37
CHAPTER
Psychogenic Nonepileptic Events
Jonathan J. Halford
Nonepileptic events (NEEs) are episodes of altered movement, sensation, or experience resembling epileptic seizures which are not associated with ictal epileptiform discharges in the brain, but which, instead, have a psychological origin (1). NEEs are listed in most diagnostic manuals (2) as a dissociative or somatoform disorder, meaning that it is considered to be an involuntary response to emotional, physical, or social distress. NEE disorder is the most common type of nonepilepsy paroxysmal disorder and 20% to 40% of patients with a diagnosis of presumed epilepsy have NEEs (3–5). NEEs occur more frequently in women (accounting for 80% of cases) and the majority of patients are 15 to 35 years of age (6). Most experts in the United States use the term NEEs to describe the condition since the terms “pseudoseizures” or “spells” have pejorative connotations (7). There is also an ongoing debate as to whether the condition should be termed psychogenic nonepileptic “seizures” versus “attacks” or “events” because of the concern that the use of the term “seizure” may confuse some physicians into thinking that patients with NEEs have epilepsy (8,9).
Studies have shown that the diagnosis of NEEs are delayed by a mean of more than 7 years and that most patients are initially thought to have epilepsy (10). This leads to treatment with antiepileptic medications, which are expensive and may exacerbate NEEs (11) and sometimes the performance of expensive unnecessary procedures such as vagus nerve stimulator implantation (12). This delay in diagnosis also leads to the performance of unnecessary lab tests and diagnostic procedures and to a delay in the initiation of appropriate treatment. The patient may also be subjected to complications from invasive procedures if they present to an emergency department (ED) with a continuous NEE (“nonepileptic psychogenic status”) (13). Early diagnosis of NEEs can lower out-of-pocket and systemwide costs by reducing the need for ED visits, hospitalizations, repeated diagnostic testing, and antiepileptic medications (14). Since many of the patients with undiagnosed NEEs are initially thought to have epilepsy, it is important to consider video EEG (vEEG) monitoring for any patient having seizures which have not responded to an adequate trial of two or more antiepileptic medications.
DIAGNOSTIC EVALUATION
NEEs may be suspected in the clinic on the basis of the patient history. There are several aspects of clinical history that hint that a patient may have NEEs, but the sensitivity and specificity of these clues for predicting whether a patient has NEEs, taken together, are not known. These clues include triggers for the events such as “stress” and “getting upset,” circumstances of the events such as the occurrence in a physician’s office, and features of the past medical history including a history of fibromyalgia or unexplained “chronic pain” (15). A florid review of systems (16) and a long list of medical allergies also suggest NEEs (17). On the other hand, a history of tongue biting suggests epileptic seizures (18).
There are several aspects of patient history that may be misleading. First, it is not uncommon for a patient with NEEs to have a history of EEGs that have been interpreted as showing evidence of epilepsy. This is usually due to the misinterpretation of normal EEGs by general neurologists in private practice who lack fellowship training in clinical neurophysiology (19). So a history of an abnormal EEG performed at an outside institution should not be taken as evidence of epilepsy. Although this is difficult to accomplish, an effort should be made to acquire and review this EEG recording to determine if it is indeed abnormal. Secondly, a history of head trauma is not helpful in making the distinction between NEEs and epilepsy in Veterans (20), and this is probably true in non-Veterans as well (although it has not yet been studied). Third, although there are characteristics to the movements of patients during a nonepileptic seizure that can help distinguish these events from epileptic seizures (if the event is recorded on video), a recent study found that these movement characteristics are not reliably reported by nonmedical eye-witnesses (21). So descriptions of patient movements during seizure by family and friends should be interpreted with caution.
Although inpatient vEEG monitoring is expensive and time consuming, it is the only reliable method for making the diagnosis of NEEs (22). Even with a vEEG recording, distinguishing NEEs from certain types of epilepsy, such as frontal lobe epilepsy, can be difficult and even academic experts do not completely agree on all cases (23). vEEG not only provides a definitive diagnosis in almost 90% of patients but also results in treatment change in 79% of patients (24). Without vEEG monitoring, the neurologist’s ability to differentiate epileptic seizures from NEEs by patient history has a specificity of only 50% (25).
There are many semiological features on video that strongly suggest NEEs. Patients with NEEs frequently close their eyes during a seizure compared to patients with epileptic seizures who usually do not close their eyes at all or for only a few seconds (26). Resistance to the eyes being pulled open by the examiner (“forced eye closure”) is a characteristic of nonepileptic seizures (27). Patients frequently cry during nonepileptic seizures (28), particularly women (29). Ictal stuttering and postictal whispering voice strongly suggest NEEs (30,31). Speech during NEEs tends to contain more emotion than speech during an epileptic seizure, which tends to be more monotone (32). The speech of patients in NEEs is often intelligible and sometimes patients answer questions during nonepileptic events. The speech of patients during epileptic seizures is more often fragmented and composed of meaningless phrases or sounds (32). The patient’s mouth is often open during the tonic–clonic phase of a convulsive epileptic seizure but it is usually closed during a psychogenic seizure (33).
The movements in nonepileptic events are often out-of-phase or side-to-side movements and are often chaotic disorganized thrashing (34). Out-of-phase movements are nonsynchronous among the extremities and/or oriented in multiple directions. The movements in frontal lobe seizures, the most common type of seizure confused with NEEs, often involve vocalization and quick tonic posturing (35). The movements in nonepileptic events tend to wax and wane (or completely stop and return again shortly) and to be less stereotyped than epileptic seizures. Nonepileptic events tend to last longer than epileptic seizures. The classic temporal lobe complex partial seizure lasts 10 to 140 seconds (36), whereas nonepileptic events have been documented to last 20 to 805 seconds (34).
Certain postictal behaviors are more associated with one type of event or another. Patients with NEEs often recover quickly after a seizure is over. This can happen with frontal lobe seizures as well, but it is unusual (32). Certain movements are more common after an event in epileptic seizures such as postictal nose rubbing, postictal cough (37), or noisy and stertorous (snoring or gasping) breathing (38). Postictal confusion is frequently seen after both epileptic seizures and NEEs (39), but patients with NEEs more frequently recall what happened during the event (40,41). Postical headache and fatigue have been reported to be more common with epileptic seizures (42). Refractory interictal headache and other pain syndromes are more common in patients with NEEs (43).
There are several common misconceptions about movements in nonepileptic events. First, pelvic thrusting was once thought to be more common in nonepileptic events, but studies have found that pelvic thrusting is as common in frontal lobe epilepsy as it is in NEE disorder (44). Second, it is also commonly believed that, unlike patients with epilepsy, patients with NEEs do not injure themselves during their seizures. But research shows that more than 50% of patients with NEE disorder have sustained an injury due to their events (45). The type of injury is helpful in differentiating epileptic seizures from NEEs. Excoriations on long bones surfaces, such as the arm, leg, or cheek, are seen in NEEs as opposed to lacerations from epileptic seizures (46). Tongue biting and incontinence were once thought to be specific to epilepsy, but they are reported by up to two-thirds of patients with NEEs (47). Tongue bites are often located laterally in patients with epileptic seizures, whereas patients with nonepileptic events bite the tip of their tongue, lip, or buccal region (26,40).
It is important to understand the limitations to vEEG monitoring in order to avoid serious diagnostic errors. Just because the ictal EEG recording is normal does not mean that the seizure event is nonepileptic. Many simple partial and frontal lobe seizures do not manifest themselves on scalp EEG (48). The ictal EEG may also be uninterpretable or difficult to read if movements generate excessive electromyographic (EMG) artifact. EMG artifact is particularly a problem with some hypermotor frontal lobe seizures. It is important to remember to ask the patient if they retained awareness during their seizure, suggesting that it may have been a simple partial seizure. Frontal lobe seizures that do not show EEG manifestations are typically brief and may involve tonic posturing or hypermotor movements (46). If a seizure occurs directly out of sleep (on vEEG monitoring), it is almost certainly an epileptic seizure, even if there is no scalp EEG manifestation. Sometimes patients with NEEs exhibit events that involve “pseudosleep” during which they appear to be asleep, but the EEG shows wakefulness (49). If a seizure is triggered by a placebo response (such as a saline injection) or by suggestion, it is probably nonepileptic.
Observing what patients bring with them to the epilepsy monitoring unit is also useful in distinguishing patients with epilepsy from those with NEEs. One study has found that patients with NEEs frequently brought a stuffed animal toy with them. In this study, 2.5% of the 834 patients (23 patients) admitted to the monitoring unit brought a stuffed animal. Of these 23 patients, 20 were diagnosed with NEE disorder and 3 with epilepsy. The three patients with epilepsy had a history of psychiatric disorder (50).
Various stimuli have been used to provoke nonepileptic events in patients undergoing inpatient vEEG monitoring. These include body part compression, verbal suggestion, placement of a tuning fork or moistened patch on the skin, intravenous administration of saline, and hypnosis (51). The use of these methods is controversial, because they may involve misleading the patient since a placebo may be used. For example, a patient may be told that they are going to be given an intravenous medication that will bring on a seizure when they are actually being given a bolus of IV saline. This ethical concern can be avoided if the patient is completely informed of the nature of the procedure. But performing additional procedures to provoke seizures in patients with NEEs may not be necessary since photic stimulation and hyperventilation, which are typically performed every day during vEEG monitoring at most centers, frequently provokes nonepileptic seizures (52).
The use of questionnaires to predict which patients have NEEs has had some success, although no questionnaire performs perfectly. A recent study showed that a questionnaire of preictal and postictal features given to witnesses of events showed a predictive accuracy of 84% (53). The most sensitive features in the questionnaire were the presence of “postical breathing loudness” and shorter duration of event to predict epileptic seizures. A questionnaire administered to patients showed that those with NEEs reported significantly greater levels of general awareness and responsiveness and more vivid subjective experiences during events (54). A long questionnaire assessing demographic, clinical, seizure-related, and psychological information was able to predict NEEs with a 94% sensitivity and 83% specificity at one clinic and a 85% sensitivity and a 85% specificity at another (55).
Various serum markers have been studied to attempt to differentiate NEEs from epilepsy. Serum creatine phosphokinase (CPK) concentrations measured 12 to 15 hours after generalized convulsive events, if higher than 160 mg/dl, strongly suggested epileptic seizures (56). One study showed that adults with NEEs had decreased levels of brain-derived neurotrophic factor (BDNF) in comparison to healthy controls but the levels of BDNF did not differ between patients with NEEs and epilepsy (57). An assessment by the American Academy of Neurology concluded that a serum prolactin level measured 10 to 20 minutes postictally could help differentiate convulsive epileptic seizures from a convulsive nonepileptic event (58). Overall, it appears that two measures, CPK and prolactin, are of use, but only after a generalized convulsive event.
Neuroimaging is not helpful in distinguishing patients with NEEs from those with epilepsy. Between 10% and 30% of patients with NEEs are found to have abnormalities on brain MRI. These abnormalities were usually nonspecific gliosis or postoperative defects, but mesial temporal sclerosis was found in few cases. Just because a patient has neuroimaging abnormality associated with epilepsy does not mean that they definitely have epilepsy (59,60). Most ictal SPECT scans are negative in patients with NEEs (61), but they can also be negative in patients with epilepsy, so a negative SPECT scan also does not help much in making the distinction.
RISK FACTORS AND PATHOGENESIS
NEEs tend to occur during the second to fourth decades of life, with only up to 20% of cases occurring after the age of 40 (62). Although the cause of NEE disorder is unknown, studies find that many patients have certain types of stressors that may lead to the disorder. Patients with NEEs are more likely to be obese than patients with epilepsy (63). In children and adolescents, difficulties at school, family discord, and cognitive dysfunction are frequently present. In adolescents, there is frequently a history of depression (64). A history of sexual abuse is very uncommon in children and adolescents (64). Adult women with NEEs also frequently report a history of sexual abuse (65). Three-quarters of patients report some type of traumatic antecedent factor such as sexual abuse (32.5%), physical abuse (26%), bereavement (18.7%), health-related trauma (8.3%), or accident or assault (8%). The stressors in these patients may not be reported initially and may be discovered only later during counseling (51). Janet first proposed the theory that traumatic memories could lead to the dissociation, or splitting off, of nuclei of consciousness, which could occasionally take over a person’s behavior (without the person’s conscious awareness) (66). This phenomenon was first noted in his studies of hypnotism, and it has been reported that patients with dissociative disorders tend to be easily hypnotized (67). But no method for testing this hypothesis has been developed. Although some patients with NEEs may have abnormalities on brain imaging, there is no specific neuroimaging finding associated with NEEs.
Functional imaging studies have begun to reveal areas of the brain that are involved in psychogenic movement disorders other than NEEs. A recent functional MRI (fMRI) study in patients with conversion disorder showed increased connectivity between the right amygdala and the right supplementary motor area, while subjects visualized faces with different emotions in comparison to normal controls (68). This suggested there was a “hyperlink” between brain areas regulating emotion and areas initiating motor movements. Another study showed that patients with psychogenic tremor had decreased activity in the right temporal–parietal junction in comparison to patients with physiologic tremor. This brain region is thought to compare internal predictions with actual events and may explain why these movements are not perceived by the patient as self-generated (69). This type of functional imaging study has probably not been performed on patients with NEEs because subjects are required to lie still while they are in the MRI scanner, which is possible for patients with ongoing psychogenic tremor but not for patients having a nonepileptic event.
TREATMENT AND OUTCOME
The first phase of treatment is communicating the diagnosis to the patient at the end of inpatient vEEG monitoring. Most experts agree that how the diagnosis is communicated is very important to improving the chance for a good outcome (70). Research shows that when patients in the epilepsy monitoring unit are not given information on their diagnosis, they have a higher likelihood of no improvement or worsening of NEEs (71). Little research has been done on exactly how the diagnosis should be communicated to patients, but there is a study that suggests that the diagnosis should be presented in a positive light (6). The following is a method that the author has found useful in presenting the diagnosis to patients.
I usually begin with telling the patient that “I have reviewed all of your seizures that we recorded and I have some good news for you—you don’t have epilepsy.” I go on to explain that epilepsy is a serious and often disabling disease that often does not respond to seizure medications and so they should be thankful that they do not have that. I next explain that they have a condition that we frequently see in the epilepsy monitoring unit, called “nonepileptic events,” and that it is a manifestation of stress or trauma. I explain that the stress or trauma may not have happened recently but could have happened in the distant past. I tell them that stress is manifested in different ways in different people—some people may feel nervous, sweaty, or tremble when they experience stress but others may manifest stress by having these types of “events.” Importantly, I reassure them that I don’t think that they are “crazy,” because I have learned that patients are often concerned with this. But I do explain that patients with “nonepileptic events” often have a history of physical or sexual abuse or trauma as well as other comorbid psychiatric disorders such as anxiety, depression, or posttraumatic stress disorder, which need to be treated. I tell them that they no longer need to take antiepileptic medications since these medications will not stop them from having their events “as you have noticed.” I then let the patient talk about how this news makes them feel for awhile. Often they will have questions. Usually the patient will say that this make sense and talk about past abuse or stressors. Some researchers suggest showing the patient the video recording of their seizures as part of this initial talk (6), although this is not something that I do on a routine basis, I do show the patient the vEEG of their events if they ask.
When explaining the treatment plan, it is important to tell patients that there is no single treatment or “magic bullet” that can stop their events, rather several treatments are usually employed. First, they will need to stop taking antiepileptic drugs (AEDs). This is not just because the medication is not helping but also because the medication is expensive, has potential short-term and long-term side effects, and can cause confusion in other treating physicians who are likely to assume that the patient has epilepsy because they are on an AED. A study of immediate versus delayed discontinuation of AEDs demonstrated that immediate AED withdrawal was not associated with greater risk to patients and improved outcome—patients with early discontinuation of AEDs used less rescue medication and were more likely to attribute their NEEs to their mental state at follow-up (72). Second, they will need outpatient counseling with a psychologist for at least several months. Two recent studies of cognitive behavioral therapy (CBT) in comparison to standard medical treatment have reported improvement in event freedom rates, psychiatric symptoms, quality of life, and psychosocial functioning by the end of treatment (73,74). Sometimes CBT is difficult to arrange as relatively few practitioners are able to provide this. Any type of psychotherapy for these patients is probably better than none. In a recent survey, the majority of physicians involved in the treatment of NEEs stated that psychological treatment was the treatment of choice for these patients (75). Third, outpatient evaluation by a psychiatrist should be arranged. Psychiatric comorbidity is common in patients with NEE disorder. The types of comorbid psychiatric disorders most often present are depression, anxiety, posttraumatic stress disorder, and personality disorders (76). Treatment of these comorbidities could improve the patient’s quality of life and their NEEs. A recent pilot study of treatment of depression with sertraline in patients with NEEs showed improvement in their nonepileptic event frequency (77). Fourth, outpatient follow-up with a neurologist is important so as to communicate to the patient that they are not being cast aside once the diagnosis of NEEs is made and also to make sure that the patient has their antiepileptic medication discontinued and gets outpatient psychological counseling and psychiatric evaluation.
Outcome studies have shown that a certain percentage of patients will have their NEEs stop after the diagnosis is communicated to them. In one study, 50% of 68 patients newly presenting with NEEs were event free after 3 months and 44% after 6 months (78). Another recent study with follow-up at 6 and 12 months showed that 38% of patients were event free and 18% had increased events (79). Factors associated with poor outcome include denial of stressors and psychosocial problems, new somatic symptoms after disclosure of the diagnosis, a history of chronic abuse, and higher rates of depression and personality disorders (80). Other factors associated with poor outcome include female sex and patients drawing social security payments (79). Several studies have suggested that a longer delay to diagnosis is associated with worse outcome (45,47,81), although this was not found in the most recent outcome study (79). Although a certain percentage of patients will become event free after the diagnosis is delivered, long-term results suggest that the decrease in nonepileptic event frequency shortly after the diagnosis is not maintained longitudinally when solely giving the diagnosis is the main intervention (45,82). Therefore, patients should be offered the additional treatments described previously.
