What is a diagnosis of epilepsy?

At an epilepsy symposium for the general public, the mother of a boy with epilepsy asked me about her son’s diagnosis. He showed me her handwritten copy of a laboratory report indicating a mutation in the SCN1A gene and asked, “What does this mean? Is this my son’s diagnosis?” Her questions are apt, but they also lead to the larger, more general challenges in the diagnosis of epilepsy. What is the significance of a genetic mutation? How does each test contribute to the epilepsy diagnosis? What is the best way to characterize her son’s condition? That is, what is her son’s Diagnosis?

Identifying the type of epilepsy (the epilepsy diagnosis/epilepsy syndrome) is complicated, and identifying the seizure type (the seizure diagnosis) also can challenge the clinician. As such, the patient also faces a challenge in understanding the disorder. Ultimately, the problem is that the epilepsy diagnosis is based entirely on whether unprovoked seizures have occurred and the seizure diagnosis often is dependent of a person’s experience and behavior during a confusing and often scary period that commonly does not last more than a few minutes. The only confirmatory diagnostic test requires recording a seizure, which reflects the lack of a marker or specific change such as occurs in almost every other medical condition. Diagnosing Alzheimer disease can challenge the clinician, but the disease has well characterized changes to the brain’s structure that are the ultimate marker for whether the disease is present. The absence of an epilepsy marker is the central challenge in the epilepsy diagnosis.

Historically, the type of seizure has been based on behavioral features, such as the patient’s sensations, movements, and impairment or loss of consciousness. Based on these features, the seizure diagnoses were defined by experts decades ago and forms of epilepsy were derived based on these seizures, the age of onset, and any concomitant neurologic abnormality. This diagnostic system was invaluable because it created a common language for clinicians and researchers and it gave the patient a name for the problem. I believe this naming process is intrinsically meaningful. Naming is a quintessentially human trait and possibly a human need, perhaps because it gives a sense of comfort. If you were lost while traveling through an unknown countryside and asked a local where you were, wouldn’t you feel some sense of comfort by a reply with the place’s name? This may be followed quickly by a need to know where that place is, but having a name gives a sense of greater security. Similarly, we, as patients, often feel comfort when the clinician gives the diagnosis, even before we hear what that diagnosis means. In contrast, consider the patient’s experience when the clinician says, “I don’t know what this is.” The clinician’s honesty is admirable, but the lack of a name increases concern even if nothing else has changed. Naming has value and it is the crux of a diagnosis.

Obviously, only naming is not sufficient and naming is intended to lead to greater understanding. This understanding is obtained through research to determine the cause, course, and best treatment for a disease. However, research is most successful when it uses a valid diagnostic system and not an arbitrary division of conditions into a collection of names. Accurate research results come from studying a group of individuals with one condition in common, and a diagnostic system that incorrectly combines or separates people according to flawed diagnostic divisions impedes advances. Therefore, diagnostic accuracy both helps the patient now and fosters better understanding that will help all patients in the future. A valid diagnostic system is one that is based on reproducible evidence, not personal opinion or idiosyncratic belief even if it comes from experts. Creating a scientifically valid system is both a specific area of research and an obstacle that must be overcome to move all epilepsy research forward.

The traditional diagnostic system has served us well. It has provided reliable information about prognosis and treatment options. However, new scientific insights has led to critical analyses of the system in recent years and the International League Against Epilepsy has released several revisions to the system, including a major revision this year. “What is wrong with the existing system?” is a common question from neurologists who have used it successfully for their entire careers. Indeed, it works and replacing a working system is a burden. With the traditional system, clinicians make treatment decisions based on whether the patient’s epilepsy is idiopathic (primary, that is, an epilepsy not caused by another condition or accompanied by a problem other than seizures) or symptomatic (secondary, that is, an epilepsy accompanied by other neurologic abnormality) and whether consciousness is lost during the seizure. However, epileptologists interested in the most accurate diagnostic system have been gradually recognizing that the old system is not entirely valid. The idiopathic and symptomatic division is blurred when brain function is examined with closer scrutiny and the question of consciousness impairment is complicated by the mysteries of consciousness. I believe every epileptologist has seen seizures that led to unanswerable questions about consciousness. Essentially, revisions are needed to improve the validity of the system, and this need impacts patient care when a maximally accurate diagnosis is wanted.

We live in a world of growing precision and this challenges our ability to always be accurate. Sometimes, accuracy must be compromised for the sake of utility. Our new medical era has powerful diagnostic technology, including genetic testing. The mother who approached me had the results of a gene test that is relevant to epilepsy and accurate for her son, but it is not his complete epilepsy diagnosis. The genetic results may impact her son’s treatment, but the treatment is still based primarily on the old seizure diagnosis system. In a way, the identification of a mutation resulted in a diagnostic island in an archipelago that includes other aspects of his epilepsy diagnosis. Hopefully, current research will bridge the islands by bringing the test results, including electroencephalographic, imaging, genetics, and others, into a more unified understanding. This would be a great step toward better treatments. At this point, we need to be clear about the limits of diagnoses, open to improvements in our naming, and recognize that the diagnosis of epilepsy may still include multiple diagnoses because the underlying problem needs to be understood much better.