Psychogenic nonepileptic seizures (PNES) is the final diagnosis in approximately 30% of patients presenting to specialized epilepsy centers for intractable epilepsy.^1–3 PNES specifically related panic and dissociative disorders are a (psychiatric) subtype of PNES and are addressed in separate chapters. This chapter deals with other types of PNES, which represent the vast majority of psychogenic attacks seen at epilepsy centers.

The advent of the widespread availability of video-encephalography (EEG) has led to the recognition that many patients treated on the assumption of having intractable epilepsy indeed do not have epilepsy. There are many different etiologies of nonepileptic spells, but this chapter will concentrate on spells with a clear psychogenic origin, which account for >50% of the misdiagnosis of epilepsy⁴ in general referral clinics and up to 90% in tertiary epilepsy clinics.⁵ Commonly called psychogenic nonepileptic seizures, these are spells that clearly have a psychogenic component and are not epileptic in origin. The literature is full of confusing terms, such as hysterical seizures, nonorganic seizures, functional seizures, stress seizures, and pseudoseizures, but PNES has become the preferred term. Such seizures are felt to be a conversion disorder manifested by spells that imitate seizures. They are subconscious in origin and very real to the patient. They are differentiated from the far less common condition of malingering or factitious disorder, in which patients consciously fake a seizure for secondary gain, as well as nonepileptic spells with a nonpsychogenic etiology, such as syncope and movement disorders.

PNES is diagnosed fairly often in the general population, with an estimated prevalence of 2 to 33 per 100,000, making it nearly as common as multiple sclerosis and trigeminal neuralgia.⁶ The diagnosis can be made with near certainty by the use of video-EEG monitoring (VEM). Establishing this diagnosis minimizes the risks from antiepileptic drugs and inappropriate medical treatment. It also helps steer the patient toward more effective treatment and minimizes the wasting of the patient’s and society’s health care resources.

Age of Onset and Gender Distribution

The vast majority of patients with PNES are women between the ages of 20 and 50 years. This is consistent across most large published series^7,8 and even across cultures.⁹ There is a growing literature about both younger and older age groups, but they are distinctly less common. A study of the Veterans Affairs (VA) population found 9.6% of PNES patients are older than 60 years.¹⁰ Patients with late-onset PNES are proportionally more often men, have more chronic medical illness, and may be more likely to exhibit trembling rather than violent thrashing.¹¹ Even though the elderly are a small percentage of PNES, PNES is found in a significant percentage of patients admitted for VEM. In one study of 94 elderly (>age 60) patients admitted for VEM, PNES was diagnosed in 13 (14%).¹² Similarly, in an elderly VA population, 10 out of the 71 (14%) patients monitored were found to have PNES.¹³

Prevalence Rates with Epilepsy

The coexistence of epilepsy in a large percentage of patients with PNES was a common concern in earlier case series. This led to the widespread belief that most patients with PNES also had epilepsy, which resulted in the continuation of antiepileptic medications despite documentation of PNES. Later series have shown 5 to 15% of patients have coexisting epilepsy when strict criteria for the diagnosis of epilepsy is used.^14–16 Another phenomenon that has been well described is the onset of PNES after epilepsy surgery. When this occurs, it typically is within 1 month of surgery.¹⁷ Risk factors include neurologic dysfunction in the right hemisphere, seizure onset after adolescence, low intelligence quotient, serious preoperative psychopathologic conditions, and major surgical complications.^17,18

Natural History

The natural history for PNES is poor for both PNES resolution and overall functioning for adults. Studies have reported seizure resolution in <50% after diagnosis.^19–21 Some authors feel that this should not be the only outcome measured, as even when PNES resolves, overall functioning is low. A British study showed that over 50% of patients remained on disability.²⁰ In the United States, approximately 50% remain employed.²¹ Quality of life measures show lower scores in patients with PNES than those with epilepsy. The primary factors appear to be depressed mood and medication side effects. There were more reported problems associated with antiepileptic medications in patients with PNES than in patients with epilepsy, even though epilepsy patients were on more drugs.²²

The outcome may vary by type of spell. Spells that are less dramatic (catatonic, motionless, or less thrashing) appear to have a better resolution.^20,21,23 Other factors that have been associated with improved outcome include shorter duration spells prior to diagnosis,²³ being female,¹⁹ leading an independent life,¹⁹ formal approach to psychotherapy with behavioral counseling by a team of psychologists and psychiatrists familiar with PNES,¹⁹ absence of coexisting epilepsy,¹⁹ educational status,^20,21 being accompanied to the first visit,²¹ fewer additional somatic complaints,²⁰ and lower dissociative scores on a psychological battery.²⁰ Children and adolescents appear to have a better outcome. This may be because of the shorter duration prior to diagnosis and treatment.^24,25

PNES is a conversion disorder manifested by epileptic-like spells. Research into personality traits and exposure to trauma has identified particular psychological profiles. Patients with PNES consistently have increased depression, but only marginally more than those with epilepsy.^27,28 Some studies show increased anxiety,²⁸ whereas others have not seen this.²⁷ There have been studies that suggest that patients with PNES have increased somatization.²⁷

There is, however, heterogeneity, and a recent publication proposed that there are three clusters of patients with PNES as defined by personality tests. These groups are depressed neurotics, activated neurotics, and somatic defenders.²⁶ The depressed neurotics were described as patients with a high degree of neuroticism, low openness, high agreeableness, and “gloomy” pessimism. They tend to keep peace by following tradition, are quite conflicted about anger, and do not want to show anger in fear of offending others. Their feelings are easily hurt, and they often feel victimized. This group was the largest subtype in this study. The activated neurotics show more extroversion and thus are less gloomy and more likely to be anxious rather than depressed. They also are more likely to show anger. Similar to the depressed neurotics, they are rigid and feel everyone should follow the rules of society. The somatic defenders are average in all facets of the personality testing other than having a maladaptive defense system by somatization in the face of psychological stress. This study suggests that patients with PNES are not a homogeneous group.

Coexisting Conditions

Older reports identified a large proportion of patients with PNES who had a history of sexual abuse, but the proportion in later series is smaller although certainly significant.²⁹ This is perhaps more important in children and adolescents.³⁰ A history of abuse may be more common in the convulsive rather than the limp type of PNES.³¹ Studies using self-report show that patients with PNES report a higher incidence of many types of trauma, both sexual and nonsexual, in childhood and as adults.³² Whether this is a causative factor in PNES is not known. There are high rates of various types of sexual, mental, and physical abuse in epilepsy patients as well.³³

Patients with other medical problems, particularly neurologic disorders such as multiple sclerosis, stroke, head trauma, or previous brain surgery, often have spells diagnosed as epileptic that have been perpetuated without reevaluation. A study of patients with traumatic brain injury diagnosed with epileptic seizures showed that 30% had PNES.³⁴ Thus, the diagnosis of PNES should be considered even in the presence of medical disease if the patient has indicators of possible PNES and a lack of response to antiepileptic medications.

Seizures are usually diagnosed based on witnesses’ descriptions, which are rarely detailed or accurate enough to separate PNES from an epileptic seizure. This is true also when the description is from medical personnel. Because of this, certain “red flags” can be very useful in helping the practitioner suspect the diagnosis. The most important red flags are a failure to respond to antiepileptic medications, spells that occur in front of an audience only, unusual triggers, and associated diagnoses of chronic pain syndromes without physical etiology.

Failure to respond to medication is an important piece of historical information that should lead to consideration of this diagnosis. Studies from specialized epilepsy centers, including both outpatient and inpatient settings, have shown that 25% of patients with pharmacoresistant epilepsy have PNES.² A high frequency of spells that are completely unaffected by appropriately chosen antiseizure medications is unusual for epilepsy and should lead to consideration of nonepileptic seizures and PNES in particular. Occurrences that are mainly witnessed, such as in the waiting room of a physician’s office or medical center clinic, may be predictive of PNES.³⁵ The documented appearance of these spells during unwitnessed moments is unusual, but it does not exclude the possibility of the diagnosis. Specific triggers that are unusual for epilepsy should be asked about because this suggests PNES. Triggers such as emotional upset, pain, certain movements, and particular circumstances should alert the physician to the possibility of PNES, especially if the triggers are highly powerful. Epilepsy triggers are not typically as predictable.

Other medical history that should lead to consideration is the presence of diagnoses such as fibromyalgia, chronic fatigue, chronic pain, and irritable bowel syndrome. In a population referred for refractory seizures, chronic pain and fibromyalgia have a strong association with PNES.³⁵ A floridly positive review of systems suggests somatization and the possibility of PNES. Psychological diagnosis with a history of maladaptive behaviors is also suggestive (depression, particularly with panic attacks, and somatization disorders). Neurologic examination can raise suspicion if the patient has an inappropriate level of concern. Overdramatization during history taking and histrionic behaviors during the exam, such as give-way weakness and tightrope walking, should raise concern for PNES. Certain historical features argue against the diagnosis of PNES. The most important one is occurrence during sleep. Occurrence during sleep is distinctly uncommon, although occurrence out of periods where the patient appears to be asleep (pseudosleep) has been described.^36,37 These patients later reported being asleep, and appeared to be asleep to observers, but their EEGs revealed normal waking activity just prior to the spell. The presence of preictal pseudosleep can be quite specific.³⁷ Of course, sleep disorders should be considered when unusual spells suspicious for PNES occur exclusively during sleep. Serious biting injury to the lateral tongue suggests generalized tonic-clonic seizures.³⁸ It is unusual for patients to have serious injuries related to the PNES episodes themselves, but this is possible, and, unfortunately, iatrogenic injuries and even death may result when medical care for pseudostatus epilepticus has been aggressive. Incontinence when reported is not as helpful, as ∼44% of patients with PNES in one survey reported incontinence during episodes.³⁹ Witnessed incontinence during VEM is unusual for PNES, but it was reported in one patient during a spell of quiet unresponsiveness.³⁹

There is a growing body of literature that borderline or incorrectly interpreted tests perpetuate an incorrect diagnosis of epilepsy.⁴⁰ In our center (University of South Florida Comprehensive Epilepsy Center), 32% of the patients diagnosed with PNES over a period of 2½ years had a reportedly abnormal EEG. When these EEGs were subsequently reviewed, all of the tracings were normal.⁴⁰ This does not reflect the inaccuracy of EEG interpretation as much as the influence of an incorrectly interpreted EEG on diagnosis. Most of the incorrectly interpreted EEGs showed sharply contoured background rhythms, wicket spikes, hypnagogic or hyperventilation hypersynchrony, fragmented alpha rhythms, or unusual fluctuations of the background rhythms. These fluctuations include the extension of the alpha rhythm anteriorly into the temporal regions and the combination of alpha rhythms with temporal theta of drowsiness.^41,42 The consequences of EEG misinterpretation are numerous. False-positive results deter the clinician from challenging the diagnosis, which leads to delay in achieving the correct diagnosis and proper treatment. This is likely a contributor to the usual mean delay of >7 years before PNES is diagnosed.⁴³ This is particularly worrisome, as it has been shown that the longer the delay in diagnosis and treatment, the worse the outcome.⁴⁴ False-positive results also complicate management, as patients often hold on to the fact that their EEG was abnormal as irrefutable proof of the epilepsy diagnosis.