New Seizure Classifications Challenge Old Ideologies
A proposal for a new way to classify seizures would modify the way they have been discussed and organized since 1989.
But the new system – put forth by a committee specially convened by the International League Against Epilepsy (ILAE) – isn’t finding friends in every corner. Its move away from expert opinion to clinical evidence has ruffled the feathers of some highly respected clinicians, who assert that the current system is well organized, logical, and has long been integrated into both education and practice.
The proposed classification and terminology scheme would follow a Darwinian-like “Tree of Life” organization. Seizures would be broadly grouped into generalized or focal, depending on their origin and propagation over hemispheres; etiology (structural/metabolic, genetic, or unknown); and finally, by diagnostic specificity.
In this new scheme, focal seizures are seen as those that originate within a neuronal network but are limited to one hemisphere. Seizures that rapidly activate bilateral neuronal networks are seen as generalized.
Generalized seizures could be further categorized by their clinical presentation (tonic, clonic, or a combination; atonic; myoclonic; or absence), age at manifestation, and into other symptoms (Epilepsia 2010;51:676-85).
The system also advises dropping some terms that could be confusing or emotionally charged. “Benign” doesn’t recognize that real problems may result from the condition, wrote primary author Anne T. Berg, Ph.D. Both “idiopathic” and “cryptogenic” can now mean “unknown cause,” but can also designate highly drug responsive, or symptomatic. “Symptomatic” currently connotes a poor prognosis, while ignoring the concept that all seizures cause symptoms, she wrote in the document.
Dr. Jeffrey R. Buchhalter
“In addition to being confusing, these current terms aren’t necessarily true,” said Dr. Jeffrey R. Buchhalter, who was a member of the classification writing committee and is a pediatric neurologist at the Phoenix Children’s Hospital. “What this new system is all about is separating what we know is true from what we do not.”
The proposed system doesn’t replace diagnostic classifications, Dr. Berg, a research professor at Northwestern University, Chicago, said in an interview.
A diagnosis can be made with the existing terms, like Lennox-Gastaut syndrome or childhood absence epilepsy. Instead, the proposed system attempts to logically organize them according to their fundamental relationships.
“What we propose looks at the individual pieces [of the disorders] separately and then asks ‘How often do these go together?’ The new system would also benefit research, because everyone – no matter what country – would be speaking the same scientific language, she added.
The committee’s proposal is based on scientific evidence, instead of relying heavily on expert opinion, as does the current scheme, which was created in 1989. The current system doesn’t allow for the enormous changes made in understanding what causes epilepsy and how it is best treated, Dr. Hans Luders said in an interview. Any descriptive method needs to include these advances in imaging, genetics, and pharmacotherapy. The new system incorporates those new understandings, but in doing so, it drops some cherished nomenclature.
“What we have now, in many cases, are named syndromes used as diagnostic labels,” said Dr. Luders, a neurologist at Case Western Reserve University, Cleveland. “Syndromes are a constellation of clinical signs that tend to uniformly occur together. As we have become more sophisticated, what is a syndrome or not has become debatable. There are so many syndromes now that memorizing them has become quite confusing,” both for neurologists and students.
“We’re moving away from consensus opinion of thought leaders, and from whom consensus opinion was derived, to evidence-based ideas. And in any evidence review, expert opinion is the lowest tier. It’s below level 4,” Dr. Berg said.
As such, the new proposal is not without its detractors, she said. A March opinion piece in Epilepsia airs some of these issues (Epilepsia 2012;53:399-404). In it, Dr. Chrysostomos P. Panayiotopoulos argues that the proposed classification scheme “does not fulfill its intent to improve the [current] classifications.”
Dr. Hans Luders
Dr. Panayiotopoulos, a neurologist at the John Radcliffe Hospital in Headington, England, is also the first to have described Panayiotopoulos syndrome, a benign childhood epilepsy characterized by autonomic seizures and status epilepticus, and a shifting or multifocal electroencephalogram.
The new classification scheme “has met with considerable protest from several expert epileptologists” and doesn’t incorporate most of the classification changes suggested in the 2010 ILAE report, he wrote. But philosophical arguing hobbled that proposed system, and it was never actually adopted by the ILAE. Because of the field’s inability to agree on the 2010 document, much of the extant classification system is still the same as it was in 1989.
In his editorial, Dr. Panayiotopoulos expressed much concern about the proposed classification of focal seizures according to their manifestations. “Such a proposition defies the essence and the principal of any classification that requires an organization and a common language for communication.”
Neglecting to rely upon expert opinion is a large error, he said. “The ILAE Commission could benefit by asking experts in basic and clinical science to provide a concise statement in their field of expertise as, for example, what are focal, myoclonic, or absence seizures, and their subtypes, their manifestations, and their possible pathophysiology.”
Areas of disagreement should be “identified and stated clearly, with documentation of the reasons for it.”
But this approach could throw the debate back into the turbulent waters of the last decades, Dr. Buchhalter said. The new proposal is just one in a long line of criteria adjustments – some of them accepted into use and many of them not.
“Before 1981, seizure classifications were based on semiology. Beginning then, though, things changed and we started to base classifications on EEG type. Then, as we gained knowledge about genetics and family history, we incorporated that information as well.”
This led to an ILAE uproar – members said the system was too bulky and complicated, Dr. Buchhalter said. “A task force met and said this was just getting too confusing and that everything but indicators of focality or generalization should be stripped out. All the data on what the seizures looked like, family history, and genetics were all thrown out.”
It wasn’t long before the complaints started. “Within 8 years, ILAE took a lot of heat,” because of this oversimplified system. In 1986, there was a “paradigm shift,” with the addition of information, such as developmental status, family history, and imaging findings. “We established measurable, objective criteria that recognized seizure types and syndromes as unique entities,” Dr. Buchhalter said.
In another failed effort to simplify classification, Dr. Luders proposed a five-faceted semiology-based scheme. Seizures could be classified, he proposed, by localizing the epileptogenic zone and by seizure semiology, etiology, frequency, and medical comorbidities.
“All cases can be classified according to this five-dimensional system, even at an initial encounter when no detailed test results are available,” he wrote in the German journal Der Nervenarzt. “Information from clinical tests such as MRI and EEG are translated into the best possible working hypothesis at the time of classification, allowing increased precision of the classification as additional information becomes available” (Nervenarzt 2006;77:961-9).
“I say forget all of these names and syndromes – you don’t have to use them,” he said in an interview. “If you’re a general neurologist or psychiatrist who occasionally sees an epilepsy patient, these labels are very confusing. It’s much easier to handle [with the semiologic system], and this is the way we do it [at Case Medical Center]. It’s the same way that we classify any of the neurological diseases.”
Dr. Luders’s suggestion was likewise unpopular, said both Dr. Buchhalter and Dr. Berg. And all three agreed on one thing: Change of any kind is difficult in a large group.
“When things like this are proposed, you have to factor in the emotional part,” Dr. Buchhalter said. “People get used to concepts and don’t want to let them go.”
Dr. Luders, Dr. Berg, Dr. Buchhalter, and Dr. Panayiotopoulos reported having no financial disclosures.