Routine neurological work-up for adults with unprovoked first seizures should include an EEG and brain CT or MRI, according to new practice guidelines by the Quality Standards Subcommittee of the AAN and the American Epilepsy Society. The panel based its findings on an extensive literature search and included 53 relevant studies or reports.
In a telephone interview with Neurology Today, subcommittee report author Allan Krumholz, MD, professor of neurology and director of the Maryland Epilepsy Center at the University of Maryland Medical Center in Baltimore, described the new guidelines, which were published in the Nov. 20 Neurology (2007;69:1996–2007).
WHY IS EEG IMPORTANT IN THE DIAGNOSIS OF FIRST UNPROVOKED SEIZURE?
Studies support EEG as a useful tool for prognosis and to predict the risk of seizure recurrence. This is similar to findings in children. In particular, epileptiform EEG activity, such as generalized spike and wave discharges or focal spikes, are associated with a greater risk for seizure recurrence. A routine EEG revealed epileptiform abnormalities in approximately one-quarter of patients, and predicted seizure recurrence.
EEGs were abnormal in 12 percent to 73 percent (average 51 percent) of the studies and were significantly abnormal in 8 percent to 50 percent (average 29 percent). Epileptiform activity — spikes or sharp waves — was deemed significant by study authors in patients clinically judged to have a new onset seizure. This yield is substantial. However, it is also clear from the evidence that a normal EEG does not exclude the presence of a seizure disorder. Indeed, on average, about 50 percent of individuals clinically diagnosed with a seizure have a completely normal EEG.
IS IT POSSIBLE TO PREDICT WHICH PATIENTS MIGHT HAVE SUBSEQUENT SEIZURES AND, IF SO, WHAT ARE THE BEST PREDICTORS?
There are some reasonable predictors of subsequent seizures, with the EEG being one, but still the best one may be the history and physical examination. For example, if the history reveals past seizures, as it does on about 50 percent of individuals who have an apparent unprovoked first seizure, this indicates epilepsy, and a high risk for recurrence if untreated. Generally patients who have had more than one seizure merit antiepileptic therapy while patients with a single seizure may not. Moreover, other features of the clinical history and physical and neurological examination, such as the presence of specific generalized or focal neurological deficits, are predictive of higher risks of seizure recurrence.
The exact degree of increased risk for seizure recurrence associated with epileptiform EEG abnormalities varies. In one study patients with idiopathic seizure disorders and generalized spike wave EEG abnormalities had an actual 55 percent rate of seizure recurrence at 60 months follow up; the expected recurrence rate for these patients was calculated in the same study to be less, 48.2 percent, a small but statistically significant difference.
Our own meta-analysis estimated post-test probability of a seizure recurrence in patients with epileptiform EEG abnormalities as 49.5 percent compared to only 27.4 percent in individuals with completely normal EEGs. The data show no significance for other more nonspecific EEG abnormalities, such as focal or diffuse slowing, for predicting seizure recurrence. So it appears that epileptiform EEG abnormalities — spikes or sharp waves — suggest a two-fold greater risk of seizure recurrences. Some abnormalities on brain imaging studies may also be predictive of higher risk for seizure recurrences.
WHAT WERE THE FINDINGS REGARDING BRAIN CT AND MRI?
CT was abnormal in 1 percent to 57 percent (average yield 15 percent) and was significantly abnormal in 1 percent to 47 percent (average 10 percent). These significant abnormalities affected patient management and included previously unrecognized brain tumors, vascular lesions, and cerebral cysticercosis. Two of the studies also indicated that abnormal CT was associated with a higher risk of seizure recurrence, but the other studies did not address this issue.
Almost all of the included studies reported on CT rather than MRI. One reason for this is that some of the studies were older, prior to the availability of MRI. Also, many studies included patients seen acutely in emergency departments, where CT is the procedure of choice because of the speed and ease of obtaining the study and its effectiveness in excluding catastrophic problems that might require immediate attention.
Brain MRI seems to be done more selectively, and most of the MRI studies are not prospective or well controlled, thus limiting them as evidence. However, one MRI study of new onset seizures indicated that the yield of MRI is at least as high and likely higher than CT.
IN GENERAL, THEN, IS NEUROIMAGING AN EFFECTIVE SCREENING TOOL?
In general, neuroimaging can determine possibly important treatable causes of seizures in a significant number of patients, particularly in older patients. For adults with an apparently unprovoked first seizure, the evidence indicates that a brain imaging study, either a CT or MRI, is probably useful. It has a yield of about 10 percent, which may lead to the diagnosis of brain tumor, stroke, cysticercosis, or other structural lesions, and may have value in determining the risk of seizure recurrence.
ARE THERE ANY INSTANCES WHERE SUCH IMAGING IS UNNECESSARY?
The decision to obtain any such test, including brain imaging, should still be based on good clinical judgment. For example, a patient who has a generalized convulsive seizure, no focal neurological deficits, a family history of primary generalized seizures, and an EEG consistent with that diagnosis probably has a primary generalized seizure disorder. In my own opinion, in that situation brain imaging is highly unlikely to show any significant or clinically useful abnormality.
SHOULD PATIENTS HAVE A ROUTINE LUMBAR PUNCTURE?
In the adult initially seen with an apparent unprovoked first seizure, there are insufficient data to support or refute recommending routine lumbar puncture, but lumbar puncture may be helpful in specific clinical circumstances, such as patients who are febrile. Some evidence supports doing a lumbar puncture after a first seizure in patients who are immunocompromised even if they are afebrile.
WHAT ABOUT ROUTINE TOXICOLOGY SCREENING?
Seizures are reported as a consequence of drug intoxication particularly with tricyclic antidepressants, cocaine, and other stimulants. Several studies of emergency department admissions for first seizures, including both acute symptomatic and unprovoked seizures, indicated that about 3 percent may relate to drug toxicity or abuse. Still, a recent evidence-based review by the American Academy of Emergency Physicians Clinical Policy Committee considering management of adults with seizures seen in an emergency department did not find sufficient evidence to recommend routine toxicology screening, and our analysis comes to the same conclusion.
WHAT LABORATORY TESTS SHOULD BE ROUTINE AS DIAGNOSTIC SCREENING FOR UNPROVOKED FIRST SEIZURES IN ADULTS?
Although some abnormal laboratory results are reported in adults with an apparent unprovoked first seizure, there is not sufficient evidence to support or refute recommending routine testing of blood glucose, blood counts, or electrolyte panels. The necessity for such studies should be guided by specific clinical circumstances based on the history, physical, and neurological examination.
WHAT STEPS THEN DO YOU RECOMMEND FOR THE DIAGNOSTIC WORK UP IN THE ER?
The physician in the ER must first determine if the event was actually a seizure; and if so, whether this was the first such event. A complete history, physical, and neurological examination are critically important to determine whether an event was indeed a seizure. The report of a reliable observer can be quite helpful in this regard.
Once an event is judged to have been a seizure, the next goal is to determine the cause. For some patients, the history, physical, and neurological examination are enough to discern a probable cause or provide information to guide the physician to consider other diagnostic testing. Some conditions causing seizures require prompt diagnosis and treatment, and others strongly influence prognosis and affect decisions about starting antiepileptic drug therapy.
In particular, provoked seizures are the result of acute precipitating disorders such as meningitis, intoxications, trauma, or metabolic derangements including hypoglycemia, and may require prompt intervention in the ER to reverse potentially damaging conditions. In contrast, unprovoked seizures may also have causes, but these are not acute precipitating conditions requiring immediate action. Their basis may have no known cause or be due to a known pre-existing brain injury or lesion such as a tumor or stroke.
SHOULD ALL TESTING BE DONE IN THE ER?
For many adults, such as those who return to a normal level of function after a first seizure, all of the tests that we recommend and consider need not be done in the ER. Indeed, some tests — EEG and even brain scans — that we identified as useful for routine testing, may actually be better reserved for on an elective or outpatient setting. In particular head CT is commonly done as an emergency procedure in the ER because it is quick and readily accessible, unlike MRI, which is more of an elective or outpatient procedure; however, brain MRI is likely to give more useful information.
Unfortunately, this issue of what to do in regard to ER testing may be influenced by the harsh medical-legal climate in our country. In my experience, ER physicians and others, including myself, are unduly influenced to do testing that may not always be in the best interests of their patient or based on their own best clinical judgment because of medical legal concerns. To me, that is a serious problem.
A TYPICAL CASE: HOW TO HANDLE A FIRST ADULT SEIZURE
A 36-year-old executive comes to your office with his wife and tells you he had a first seizure eight days earlier. He is in excellent health and is an avid jogger. He says that on the morning of the incident — while running on a treadmill at his health club — his vision seemed a bit blurred and diffused, but he felt well. A physician working out at the same club observed the patient collapse, lose consciousness, and convulse violently. The patient gradually awoke but was confused for about 30 minutes and noted that he had bitten his tongue. That morning he had eaten his usual breakfast and felt great.
He had been taken to a local hospital emergency department (ED) where he had blood tests, an electrocardiogram, urine analysis, chest X-ray, and brain CT. He was told that all these tests were normal and was referred to his primary care physician who then referred him for a neurological opinion. The patient is back to his previous level of function, but notes that his muscles are still a bit sore.
His past history is significant for alcohol abuse as a teenager, but he stopped drinking as a young adult, except for a brief relapse in his twenties. His family history indicates no other relevant problems or history of seizures. He is married and has a month-old daughter. He takes no medications, not even vitamins.
General physical and neurological examinations are entirely normal except for an abrasion on one side of his tongue. He has all the results of his ED visit and tests with him.
You advise him about his condition, the risks of recurrence of his seizure, and the rules and his responsibilities regarding driving in your state. In particular you tell him that he should not drive, and that he should follow his state's rules on this, which require him to report his seizure to the state motor vehicles administration (MVA). The MVA will then determine when he can legally resume driving, and you document this discussion with him in his chart. You schedule him for a routine EEG and MRI and a return visit in one or two weeks. His EEG is completely normal as is MRI.
You inform the patient and his wife that his risk of seizure recurrence with these findings is estimated to be less that 50 percent in the next few years, but there is still some risk of recurrence even after many years. You discuss the benefits and risks of antiepileptic medications for his condition. To illustrate, you remind him of the media reports of the recent seizure suffered by Supreme Court Chief Justice John Roberts after being seizure-free for many years after an apparent first seizure.
After some consideration, the patient informs you that he chooses to remain off seizure medications, but agrees that he would take medication should he have another seizure. He returns to see you in three months and has resumed all his activities, including jogging and running at his health club. At visits six months and one year later, he continues to report no seizures and that he is well.