which ones, but at this point, we do not know yet. The discussion is leaving me with a sense that the authors believe complex febrile seizures are at least a "forme fruste" and this is contrary to the epidemiological evidence. Complex febrile seizures are a lesser risk factor than encephalitis, significant head trauma or stroke. They do warrant special discussion because they are common, but it needs to be emphasized more that the vast majority will NOT develop epilepsy.
  5. Under implications for treatment – if this is to be discussed it should be clearly stated that the data from several well designed large scale epidemiological studies (the Italian first seizure study and the UK MESS study) show that while treatment reduces the risk of seizure recurrence it does not influence long term outcome.
  6. Of the available data on "remission", the most rigorous are for 5 years off medications from several studies. There are very few data on 10 years off meds remission. So, while I support the new "no longer present" definition and agree it is different than a cure as some of these do relapse, am confused by why 10 years was chosen which should be explained better. Certainly, in the example given of BECTS that is not necessary and clinically, likely a bit too stringent. Keep in mind that ILAE requiring 10 years off meds before saying "no longer present" may have negative impact on the ability of people who had childhood seizures and who outgrew them in adolescence to serve in the military or other occupations where "no longer present" is a requirement.

Hope these are helpful,
Shlomo Shinnar

Return to top

18 August, 2013

I write in continuation of my earlier email in regards to the paper 'An Operational Clinical Definition of Epilepsy'.

My comments are as follows:

1. Disease:
The paper says, "Epilepsy has traditionally been referred to as a disorder or a family of disorders, rather than a disease, to imply that it is comprised of many different diseases and conditions. 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".

The main contention of the esteemed members of Task Force is that 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".

There can be certain advantages of retaining the term 'disorder' as compared to 'disease' in the new definition. One recent example will bring forth what we want to convey.

We quote a press coverage in the Times of India (today, dated August 18, 2013).

Diabetes no grounds for denial of job: HC by A Subramani, TNN.

Chennai: Can diabetes and a blood sugar level of 140 mg/dl be cited as reasons to deny employment to a person?

No, the Madras High Court has said, adding "There cannot be a blanket ban on giving public employment to persons with diabetes. Medical experts opine that diabetes is a condition where the body fails to utilize the ingested glucose properly. Further, there is a strong school of thought that a diabetic is not suffering from a disease, but only a disorder that could be managed."

Manikandan, a clerk in Southern Railway, died in 2010. His wife Arockiammal applied to the railways for a job, but she was told she was unfit as her blood sugar level was 140 mg/dl. She moved the Central Administrative Tribunal, which directed the railways to appoint her. The railways then moved the HC.

Defining all people with epilepsy as having a 'disease' can have medico-legal implications and also lead to social discrimination. As highlighted by the recent High Court ruling mentioned above, that "a diabetic is not suffering from a disease, but only a disorder that could be managed." Going with the High Court's judgment it can be proposed that those diabetics whose blood sugar can be controlled can be considered to have a 'disorder' while those whose blood sugar can't be controlled with adequate oral medications or those having insulin-resistance may be classified as having a 'disease'.

In view of the above, we suggest that all those having epilepsy, whose seizures are controlled on anti-seizure drug(s) should be classified as having a 'disorder'. The term 'disease' could be reserved for only those whose seizures are difficult to control or are having a type of drug-resistant epilepsy.

We hope these comments are useful.

Satish Jain
Menka S. Jain

18 August, 2013

It is a great work what the task force is doing.

I think that the definition of two or more Seizures plus and the Other posibility of one seizure but a risk of 75% improves the definition we have these days.

I have more doubts with the term "no longer present" and not to use "cure "... Perhaps because I am a child neurologist. I think if the risk is so low after X number of years we must consider the term "cure".

Thank you all of you
Clarisa Maxit

Return to top

18 August, 2013

I have read the paper. Congratulations, It was an excellent work.

In my opinion when a child is seizure free for more than 10 years, he or she has not a recurrences risk of >75 %.

With best regards
Parvaneh Karimzadeh, M.D

18 August, 2013

I appreciate and congratulate the task force for defining epilepsy in more convincing way. My doubt is about the 'partial seizures' which are less symptomatic, but occurs recurrently, how will authors consider this aspect while defining epilepsy.

with regds

16 August, 2013

ESES is an excellent example of why a number of seizures should not be included in the definition of Epilepsy.

If you diagnose a patient with acquired epileptic aphasia, etc, who has not had motor seizures, still the diagnosis of epilepsy is warranted.

One should not make the point that the aphasia is a behavioral paroxysm.

Seizure: paroxysm of abnormal behavior.

Epileptic seizure: paroxysm of abnormal behavior secondary to abnormal, excessive, hypersychronized neuronal activity.

Epilepsy: Medical condition characterized by a tendency for recurrent un provoked epileptic seizures or epileptic seizures provoked by a stimulus by medical consensus not accepted as a precipitant for epileptic seizures.

Erick Sell

16 August, 2013

This is a great effort to make it more practical definition of epilepsy, aiding in decision making for the treatment of patients with epilepsy.

I agree that if Epilepsy is a brain predisposition to the generation of unprovoked seizures, then it is a disease, more than a "simple" disorder.

Still remains the issue of determining more accurately the individual risk of recurrence of a particular patient, since the number "75%" is somewhat elusive in daily practice. I think the level of evidence used to support the new definition need not appear in the definition, and this should be at the conceptual level, as "especially high risk of recurrence."

Epilepsies with low rate of seizures, such as Landau-Kleffner, would still be included in the new definition, since they were kept in the session of "special situations" of the Classification of the Epilepsies.

All the best
José Augusto Bragatti

Return to top

16 August, 2013

I have read with great interest the paper about "An Operational Clinical Definition of Epilepsy" that the task force has done.

I have some uncertainty about when the epilepsy have to be considered "no longer present" reported in Table 2.

I am worried that this sentence , included in Table 2 , may create some misunderstanding even if the Authors have clearly explained the concept in the whole text of the article .

I think that, if a definition of "no longer present" was made, it is essential ,to avoid misunderstanding ,provide some concepts for other situations such as the "surgically remediable epilepsy syndromes".

In fact , the definition of "no longer present" reported in Table 2 (that is the take-home message) extrapolated from the analysis of all the text, may have advers effect on motivation for epilepsy surgery referral both for physicians, neurologists, families, and patients.

I think that may be useful and clarifier to add in Table 2 the "key" sentence reported in the text:

"Clinicians will have to individualize other circumstances in which epilepsy is definitively no longer present or cured".

With Best Regards
Marco Giulioni

16 August, 2013

One challenge will be the 75% chance of another seizure. The example you give at the end of the stroke patient who has a greater than 70% chance of a recurrent seizure after one post-stroke already violates your rule? If they need a 75% chance, how can >70% equal ≥75%?

It is often hard to give a precise risk. Rather, there are ranges and criteria from literature studies may not be precisely applicable to specific patients. So at the end, what I call epilepsy and my colleague in the room next door may differ for the same patient. Understand a specific percentage is meant to avoid ambiguity, but it creates another set of uncertainty.

I agree with the concept, but would probably have chosen some thing like 'greater than a 50% chance of a recurrent seizure' – but there is no right or wrong here. So fine with the 75%. Might try to give a lot of specific literature citations to identify cases that clearly hit this mark.

Orrin Devinsky

16 August, 2013

I agree with a previous comment regarding the percent probability of seizure recurrence, which seems arbitrary. I would argue that any chance of seizure recurrence higher than the general population be considered, as some patients are treated with AEDs for seizure recurrence probabilities that could arguably be less than that (for many patients, there's not enough literature to even decide on a percentage risk). Then being unable to diagnose them with Epilepsy or probable epilepsy will cause problems getting insurance approval for treatments. I think it is better the determination rest with the physician for each individual patient.

Carla F LoPinto-Khoury

16 August, 2013

I am in general agreement with the document and I appreciate the work.

1) However, I disagree with the following statement: "Data must be available to support an approximate 75% risk or more for another lifetime seizure."  Why must we "prove" that the risk is greater than 75%.  On which data will we relay?  What if I believe that the diagnosis and treatment of epilepsy is indicated but cannot prove this risk >75%?  Am I at legal risk given the ILAE statement?  It is the clinician and patient/family who decide on the degree of acceptable risk!  The printing  of percentages and cut-offs do nothing for the art of medicine (insert epileptology).

2) In regard to epilepsy no longer present:   Why 10 years?  Is the recurrence risk not quite low after 5 years?  In the absence of data, is it not the clinician's (perhaps insert epileptologist's) decision as to whether a remission (from epilepsy) has been achieved?  The 'number of years' required for "epilepsy no longer present" or "remission" will not be the same for every patient.  I agree that one must consider the benefit and harm (including legal) of even having the designation "epilepsy no longer present."

3) Infantile spasms is a seizure type that does not require 24 hours between spasms for the diagnosis of epilepsy.

John R. Mytinger

Return to top

16 August, 2013

Congratulations for your work!!! I think is a very interesting and useful tool for our daily job...

I have have some comments:

1. The first one is about the 75% recurrence risk mentioned to make the diagnosis of epilepsy: How would that apply to certain epilepsy syndromes in childhood (BECTS for example): a boy with a focal seizure involving the face and the arm out of sleep, which EEG shows rolandic spikes has epilepsy in terms of the "old" definition, which to my opinion is correct. What do you think about it?

2. Do you think it could be interesting to mention something about pharmacological treatment duration? I mean some advertisements, I think it could be interesting for physicians...

Salvador Ibáñez-Micó

16 August, 2013

I think it is very precise and important change in the erstwhile definition. It helps clarify many doubts in specific situations like Reflex epilepsies and Age dependent epilepsy syndromes. The proposal of defining a "cure" from epilepsy is also unique.

Pamela Correia

16 August, 2013

I like the present definition in view of current scenario, however, I have certain reservation.

1.The term 'probability of further seizures similar to the general recurrence risk after two unprovoked seizures' will be subjective and depends on the understand-ability of various physicians. It should be more objective in nature.

2. A period of 10 years universally may be too long in certain cases. It has been seen and documented that post-five years seizure free period after stopping medication is sufficient enough to declare a case as 'no longer present'. Moreover, patient would like to hear from the physician 'cure' rather than 'no longer present'. Even after 10 years if we say to patient 'no longer present', it may demotivate patient but also physician.

I think we should be more objective in defining epilepsy.

Sonu Goel

16 August, 2013

Great improvement. These definitions for epilepsy are acceptable to me.

Yanlei Huang

16 August, 2013

Thank you for sending this positioning statement on definition of epilepsy. As a general neurologist in practice what you propose will help with some inconsistencies. I have always favoured using statistical risk in discussions with patients - it might be worth finding out from patients what level of risk they think seizures should be treated - if they agree that a 70% risk or higher is sufficient for treatment you have an additional rationale for this cut off - I suspect if you asked patients they might want a different level of risk!

Wishing you every success with your work
Raeburn Forbes

Return to top

16 August, 2013

Generally, operationalizing a diagnosis as much as possible is essential in a scientific context and improves reliability of communication.

I appreciate wide parts of the new definition,but I would like to point out that the introduction of a "75 %" seizure recurrence probability as a necessary precondition for calling a state epilepsy may lead to practical problems and in my view is not helpful.

First, this cutoff value is not evidence-based, and it is unclear why it should not be 50% or 80%.

Second, in many cases the necessary information is not available, thus prohibiting the diagnosis of epilepsy (case 5) even if considered appropriate by the expert physician.

Third, prohibiting a diagnosis of epilepsy in patients due to a lower or unknown risk for future seizures may have unwanted medicolegal consequences. It is correct that scientifically, diagnosis and treatment can be separated. In practice, antiepileptic treatment in patients not fulfilling the criterion of having epilepsy may render coverage of medical and other treatments by insurances problematic, and may preclude patients from access to social support.

I would thus suggest to replace this strict criterion to a more pragmatic approach, leaving the physicion the option to judge if an increased recurrence risk for seizures is sufficient to justify the term epilepsy,
even if exact data are not available or if the rist of seizures may be below 75% according to available evidence.

Andreas Schulze-Bonhage

16 August, 2013

Firstly, I congratulate the authors for their excellent and important work.

A good operational clinical definition of Epilepsy is mandatory, for scientific and also for the diary clinical pratice of epileptology. As clearly as is it, more usefull it will be.

In my opinion, the second condition, "one unprovoked seizure and a probability of further seizures similar to the general reccurrence risk after two unprovoked seizures (approximately 75% or more)" not looks operational.

As a suggestion, instead to say "risk of 75% or more", a list of clinical and laboratory features that can rise the risk of recurrence should appear.

Regarding the text body, the way of explain using examples, there is nothing to add.

Once  more, thanks a lot for this opportunity,

Best regards,
Carina Uchida 

16 August, 2013

I found it very instructive to read the manuscript on proposed new definition of epilepsy. Here are my thoughts:

"Cure" is defined as a state when the probability of seizure recurrence is no greater than that for general population. When epilepsy is no longer present in a patient according to new definition- even in that state he has some baseline risk of having a seizure- which will be called a recurrence- since he has had epileptic seizure earlier. This risk is "baseline" risk conferred on him by being a part of the general population. Hence, there is virtually no difference between the terms "cure" (which has not been elaborated further in the document) and "epilepsy no longer present".

Extending the same argument, the time point of 10 years for the definition of "epilepsy no longer present" has not been sufficiently justified. In the Lossius et. al. (Epilepsia 2008) paper quoted in the document, the 12 month seizure recurrence rate of 15% in withdrawers was not statistically significantly different from non-withdrawers (7%). At the end of the double-blind period, the combined group who chose to taper their AED had declining monthly risk of seizure, with a figure of 0.003 at the end of 36 months. Hence, I think that to wait 10 years before calling a patient "cured" or "epilepsy no longer present" (in either case, a 'baseline' risk of seizure occurrence/recurrence which is not higher than general population) is a bit excessive and the same endpoints may be defined to have occurred earlier at 4-5 years.

Ravindra Arya

16 August, 2013

I see the definition as a work in progress and this is a substantial advance. Three things:

1. I agree with Christian Korff – the 75% seems a bit arbitrary and may not precisely work in all circumstances. Rolandic Epilepsy is a good example.

2. It might be good to remind the reader that when evaluating the child with a single event the specifics of the EEG features should match the clinical presentation before diagnosing epilepsy. Not every epileptiform discharge is the same – some degree of rigor is required before asserting the child has epilepsy. The two seizure rule was a good safety valve in this regard. I am worried about too many children getting treatment for "epilepsy" that they don't have based upon non-specific EEG features found in the setting of one paroxysmal event. The EEG features must match the clinical semiology.

3. I understand the important concept, but the specific application of the outgrowing of the electroclinical syndrome is a little problematic and needs some clarification. When does it apply? The example of Rolandic Epilepsy is fine, but what about early onset occipital epilepsy? One might "outgrow" the clinical susceptibility for that, but then manifest features consistent with Rolandic Epilepsy a few years later. Same applies for West. One could "outgrow" West, but then develop LGS.

Best regards,
Doug Nordli

Return to top

16 August, 2013

The task force can be complimented for having brought the definition a big step forward by successfully reconciliating the conceptual definition of 2005 with the old operational definition of the epidemiologists. It is very welcome to see that it is now recognized that epilepsy is a dynamic condition which can undergo developments including remission – an aspect missing in the 2005 definition.
There are, however, a few points which seem to need further consideration.

  1. Definition of epilepsy as a disease:

This is more problematic than it appears at first glance. To have a well-formulated definition that justifies using the same term, epilepsy, for so many heterogeneous conditions is welcome but from there it's a long way to the concept of a unitarian disease. Such a concept is perhaps not what the Task Force intends to present, but the language can easily be understood this way. The attempt to distinguish disorders as functional disturbances "not necessarily lasting" from diseases as "more lasting derangements of normal function (but not always)" is weak and unconvincing. This reads like the description of a continuum rather than a distinction. The patronizing comment that "the public" (whatever that is) poorly understands what a disorder is does not belong into a scientific paper. The decisive consideration is missing that a disease always is qualitatively different from the healthy state whereas epilepsy is quantitatively different because seizures are possible expressions of any organized nervous system and epilepsy essentially is characterized by a low seizure threshold. It's very questionable if this can qualify as a "disease". The Task Force actually fully recognizes this in the following paragraph where also they consequently drop their own recommendation and speak of a "condition" rather than a "disease". Here the text for good reason is contradictory.

  1. Epilepsy not longer present:

It is not quite clear why the Task Force abstained from using the well-established medical term remission, favouring a new concept which seems not to have a precedent in any other field of medicine. It seems doubtful whether the rather awkward concept of somebody having a "disease" which is not present can be accepted by the legal community at which it seems largely to be addressed. I have difficulties imagining myself as an expert witness explaining this to a court.

Furthermore, the task force unfortunately has neglected the well-known fact that many patients who have long been seizure free decline proposals to attempt termination of treatment because they find the risk of a relapse too high, not the least out of fear to lose their job or driver's license. This large group of patients includes without doubt a substantial number of individuals whose epilepsy has fully remitted; only at present it is impossible to identify them. It is difficult to find a solution to this problem, but it should be addressed because it is of course grossly unfair that, operationally, only people can reach the status of "epilepsy not longer present" who at a certain moment accept the risk of an AED withdrawal. This has sometimes grotesque consequences. Thus, driving regulations in several countries request now that a lorry driver is seizure free without drugs for 10 years or more. Even in a small country like Denmark dozens of lorry drivers who have been seizure free for many years and decided to continue with drug treatment to avoid the risk of a relapse and consequent loss of license, now lost it for the very reason that they still are taking drugs. The spokeswoman of the Danish Health and Medicines Authority defending the rule in a public discussion in the media took the viewpoint that these persons' still taking drugs proved that their epilepsy still was active. The Task Force's paper needs to include an unmistakable statement preventing this kind of circular reasoning which damages our patients.

  1. Necessity of lifelong treatment:

It is rather surprising to read: "uvenile myoclonic epilepsy is known to be subject to lifelong elevated risk for seizures (Geithner et al 2012)". The cited paper concludes exactly the opposite: "Life-long AED treatment [in JME] is not necessarily required to maintain seizure freedom". This agrees with the only two other papers which investigated the long-term prognosis of JME (Baykan et al 2008, and Camfield & Camfield 2009). With these three concurrent publications the hypothesis that this syndrome per se requires lifelong treatment is stone dead, and references to JME should be taken out of the paper.

  1. Imperfect information:

If this aspect is addressed at all some more thoughts should be spent on it to reach a more tangible statement. What we have here looks half-finished, a bit wishy-washy, if you forgive me.

  1. Some minor points:
    1. The last sentence of "Epilepsy no longer present": "Hence, the old definition remains the 'default' definition": could this also be said in plain language which everybody understands? I for example don't.
    2. Consequences of the operational definitions, 3rd para: are there examples where AEDs are prescribed to treat seizures without a diagnosis of epilepsy and therefore are not reimbursed? Difficult to imagine.
    3. ditto 4th para: I have never met a person who felt "disenfranchised" and outside "the epilepsy community" by not receiving a diagnosis of epilepsy.
    4. Case examples, nr7: there are quite some patients with a "first" seizure at 70 who have forgotten all about a single seizure they had in puberty. Sometimes this information somehow turns up, sometimes it doesn't. So the diagnosis of epilepsy in this case would depend on the person's memory? Come on!

Baykan B, Altindag EA, Bebek N, Ozturk AY, Aslantas B, Gurges C,
Baral-Kulaksizoglu I, Gokyigit A. (2008) Myoclonic seizures subside in the fourth decade in juvenile myoclonic epilepsy. Neurology
Camfield CS, Camfield PR. (2009) Juvenile myoclonic epilepsy 25 years after seizure onset: a population-based study. Neurology 73:

Respectfully submitted
Peter Wolf

Return to top

16 August, 2013

I think the ILAE task force, under the leadership of Robert Fisher, produced a new operational definition of epilepsy that is sophisticated, more accurate than prior versions and thus very useful.  Guided by their work, my own simplified, one-sentence definition became as follows:

Epilepsy is a chronic brain disease defined by the genetic or acquired predisposition to permit the occurrence of electrophysiologically verifiable seizures at least twice in a period longer than 24 hours either in the absence of conditions known to cause seizures even in non-epileptic individuals or in the presence of sensory stimuli harmless to healthy people.

Nandor Ludvig

16 August, 2013

It's a good things to try to clarify some points on the definition of epilepsy on basis of the lot of litterature but I have some concerns about the great work you did.

The conditions that predispose to the "greater than 75%" of risk. What are they in clinical practice?

The case 2 is about seizure occuring 6 week after a stroke but in the Item "High recurrence risk", it's wrriten : a single seizure plus a lesion ...does not automatically satisfy criteria for this operational definition. 

Could the frame time be precise?

What about other lesions (glioma, meniongioma, ....) accompanied by a single seizure?

What should be the risk for those lesions (> 75 % ? or less)

Do "probable epilepsy" should be councelse as for legal '(driving, ....) issues concerning epileptic patient?

Great job.
Fatai Radji

16 August, 2013

Thank you for asking my opinion on a new operational definition of epilepsy by the ILEA.

It seems to me that the new Epilepsy definition is clearly an improvement.

  • I wonder whether the word disease has a lesser impact than the word disorder. Besides, in some conditions like brain tumors, the propensity to develop seizures is a consequence of the underlying disease i.e. the cancer, and likewise with many other focal brain lesions the ensuing epilepsy may be regarded as a consequence of failing local control mechanisms. For that reason, to define epilepsy as a disorder, condition, or situation rather than as a disease, makes sense.
  • The text rightfully stresses the importance that many individual factor would decide whether or not apply anti-seizure therapy. For that reason, it would be useful to add to table 2, a sentence that a diagnosis of epilepsy does not necessarily imply a need for therapy, as illustrated by the presence of rare, occasional or (very) mild seizures or existing uncertainties on the diagnosis. By doing that the last sentence on when epilepsy is considered no longer to be present, may no longer be necessary.

Charles Vecht

16 August, 2013

The definition is appropriate but in clinical settings some patients may still remain untreated as the diagnosis could be confounded by co-morbid disorders or medications.

For instance - If a patient with early-onset alcohol dependence, treated for withdrawal, were to have an isolated seizure after two months of abstinence then would it remain pertinent to consider epilepsy. (Also provided that there has been no personal history of epilepsy and preliminary assessments reveal no cause for seizures). This is considering the fact that they remain vulnerable to alcohol use during their lifespan.

Similarly there are considerable number of patients who have seizures during various phases of alcohol withdrawal. Among these it may remain inconclusive if all seizures were linked to withdrawal state or some reflect an underlying epileptic diathesis that has been unmasked by alcohol use. Often in clinical practice they are given topiramate as anticraving medication.<.

Vikram Singh Rawat

16 August, 2013

Dear Task Force Members

The new definition of epilepsy seems sufficiently reasonable as far as the term epilepsy is applied in a purely medical context. However, unfortunately, the term is used also for social issues such as driver license as well as health insurance. In not a few cases, people judged as suffering from epilepsy often have difficulties in finding insurance companies to contract with (at least in our country) or, even if they make it, have to content themselves with unfair conditions. In the worst case scenario, they even lose their job. ILAE should be keenly aware of these social consequences possibly resulting from being diagnosed as having epilepsy and, if the concept is really widened, take necessary countermeasures (for example supplementing limitations or reservations under which condition it is recommended to use or not to use) to avoid unexpected collateral social damage among people otherwise (or based on the previous definition) not counted among patients with epilepsy.

Kousuke Kanemoto

16 August, 2013

Thank you to the authors for this helpful piece of work on clarifying definitions of epilepsy.

I was intrigued at the example of the 85 year old man:

6. Two seizures long ago. An 85 year-old man had a focal seizure at age 6 and another at age 8 years. EEG, MRI, blood tests and family history were all unrevealing. He received anti-seizure drugs from age 8 to age 10 years, when they were discontinued. There have been no further seizures. Comment: According to the new definition, epilepsy is no longer present, since he has been more than 10 years seizure-free and off seizure medication. This is not a guarantee against future seizures, but he has a right to be viewed as someone who does not currently have epilepsy.

Clearly he was one of the very first patients to have an MRI scan as it appears that he had it 77 years ago. I would also concur with the suggestion that he has a right to be viewed as someone who does not currently have epilepsy, given that his last seizure was in 1941.....

Duncan Cameron

16 August, 2013

The task force should be congratulated on this work. My only comment is regarding the debate of disease and disorder. As a GP I meet a lot of people who describe their health and illness in a variety of ways and I have never met anyone talking about epilepsy who think it is not serious because a doctor called it a disorder. Likewise I have talked to hundreds of people with diabetes and rarely describe that as a disease. People have to live with a condition (whether epilepsy, diabetes, asthma or whatever) and to call it a diease is stigmatising and suggests to some that it is more like a contagion and cannot be controlled. I would urge keeping the term 'disorder'.

Henry Smithson

16 August, 2013

I have reviewed the paper by Dr. Fisher's group on the definition of epilepsy.


Congratulations to the authors, and thanks for their service,
Francesco Natale.

16 August, 2013

My comments are as follows:

1. In my opinion, the paper has covered two very important aspects from a clinician's point of view: a) when to classify 'single seizure' as epilepsy, and b) to have a provision for considering 'epilepsy to be no longer present.'

2. Single unprovoked seizure: To be able to classify single unprovoked seizure with a high risk of recurrence of an unprovoked seizure (approximately >75%) as epilepsy is indeed a very useful and practical aspect of epilepsy management. Most clinicians would treat such a patient with anti-seizure medication but such patients were not classified as having epilepsy using the old definition. Since we are treating such patients as 'epilepsy', it may be better to classify them also as having epilepsy. However, the data required and criteria for estimating the risk of recurrent unprovoked seizure after the first unprovoked seizure need to be defined clearly so that those can be used by an average clinician. This will avoid unnecessary stigma due to over diagnosis of epilepsy.

3. There is another important point to be considered here. Such a definition that includes single seizure with a high risk of recurrence of an unprovoked seizure (approximately >75%) as epilepsy can have medico-legal implications. By old definition, a person with single seizure was not classified as having epilepsy but the same person now can be classified as having epilepsy by the new definition. There are many career options and daily activities like driving that may be closed to people having 'epilepsy' but not for those having had a 'single unprovoked seizure'. This may have implications even in finding a marriage partner in countries that still follow the practice of 'marriages arranged by the parents'. The situation can get complicated because a large population of 'people with epilepsy' today live in countries where the practice of 'marriages arranged by the parents' is still followed!!

4. Long-interval seizures. In the example given, a 70 year-old woman had unprovoked seizures at ages 15 and 70. EEG, MRI and family history are unremarkable. Both old and new definitions consider this woman to have epilepsy. Despite the diagnosis, many clinicians would not treat because of the infrequency of seizures. It has been very correctly pointed out that should investigations somehow show that the causes of the two seizures were different, then epilepsy would not be considered to be present. In view of this, there is a definite need to define the criteria when a person with two unprovoked long interval seizures may be classified as NOT having epilepsy. This could be done in a manner similar to that has been done for defining 'epilepsy to be no longer present'.

5. Epilepsy is no longer present: The Task Force has defined 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. Taking the example of a girl now aged 27 years who had absences and a single generalized seizure between age 6-7 years and has been seizure free and off anti-seizure medications for the last 12 years. If her parents or she were to disclose that she had epilepsy, then they will not be able to find a partner for her. By new definition, in her case epilepsy is no longer present and she gets married without disclosing her previous seizures history. Unfortunately she gets an unprovoked seizure soon after marriage and then the previous seizure condition also gets disclosed. Such a situation in a country like India will result in prolonged legal battle with the accusation of 'cheating' on the girl's family that many times ends up in a divorce. Similarly, there can be problems with insurance, applying for a job, while choosing a career etc.

6. Finally, I have always maintained that any such Task Force of the ILAE/IBE, must have adequate representation from the developing world where majority of the people with epilepsy unfortunately live today and where the impact of the revised definitions, concepts and possible stigma will have the maximum impact in the near future. The ground realities and socio-cultural milieu in the developing world are sometimes very harsh and totally different from those of the developed world!!

I hope these comments are useful.

Satish Jain
Menka S. Jain

[see continuation, August 18]

Return to top

16 August, 2013

Many thanks for this very careful effort! I have 2 comments.

  1. While the proposed recurrence risk (75%) has the great advantage of quantifying a threshold, as the authors correctly state, no information is (yet) available for several clinical situations. This might generate some problems in medico-legal issues, at least in some geographical settings, for example for reimbursement of AED, or for driving. Maybe, a formulation that leaves some flexibility would prevent this sort of potential issues (instead as: “One unprovoked seizure and a probability of further seizures similar to general recurrence risk after two unprovoked seizures (approximately 75% of more)” , to: "One unprovoked seizure and a probability of further seizures which is considered high for the patient in the given clinical situation; as a broad guideline, similar to general recurrence risk after two unprovoked seizures (approximately 75% of more)”). Alternatively, a clear statement that this is a definition that does not imply medico-legal consequences – at least at this time – could be evaluated.
  2. As some pediatric neurologist correctly point out, the 10-year frame in order to consider an epilepsy “no longer present” may be too long; an example would be a patient with rolandic epilepsy and seizures between age 7-12, which has a normal EEG and no seizures at age 18. When he/sh applies for the driving license, the vast majority of clinicians would agree that this person no longer has epilepsy, but maybe (with the new definition) not the authorities… This could also apply to adult patients, e.g. after epilepsy surgery and a successful withdrawal of medication (with normal EEGs) after several (but not 10) years.
Warmest regards,
Andrea Rossetti

16 August, 2013

I like the new definition. The new defintion helps to precise diagnosis for clinical and scientific purposes. I agree with other reviewers that the issue of 'probable epilepsy' is problematic. The epileptic seizure should be clarified as distinct from non-epileptic behavioral events. The current text seems to indicate that people in the former category should not be labeled as epileptic, yet in many cases these patients have work or other lifestyle restrictions that require some type of diagnostic label. A "probable epilepsy", while certainly not perfect and perhaps fraught with controversy, would at least improve that situation. They d 'agreement with other colleagues that it is better to use the term "epileptic seizure" routinely instead of "seizure", I think that will help people consider the differential diagnosis of seizures e.g. convulsive syncope more readily. They d 'agreement with other colleagues that epilepsy requires a comprehensive approach. Study of the definition also from the viewpoints of social implication and impact is required.

Michele Roccella

16 August, 2013

The purpose of the paper is to provide epileptologists and other specialists dealing with seizures and epilepsy with a practical (“operational”) definition of epilepsy. This sounds very useful, given the need for clinicians to challenge the introduction of a therapy in patients suffering from unprovoked seizures. I read the paper and the comments made until the 15 august.

  1. The experts comprehensively pointed the difficulty in setting a threshold for determining a “high risk” for recurring unprovoked seizures. High probability of seizure was defined as a recurrence risk greater than 75%. This is unclear for me. Does this mean that among 100 people, 75 will have a future seizure? The question is, in which time delay? One year? Life long? In practical terms, an artificial determination of a threshold higher than 75% appears unrealistic. I would prefer the use of a “high risk of”, without more precision. Risk should be considered as “low”, “moderate” or “high”. We all have (and the patients do, also) a representation of what is a “low risk”, a “moderate risk” and a “high risk”, and decision could be challenged on these qualitative considerations, more than wrong and illusory quantitative assumptions.
  2. The experts wrote “a third patient might have photosensitive epilepsy, yet not be considered to have epilepsy because the seizures are provoked by lights”. This contradicts the so-called “conceptual” definition of epilepsy (2005) where it is the “enduring brain predisposition to generate epileptic seizures” which is pointed as the definition of epilepsy. Indeed, Photosensitivity is an “enduring brain predisposition to generate ES”, so could be called “epilepsy”, even if seizures are provoked by lights. This is also an important conceptual point: One should admit that in patients with epilepsy, seizures are not occurring by chance, randomly, but could be related to specific factors which are only too complex to be acknowledged by patients and care-givers. This is an important difference to be underlined between “true” provoking factors (toxic-, metabolic-, structural- acute conditions) and “predisposed brain-specific” provoking factors such as in reflex seizures. It could be argued that reflex epilepsies only represent the emerged tip of an iceberg where all seizures occurring during epilepsy are related to “hic et nunc” provoking factors.
  3. This is why I do contest the “3rd Condition” of definition proposed in table 2. I mean that reflex epilepsy could be regarded as a subcondition of “2nd Condition”, where the sensitivity to some reflex factors (lights, eating, reading...) could be proved to increase the recurrence risk to what is observed after two unprovoked seizures.
  4. This led me to emphasize the need to clearly distinguish between reflex seizures and provoked seizures. To my opinion, the term “acute symptomatic seizure” should be definitively preferred to “provoked seizure”, in order to limit the risk of confusion. “reflex seizure” should not be considered as “acute symptomatic seizure”, since “reflex seizure” reflects a specific form of “enduring brain predisposition to generate seizures” (as mentioned in the manuscript). The Experts pointed the importance of the reversible characteristics of the cause leading to acute symptomatic seizure. Reflex “unprovoked” seizures are, indeed, not related to a reversible cause.
  5. "Antiseizure medicine" is unusual. Prefer “antiepileptic drugs”

Warm regards,
Laurent Vercueil

16 August, 2013

This made interesting reading.

On the whole I think this probably reflects what a lot of neurologists/epileptologists are currently diagnosing in day to day practise. I also think it won't change practise or teaching, but will reaffirm many people's clinical decision making.

I also have a bit of a problem with taking the disorder to a disease status, when as well outlined in the paper the causes of the disorder are many and thus not a single disease state. I think it makes more sense to be a disorder with various influences that can then spontaneously remit and become “no longer present”, then the concept of disease versus cure is not required, what is wrong with remission, which is clearly reserved for cancers and thus has connotations with this. I don't believe that calling something a disease gives it greater gravity, is there evidence that supports this comment? The problem with epilepsy and society is the stigma, driving and work considerations that go with the disorder.

The concept and diagnosis is difficult entirely because of the heterogeneity and expression of the disorder.

Emma Whitham

16 August, 2013

Dear Dr Fisher and co-authors,

The 2005 conceptual definition of epilepsy, i.e. "a disorder of the brain characterized by an enduring predisposition to generate epileptic seizures" was an important and thoughtful progress which required to be translated into an operational clinical definition. In our view, the proposal of the ILAE task force best reflects current epidemiological knowledge and is thus very appropriate. We similarly agree with the proposal to move from a disorder to a disease, as well as for the cautious note regarding relation between diagnosis and treatment. A large part of epilepsy burden derives from undue or enforced treatment, and lowering the threshold for epilepsy diagnosis (which is expected to result from this definition) carries the risk of increasing both stigma and burden of unnecessary treatment (provided we agree that a patient with a 25% to 30% chance of never suffering another seizure might well be better off treatment if he wishes to take that risk).

These issues reflect the gap between the conceptual and operational definition, and might be partly solved by attempting to fill this gap. The conceptual definition suggests a predisposition to generate seizures, but does not translate this predisposition into a risk of having another seizure. The operational definition propose criteria partly justified by statistics (> 75% risk of seizure recurrence), but without clearly relating these criteria to the conceptual definition. In fact, the 75% figure, placed in brackets, seems secondary to the more theoretical (and thus less operational) notion of presenting a risk equal to that of patients who suffered two unprovoked seizures. In our view, it should be the opposite, since all this framework relies on the well-established observation that the probability of further seizure(s) following two unprovoked seizures more than 24 hours apart is about 75%. We need physicians and their patients to be better informed about specific figures in epilepsy. It is preferable to quote an approximate number (about 75%) than to let patients and physicians fantasizing about the implication of having suffered two seizures.

Thus, we feel it would be useful to fill the gap between the conceptual and operational definitions by being more explicit about the link between the two. This could include sentences such as:

"For clinical purpose, the enduring predisposition to generate epileptic seizures is arbitrary defined as a probability of suffering one or several unprovoked seizure in the order of 75% or more"

or/and revised first sentence of the proposed operational definition: "Epilepsy is a disease defined by any of the following conditions, which are all associated with a risk of further seizures in the order of 75% or more".

Best regards and congratulations for the work achieved
Philippe Ryvlin

16 August, 2013

When the new definition is settled on, it seems likely that it will go beyond epileptologists and neurologists and be used by the medical profession generally and also probably by lawyers. The legal implications and consequences may differ in different legal systems and it might be prudent to obtain legal input from various countries.

There could be issues regarding what 'provocation' provokes a seizure. Patients and non-epileptoloists may stretch this idea a considerable distance, e.g. that feeling a little stressed or tired is enough to explain a seizure. I wonder if a word like 'adequate' should qualify 'provocation'.

The proposal does distinguish between 'diagnosing' epilepsy, and 'treating' someone with a seizure disorder, but I wonder if this point should be made in bold type or otherwise emphasised so that not making a diagnosis of epilepsy is taken automatically to mean treatment is unnecessary without further evaluation of the individual situation.

Thank you for providing opportunity for comment.

Mervyn Eadie

Return to top

16 August, 2013

I've attached the Definition PDF with my comments. I also have a few general comments:

  1. I'd like to commend the authors - it appeared they used first-person language throughout.
  2. There is little discussion of magnitude or severity of seizure activity here. (To fully-disclose, I skimmed several sections, so I may have missed it, but...). As science hones in on the functions and dysfunctions of neural synchrony, it seems likely that seizure activity of less severity than that usually considered "epileptic" may be relatively common and related to other problems. It may be useful to make clear why severity is not currently an integral part of the definition. For now, it seems reasonable that it is not.

Thank you!
Charlie A. Davidson

16 August, 2013

Dear all,

My comments are as follows:

  1. Epilepsy is a syndrome or disorder rather than a disease, as stated in the operational definition. There are still cryptogenic seizures and not all etiologies of epilepsies can be found or defined.
  2. Since you stated that the term "unprovoked" is imprecise, why do no use this term?

Shih-cheng Chen

Return to top

16 August, 2013

The document is a positive step forward in evolving an operational definition of epilepsy and the task force is to be congratulated.

With regards to age dependent epilepsy syndromes, the task force has wisely laid this out to be different from adults or less defined constellations, for which a 10 year seizure free off anticonvulsants guide applies. There may be difficulties in determining the upper limit of these age-dependent syndromes. For the epileptic encephalopathies, this is problematic. In the example of West Syndrome, the age of diagnosis is typically in the first year but may extend into the second year, particularly if not recognized by the parents. In patients who respond well to treatment, it is difficult to draw a line when it is possible to say confidently that patients are free of epilepsy.

Even for "benign" epilepsy syndromes, it would be difficult to draw a line. For Benign Rolandic Epilepsy, seizures can occur in the teenage years, many years after initial diagnosis. Thus a single seizure 10-11 years after initial diagnosis and last seizure with BRE could lead to prolongation of the period considered at risk. As mentioned certain epilepsy syndrome have a known prolonged increased risk e.g. JME. Further development for other specific syndromes would be most helpful.

Unwitnessed seizures also form a real problem. Many parents at review will report that they have seen no seizures in their child. However domestic circumstances may preclude detection of all seizures. The same holds true for adults who live alone. Additional modalities may be required to assist in accurate ascertainment to assist the clinician in deciding if a patient is truly seizure free.

Derrick Chan

15 August, 2013

1. It is not clear whether “Epilepsy no longer present” can be interpreted as “cure”. When the phrase is translated into Japanese, it would probably be “cure” (or “no suitable translation”). A term that is understandable to layman (non-epilepsy specialist) should be used.

2. If “Epilepsy no longer present” can only be applied when a person has remained seizure-free for at least 10 years off anti-seizure medicines, then this term can hardly be used in pediatrics. For pediatricians, this could drastically reduce motivation, and explanation to pediatric patients (as well as understanding on the patients’ side) will become much more pessimistic than now. In the first place, the estimated time when treatment would end changes greatly. BECT etc. should be considered separately.

3. This has important links with the period of self-declaration required to obtain a driving license in Japan. Does it mean that self-declaration becomes unnecessary if a person has remained seizure-free for at least 10 years off anti-seizure medicines, or, since it does not imply cure, a person has to self-declare for life? The same applies to patients’ resumé and insurance declaration.

I accept items 1, 2 and 3 of Table 2, but do not agree with the latter items due to the reasons given above.

Epilepsy requires a comprehensive approach. Study of the definition also from the viewpoints of social implication and impact is required.

Dr. Masako Koide

Return to top

13 August, 2013

I like the new definition, but it strikes me as redundant. Referring to Table 2, one could easily argue that point 1 is merely a special subset of point 2. One could therefore potentially define epilepsy as something like “The occurrence of one unprovoked seizure along with a substantial probability (≥75%) of further unprovoked seizures”. (This is meant to be an example, not literal.) This would roll points 1 and 2 together. Tweak it further and you could probably roll the reflex epilepsies into this single sentence too.

In addition, the >75% threshold strikes me as somewhat arbitrary. I personally believe that something like that is a good threshold, but other clinicians may have other views, and this whole business is about predicting the future in a stochastic way that is necessarily imprecise and judgmental. Thus, one could take the definition example I gave above and leave the “≥75%” out of it, thus leaving “substantial probability” to the judgment of the practitioner. This deliberate vagueness would be an acknowledgement of the imprecision and ambiguity that are a part of our (lack of) understanding of seizure generation at the present time.

Scott Mintzer

12 August, 2013

I have read the paper and I have a little problem with the 75% recurrence risk mentioned to make the diagnosis of epilepsy. How would that apply to certain epilepsy syndromes in childhood in which the risk of recurrence has not been well established? I am thinking BECTS for example: a healthy 7 year-old with a focal seizure involving the face and the arm out of sleep, which EEG shows rolandic spikes has epilepsy in terms of the “old” definition, which to my opinion is correct. I may be wrong but I could not find much evidence about the recurrence risk in such a situation. A paper by Baznec in 2005 mentions 66%, which would not “allow” to use the term for that patient after this initial event, according to the new definition. Same applies for syndromes such as Landau-Kleffner, in which the seizures may be absent, or in which the risk of recurrence after an initial event may be low.

A time limit is a good idea.

Warm regards,
Christian Korff

12 August, 2013

It is a great work the task force has done on the definition on epilepsy.
My comments refer to the time defined for “Epilepsy considered to be no longer present … for 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.”

Are there real significant statistical differences in the time lag between 3,5,7or 10 years to declare that epilepsy is no longer present? As a clinician I would prefer to attach to the possible reliable shortest time interval for stating epilepsy is no longer present. The reasons are that may be with shorter time interval definition the psychosocial consequences and burden like prejudice, discrimination, stigma, rights restrictions etc. could be lower.

As an example most cancer patients are told to be free of recurrence after 5 years of successful treatment, and I think those patients and their families are liberated from a conscious or “unconscious” sense of anxiety, and of still being ill. Health can also be defined as the capacity to adapt to a chronic disease (definitions similar to resilience), and may be this adaptation can be fostered with a shorter time definition of epilepsy no longer present than 10 years.

Best regards,
Jaime Carrizosa

12 August, 2013

Thank you for the sophisticated work. In my opinion, the new defintion helps to precise diagnosis for clinical and scientific purposes. The examples are very helpful for clinicians. I would appreciate a table listing conditions with greater than 75% risk of seizure recurrence.
Best regards Dr. H.-F. Petereit

12 August, 2013

I wanted to say that I found the text on “An operational clinical definition of epilepsy” by Robert Fisher and colleagues well done and convincing.
I therefore approve, and am ready to adopt the new definition.

Charles Raybaud

12 August, 2013

There is in the proposal a recommendation to refer to epilepsy as a disease rather than a disorder. The medical reasons seem impeccable, but there are also social /stigma problems with epilepsy that may be exacerbated by this change as to many lay people disease and infection are strongly linked. Have the organisations concerned with social effects on the lives of people with epilepsy been consulted (eg British Epilepsy Association and equivalents in other countries)?

Regards, Jane Juler

12 August, 2013

I have two comments:

  1. How should a patient with probable epilepsy be counseled concerning legal issues? Do all patients with different certainties of epilepsy have the same disease from a legal point of view? Or does treatment per se, even when started in an ambiguous clinical situation, imply an epilepsy diagnosis?
  2. The commonly used definition of having active epilepsy is when an individual has experienced at least one epileptic seizure during the last 5 years and/or is on antiepileptic drugs. The task force proposes that “no longer present” epilepsy is defined by 10 y of seizure freedom off anti-seizure medication. This seems to be a quite conservative approach. Recent data imply that time delayed relapses are very rare after 5 y off treatment (Lossius et al., 2008). The rational behind the proposed 10 year definition of no longer present epilepsy is somewhat unclear to me.

Kristina Källén

Return to top

11 August, 2013

The one area that does not seem to have been fully considered is the important issue of epilepsy, driving and the law. For the individual who now has “epilepsy no longer present” does this mean that on completing a form to obtain a driver’s licence that individual indicates that (s)he does not have  epilepsy.  If it is “no longer” present then it would imply the person does not have it. If subsequently that same person is involved in motor vehicle accident that may kill someone or more than a single person and it is demonstrated that the driver previously had epilepsy, knew the diagnosis but failed to declare it, because it “was no longer present”, is that person guilty of a “false utterance”, has the driver falsely completed the licence application form etc. Has this new definition been subjected to legal scrutiny? Similar comment may arise within the work place in relation to hazardous employment or risk to others. Has this been subjected to legal opinion? Can such an individual deny epilepsy when applying for a pilot’s licence or the right to be front line forces within the military, as the epilepsy “is no longer present”?

There is a case within the Australian context where a bus driver was allowed a commercial licence despite a previous history of epilepsy and ultimately killed himself and others and the coroner threatened the neurologist with a charge of manslaughter on the basis that the neurologist was supportive of the driver’s right to drive the bus. The neurologist claimed that epilepsy was “no longer present” thus indicating that the bus driver had the right to deny the presence of epilepsy and respond in the negative without “false utterance”. In the end the assumption was not legally challenged due to, I suspect, policy decision, rather than setting legal precedence and thus removing the risk into the future. This leaves these questions unanswered.

Roy G. Beran

11 August, 2013

The use of the term disease is potentially harmful - 'disorder' usefully implies possibilities that include a transient disturbance in function with the capacity for the individual to have normal function; disease may have many more harmful meanings to the public - with implications for contagion (diseases are often conceptualized as this by the public) and  that the individual has a more persistent illness (when we all know promotion of the periods of normality between seizures is essential to minimize impact of epilepsy). I assume that the word disease has had to be used as in the current proposed definition the attention to the co-morbid neuropsychobiological consequences of the underlying disorder causing epilepsy have been lost. 

Kate Riney

11 August, 2013

I think it is a helpful definition.

However, I am not sure I can think of any condition in which a patient is seizure free for 10 or more years and still has a recurrences risk of >75 %. What would such a known risk factor associated with a higher probability (>75%) of future seizures be? Certainly not JME (see below)

Under “Epilepsy no longer present:” the sentence and certainly the quote “Juvenile myoclonic epilepsy is known to be subject to lifelong elevated risk for seizures (Geithner et al., 2012).” are wrong. Geithner et al. actually report on 9/31 JME patients that were tapered off AED and 6 of these 9 (66.6% of the 9 or 20% of the 31) remained seizure free. Only in 2/3 patients with recurrence the recurrence occurred after 10 (1) or 38(1) years, So together seizures recurred in only 2 of 9 patients (22%) seizure free off meds for >10 years. Geithner et al actually concluded: “Life-long AED treatment  is NOT necessarily required to maintain seizure freedom.”

Therefore. I would suggest to leave the 2nd part of the sentence out and just put: “ Epilepsy is considered…for at least 10 years off anti seizure medicines.”

Kind regards
Felix Rosenow

8 August, 2013

Great improvment....
However, ONF sentence is missing for some rare situation with epilesy without seizure like Landau-Klefner and related syndromes

Edouard Hirsch

Return to top

7 August, 2013

This comment is coming from a patient’s viewpoint concerning epilepsy. First, I am one of many individuals who think it will be a miracle if epilepsy is ever talked about as much as breast cancer. But we, as patients, are obviously still stuck in a situation where physicians cannot come to an agreement of a “medically correct” definition for epilepsy.

My suggestion is to come up with a short and easy to understand definition, so when people are going to be informed about the definition of epilepsy, there should be fewer problems, if any, about the definition of epilepsy being edited as it goes through the media. How short and simple should the definition of epilepsy be? How about something no more than 2 sentences long, with words that can be clearly understood and easy to remember? Simple and understandable so if a 6th grader with epilepsy develops the encouragement of wanting to do a simple essay report on epilepsy, then the teacher and fellow classmates will have a better opportunity to understand and remember what the primary definition of epilepsy really is.

As for the question about whether or not to call epilepsy a “disorder” or a “disease”; several years ago I did some online search investigating to see what the results would be by using the words “seizure disorder” verses using the words “seizure disease”. Even by using the words “seizure disease”, a lot of the results show the words “seizure disorder”, and other results with the words “diseases that cause seizures”. So my suggestion is to let epilepsy remain as being a “seizure disorder”.

If the ILAE Task Force would like to have something important enough to show how much they care about the lives of people with epilepsy, now and in the future, then please do some investigating about a video titled ‘Diagnosing Epilepsy’ on the ABC News website, which is about a Good Morning America interview that took place back in November 2009. About 5 minutes and 26 seconds into the video, the primary care physician who’s being interviewed states, “Our basic instinct when someone has a seizure, is to try to protect them. So when someone is having a full on grand mal seizure, which is when they’re shaking, first thing to do, old wives tale by the way, people are not going to swallow their tongue. But what you want to do is get a hard object, a spoon or something, in their mouth so they don’t bite their tongue.” My questions are: 1. If people in the field of neurology are repeatedly saying “do not stick anything in a person’s mouth when they’re having a seizure”, then do you suppose the physician interviewed on Good Morning America gave the wrong information about getting “…a hard object, a spoon or something, in their mouth so they don’t bite their tongue” ?; and, 2. If this one physician is not up-to-date on what to do, and what not to do, when someone is having a seizure, do you suppose there are more physicians who are not up-to-date as well? If people working in the field of neurology are serious about not wanting anything inserted inside a person’s mouth when they’re having a seizure, then my suggestions are to not only force ABC News to delete that video from their website, but to also have a neurological specialist who’s experienced in the field of epilepsy, interviewed by an ABC News journalist, to help bring more people, including primary care physicians, up-to-date about what to do and what not to do when someone is having a seizure.

Bruce Moreau

7 August, 2013

I applaud the intention to expand and operationalize the definition of epilepsy in this manner. Overall, I also support the statement as currently written.

I agree with other reviewers that the issue of 'probable epilepsy' is problematic. The current definition seems highly dependent on available evidence. Evidence based medicine is, of course, laudable, but not always practical. Lack of published evidence does not mean lack of recurrent seizure risk, as I am sure the task force recognizes. The current text does not clearly distinguish between 'lack of available evidence regarding risk' and 'evidence that the risk is below 75%'. The current text seems to indicate that people in the former category should not be labeled as epileptic, yet in many cases these patients have work or other lifestyle restrictions that require some type of diagnostic label. A "probable epilepsy", while certainly not perfect and perhaps fraught with controversy, would at least improve that situation.

There may be a discrepancy between the age-related epilepsies and the more general achievement of "epilepsy no longer present".  To achieve epilepsy no longer present, the text states 10 years of seizure freedom off of medications is required. Does this also apply to age related epilepsies, such that somebody with Childhood Absence Epilepsy who has their last seizure at age 11 has epilepsy until at least age 21 (depending on when meds were stopped)? If the time period of seizure freedom is different for age-related epilepsies, perhaps this can be more clearly indicated.

I concur that epileptic seizure should be clarified as distinct from non-epileptic behavioral events.

Gabriel Martz

7 August, 2013

These definition(s) for epilepsy are acceptable to me.

Ruth Henchey

7 August, 2013

I applaud the work of the ILAE Task Force and support their operational definition of epilepsy.  While expanding the conceptual definition of epilepsy to an operational definition, the Task Force implies that they support the existing ILAE conceptual definition of seizure and retain the use of the term epileptic seizure. I completely agree with the term epileptic seizure, especially when contrasted with the term non-epileptic seizure, and recognize the controversy surrounding its use. However, the report does not present the term epileptic seizure outside Table 1, the conceptual definition, and uses only the simpler term seizure in Table 2, the operational definition. I am in the camp that prefers the term non-epileptic seizure to non-epileptic attack and understand the utility of the use of the term seizure by my patients with conversion or dissociation.  Although Table 2 is delightfully compact, I would support indicating that the clinical definition is referring to epileptic seizures.

Robert B. Duckrow

7 August, 2013

An excellent piece of work.

Consider adding tiers of certainty, similar to that used for Multiple Sclerosis:

Definite – Probable – Possible – Unlikely – Definitely not a seizure

And propose criteria for each level of certainty.

David Blum

7 August, 2013

A better definition of epilepsy would have to include the risk for recurrent seizures in a given of period of time. For example, 15% per year on average instead of just stating 75% risk in the future. This is very vague. For example, some patients with primary generalized epilepsy have rare seizures and they may not have greater that 75% seizures in let’s say a year or even two years. But they have epilepsy. Because they have a risk of greater than 75% if you follow them in a longer period of time. So it is important to indicate the time frame for which the risk applies to. Just like the risk of rupture of a large intracranial aneurysm may be 4-5% per year on average and so on.

Erhan Ergene

7 August, 2013

I like the new definition, I remember one of the old ILAE documents trying to describe how provocation was not the same as specific modes of precipitation, but the new criteria makes that clearer operationally.

I agree with the document and would avoid "probable epilepsy" as a diagnostic term, but use "you may have epilepsy" or "have a high risk of developing epilepsy" for those with recurrence risk below our notional 75% in the next 5 years or so. That risk can be managed. In research studies those patients can be grouped as probable epilepsy or high risk of developing epilepsy but it is difficult for the patient to have probable epilepsy as a diagnostic label. The situation is similar but different for probable MS, as MS refers to a pathology or group of similar pathologies, but epilepsy refers to the risk of occurrence of a future epileptic seizure.

If we said risk of a second epileptic seizure 75% or more in the next 5 years that would help, e.g. with an elderly man who had an epileptic seizure age 15 years and 70 years, the contribution of the seizure age 15, I would have thought, to his future risk now aged 70 was very little. I would think almost all his risk was related to the seizure age 70, his MRI and other risk factors evident now aged 70, so if they were low risk he would not now have epilepsy just because he had a seizure long ago. Also events long ago will be more uncertain, with less reliable memory and less well investigated and documented.

I suggest we use the term "epileptic seizure" routinely instead of "seizure", I think that will help people consider the differential diagnosis of seizures e.g. convulsive syncope more readily.

William Whitehouse

7 August, 2013

I think is difficult to apply the second criteria of one unprovoked seizure and a probability of further seizures similar to the general ocurrence risk after two unprovoked seizures, because we don´t have always this kind of information, there can be discrepancies between diferent studies about the recurrence risk or the studies published can´t be applied in other countries which don´t have studies of recurrence risk because de pevalence or other factors in the diagnosis or treatment o seizures is diferente. The 75% of recurrence risk seems to me rather arbitrary

Ignacio Rojas Flores

Return to top

6 August, 2013

I have reviewed the paper by Dr. Fisher's group on the definition of epilepsy.

In brief, well-done. Congratulations to the authors, and thanks for their service.

Ed Trevathan

6 August, 2013

The announcement is expressed beautifully and it is, therefore, a surprise to come across the latter part of this statement: "It is important to note that a single seizure plus a lesion or a single seizure plus epileptiform EEG spikes does not automatically satisfy criteria for this operational definition of epilepsy. Data must be available to support an approximate 75% risk or more for another lifetime seizure. This operational definition makes no attempt to enumerate the conditions that would increase risk for a second unprovoked seizure above the threshold cited above."

The discussion is unconvincing, particularly when it says that  "Data must be available to support an approximate 75% risk or more for another lifetime seizure" will be needed, but leaves open what this might be. Determining what data must be available is said to be the 'task' of the physician. Inevitably this will be interpreted as "Because you (the patient) and I (the physician) are very concerned that you will experience another seizure, and its consequences, and you would prefer treatment, let's assume you need treatment, therefore I think the risk of you having a seizure is over 75%". Better would be to say "single epileptic seizure, the patient has deep concerns and wants treatment, manage as if 'epilepsy' ".

John Willoughby

6 August, 2013

An operational definition should be short, to the point and concise.  For a definition this is overlong and overconvoluted.  The longer the definition the more it invites amateur lawyers and professional lawyers to quibble about whether an actual clinical case falls under the definition.  I would suggest using sentences 1,2,3 of the conclusion preceded by a concise definition of epileptic seizure to remove the confusion of all seizures or paroxysmal events (particularly psychogenic non-epileptic seizures with epileptic events) as the whole definition and publishing the rest as a separate explanatory paper if a fuller discussion of the definition is desired.  I would assume that the definition will be consulted by non-epilepsy physicians, non-physician professionals and even patients for guidance.  Epilepsy professionals might  use a longer, academic discussion of what is and is not epilepsy; whatever the intention it must be kept in mind that the definition will be consulted by communities other than the epilepsy community and this use must be taken into consideration.

William Nowak

6 August, 2013

I welcome what seems to be an improved operational definition of epilepsy.  I have the following comments:

  • the determination whether someone has epilepsy or not after a single epileptic seizure will need the recurrence risk to be routinely articulated.  There will be situations where the risk is not articulated and this will be an indeterminate group.
  • There will also be situations where there is disagreement about this risk.  This will create some difficulties because of the subjectivity of this risk and therefore another indeterminate group.
  • Because of this and the issues discussed under the heading of 'possible or probable epilepsy' I think that it is essential that there is developed a operational definition of uncertainty.  In practice this uncertain group is an important group who require quite distinctive care, advice, investigations and treatment such that the uncertainty is reduced while pursuing an appropriate outcome.

I also conceptually sometimes wonder whether there may be a group where we should diagnose epilepsy but they have not yet had any seizures!!  I.e their seizure risk is high enough to make future seizures inevitable and in need of management.  The emergence particularly of early 'genetic diagnoses of epilepsy' may mean that this needs some consideration in the future! 

Colin Dunkley

Return to top


Home | Contact Us | Privacy & Security | Login | Sitemap

© 2014 International League Against Epilepsy