Seizures are caused by abnormal electrical discharges in the brain. Epilepsy is the condition of recurrent unprovoked seizures. The definition of “provoked” here is more precise than in common parlance and refers to acute, reversible provoking factors causing seizures. For example, acute hypoglycemia, alcohol withdrawal, high fever, and medication or drug toxicity are all acute reversible factors that can provoke seizures (Table 18–1). In these scenarios, the brain may be structurally normal, but exposure to the acute provoking factor leads to seizures. When the cause is treated, the seizures typically improve and the patient is not necessarily at risk for future recurrent seizures.

Brain tumors, prior stroke, prior head trauma, prior CNS infection, and cortical malformations can all cause seizures, but these entities are neither acute nor reversible, and so recurrent seizures due to any of these causes are considered unprovoked. A patient with a brain tumor (or any of the prior CNS insults listed above) who has a first seizure has had the underlying potential seizure focus for some time. If no acute provoking factor is present (e.g., infection, metabolic derangement), one may ask why the patient seized on that particular day and not the day, week, or month before? Seizures in this context are considered to be unprovoked because they can occur at any time without any provoking factor, just like the unprovoked seizures of idiopathic genetic epilepsy syndromes. Therefore, patients with recurrent seizures due to brain tumors, prior trauma, prior stroke, prior neurosurgery, prior CNS infection, or any other irreversible underlying seizure focus (see Table 18–1) are considered to have epilepsy and should be treated as such.

Some causes of acute symptomatic (provoked) seizures such as acute stroke or hemorrhage, head trauma, or meningitis can increase the risk for development of epilepsy in the future since they can lead to irreversible brain damage, creating an epileptogenic focus.

A patient with seizure(s) will generally present for evaluation in one of three scenarios:

1. 
   

   After a first seizure (or other type of spell)

   
   
2. 
   

   With a history of seizures (or other type of spell)

   
   
3. 
   

   Actively seizing

When a patient presents for evaluation after a seizure in any scenario, the goals of the clinical encounter are to determine the following:

• 
  

  Was the event truly a seizure? The differential diagnosis for a transient alteration in neurologic function includes migraine, syncope, transient ischemic attack (TIA), cardiac arrhythmia, and psychogenic nonepileptic spell. A clear description of the event by witnesses should be obtained. The clinical features that can be used to aid in distinguishing between these are discussed below.

  
  
• 
  

  Were there any clear provoking factors? A detailed medication and drug/alcohol history should be obtained, laboratory tests should evaluate for potential metabolic or infectious etiologies, and neuroimaging should be considered. If the patient has been on an antiepileptic drug for prior seizures, it should be determined whether the patient is taking the medication(s) properly and consistently. Even patients with known epilepsy should be evaluated for potential provoking factors that may have caused them to have a seizure at that particular moment, such as an infection or a new medication that could lower the seizure threshold or alter the metabolism of their antiepileptic medication(s).

  
  
• 
  

  Was this the first event or have there been others? If there have been other prior events, was the patient ever evaluated for these? If so, was the patient ever on an antiepileptic, and if so, did it help?

  
  
• 
  

  Are there any known risk factors for seizures? These may include:

  
  
  
  
  • 
    

    Prior stroke, head trauma, or CNS infection

    
    
  • 
    

    Pediatric febrile seizures

    
    
  • 
    

    Other first-degree relative(s) with seizures

    
    
  • 
    

    Abnormal gestation, birth, or cognitive development

  
  
• 
  

  Has there been emergence of any neurologic deficit prior to or since the onset of seizure(s) to suggest a focal lesion (e.g., progressive weakness, numbness, visual changes, personality/cognitive changes)?

The clinical manifestations (or semiology) of seizures depend on the area(s) of the brain from which seizure activity arises and/or to which this activity spreads. Seizures are broadly categorized as generalized seizures or focal (also called partial) seizures.

Generalized Seizures

Generalized seizures are characterized by impaired consciousness and bilateral motor manifestations if motor manifestations are present. Motor manifestations can be tonic (stiffening of involved body parts), clonic (rhythmic movements), tonic-clonic (mix of tonic and clonic), myoclonic (brief jerks), or atonic (loss of postural tone causing drop attacks). Rarely, seizures of frontal origin can produce more complex motor manifestations (e.g., bicycling, pelvic thrusting). Self-injury (e.g., tongue bite, shoulder dislocation), bladder/bowel incontinence, and altered level of consciousness commonly occur with generalized tonic-clonic seizures.

A postictal state is common after a generalized seizure, and is characterized by altered consciousness, which can range from confusion to coma depending on the severity and length of preceding seizure activity. In a patient with altered level of consciousness following seizure(s), continued nonconvulsive seizures may be occurring and must be distinguished from a postictal state. Subtle signs such as twitching of the eyes or eyelids, gaze deviation, or twitching of the extremities can be clues to nonconvulsive seizures, but overt clinical manifestations may be absent, and only electroencephalography (EEG) can definitively distinguish between ongoing electrographic seizure activity and a postictal state. Therefore, it is prudent to consider EEG monitoring after prolonged seizures with continued altered state of consciousness.

If a patient is found comatose or confused without clear cause and recovers without specific intervention, unwitnessed seizure with a subsequent postictal state should be considered in the differential diagnosis.

Some generalized seizures produce altered consciousness without motor features, such as absence seizures. Absence seizures are characterized by brief periods of altered awareness in which patients are unable to communicate or engage with the environment. Absence seizures are more common in children.

Focal (Partial) Seizures

Focal seizures can occur with impairment of consciousness (complex partial seizures) or with consciousness preserved (simple partial seizures). The clinical manifestations depend on the origin of seizure activity within the brain and can include focal motor symptoms (tonic-clonic movements, posturing, head and/or eye deviation), focal sensory symptoms (paresthesias that tend to spread over seconds), visual, auditory, or olfactory hallucinations, and/or psychic phenomena such as déjà vu (a sense of already having experienced a new place or event), jamais vu (a sense of never having been in a familiar place or situation), or a sense of fear. The aura that may precede a seizure arises from a focal origin of seizure activity before it spreads to involve more of the brain. A seizure may begin with focal manifestations and secondarily generalize, such that a focal seizure evolves into a generalized seizure.

Postictal weakness (Todd’s paralysis) may occur in the limb(s) affected by seizure activity. If a seizure is unwitnessed and a patient is found with focal postictal weakness, the patient may be initially thought to have had a stroke. Therefore, seizure with subsequent postictal paralysis should be considered in the differential diagnosis of acute stroke and transient ischemic attack (TIA) and vice versa.

When available, eyewitness accounts of what happened during a possible seizure can be helpful in distinguishing seizures from other transient neurologic events. However, in many instances, the distinction remains challenging based on historical clues alone.

Distinguishing Seizure From Syncope

In a patient who presents after an unexplained loss of consciousness, some features of the clinical history may be more suggestive of seizure as opposed to syncope. Although a few tonic clonic jerks or brief posturing may be seen in syncope, sustained tonic clonic activity is suggestive of seizure. Tongue biting (particularly on the lateral tongue) is more common in seizure, but can be caused by syncope if the patient hits the chin/jaw when falling. Urinary and/or fecal incontinence is more commonly seen with seizures than with syncope, although in a patient who becomes orthostatic en route to the bathroom to urinate, the full bladder may empty during syncope. After syncope, patients generally return rapidly to consciousness (unless syncope leads to head trauma causing more sustained loss of consciousness), whereas a more prolonged state of confusion or altered consciousness is more characteristic of seizure with a postictal state. Patients who have had a syncopal episode may recall the feelings of presyncope preceding the event (e.g., feeling of “blacking out,” “becoming warm all over”), whereas patients with seizures may recall a preceding aura (e.g., foul odor, sense of déjà vu) or have no recollection of the event whatsoever. Although some practitioners obtain a serum prolactin level as part of the diagnostic workup to distinguish seizure from syncope, this test is neither sensitive nor specific—it can be normal after seizure and it can be elevated after syncope, so it does not reliably distinguish seizure from syncope.

Distinguishing Seizure From Transient Ischemic Attack

Seizures generally cause “positive” symptoms (i.e., abnormal movements, paresthesias, visual phenomena), whereas transient ischemic attacks (TIAs) generally cause “negative” symptoms (i.e., weakness, loss of sensation, visual field deficits). The symptoms of both seizures and TIAs may arise suddenly and improve gradually over minutes to hours. With recurrent events, seizures are more likely to be stereotyped, whereas TIAs are more likely to be varied (unless the TIAs are all due to a particular stenotic vessel, in which case TIAs may also be stereotyped). If a patient has vascular risk factors, this may lead one to err on the side of evaluating for causes of TIA (see “Transient Ischemic Attack” in Chapter 19) when the clinical history is ambiguous, but prior stroke can also be a risk factor for seizure. Therefore, when in doubt, it is prudent to consider evaluating for causes of TIA and causes of seizure in parallel.

Distinguishing Seizure From Migraine With Aura

Both migraine and seizure can produce “positive” focal symptoms (e.g., paresthesias, visual auras), although those of migraine tend to evolve/spread over minutes, whereas those of seizures generally evolve/spread over seconds. Migraine does not lead to alterations in level of consciousness, although it can cause mild confusion in some patients. A headache is a common (but not universal) accompaniment to migraine with aura, but a headache can also be a component of a seizure aura or postictal state.

Distinguishing Seizures From Psychogenic Nonepileptic Spells (Pseudoseizures)

Paroxysmal spells may be psychogenic in origin rather than neurologic in origin. This does not mean that the patient is deliberately producing the spells. Rather, psychogenic nonepileptic spells (also called pseudoseizures) are generally a manifestation of an underlying psychiatric condition (e.g., conversion disorder). Clinical features suggestive of psychogenic nonepileptic spells rather than epileptic seizures include prolonged bilateral movements with preservation of consciousness (for example, preserved ability to communicate and/or respond to external stimuli; generalized bilateral tonic-clonic seizure activity will be accompanied by altered state of consciousness), forced eye closure during episodes (eyes are generally open during seizures), episodes provoked by stress or other emotional circumstances, and continued events in spite of multiple antiepileptic medications (though this can also occur in refractory epilepsy). However, patients with epilepsy can have both seizures and nonepileptic spells, and patients with underlying psychiatric conditions can have seizures. The only definitive way to diagnose spells as being nonepileptic is to obtain prolonged video EEG monitoring that captures spells and their electrographic correlates to determine if the EEG demonstrates seizures or remains normal during spells.

As should be clear from the above discussion, it can be very challenging to distinguish between seizures and other types of paroxsymal spells based on history alone. The diagnosis of seizures is sometimes even difficult when the clinician is observing an event or observing abnormal movements in a patient in coma. The gold standard is continuous EEG to capture an event or during an ongoing or fluctuating state of altered consciousness to see if there is associated electrographic evidence of seizure activity.

There is a growing encyclopedia of EEG findings and associated acronyms, which can broadly be divided into five categories:

1. 
   

   Normal variants not associated with epilepsy but of unclear significance (e.g., small sharp spikes, wicket spikes)

   
   
2. 
   

   Findings associated with focal or global cerebral pathology but not necessarily with epilepsy (e.g., focal or generalized slowing, triphasic waves, frontal intermittent rhythmic delta activity [FIRDA])

   
   
3. 
   

   Findings indicating cortical irritability and risk of seizures (e.g., spike-wave discharges, periodic lateralized epileptiform discharges [PLEDs], generalized periodic discharges [GPEDs]).

   
   
4. 
   

   Seizures

   
   
5. 
   

   Artifact, which can be due to:

   
   
   
   
   • 
     

     Patient factors (e.g., blinking, movements)

     
     
   • 
     

     Technical factors (e.g., electrocardiographic [ECG] artifact, interference from electrical hospital equipment, issues with EEG leads)

EEG is most useful in the following scenarios:

• 
  

  Determining if a particular type of event is a seizure (e.g., versus a nonepileptic spell) by capturing EEG data during a typical event (usually requires continuous EEG monitoring)

  
  
• 
  

  Determining if a patient with an altered level of consciousness is having nonconvulsive seizures contributing to the altered state of consciousness (also usually requires continuous EEG monitoring to capture intermittent seizure activity) (see “Nonconvulsive Status Epilepticus” below)

  
  
• 
  

  Determining whether a patient with seizures has a particular electrographic signature of a certain epilepsy syndrome so as to guide prognosis and management

  
  
• 
  

  Determining the risk of future seizures in patients after first seizure (see “Evaluation and Management of Patients After a First Seizure” below)

  
  
• 
  

  Precise localization of a seizure focus if patients are being considered for epilepsy surgery (see “Refractory (Drug-Resistant) Epilepsy” below)

EEG may be obtained as a routine EEG (generally 20 minutes of recording) or as continuous EEG (hours to days of monitoring). The routine EEG is often obtained to assess for interictal epileptiform discharges (spikes and sharp waves) in patients who have had one or more episodes concerning for seizure. The sensitivity of a single 20-minute routine EEG recording to detect interictal epileptiform discharges is only around 50%, although sensitivity can be increased by performing the EEG in the sleep-deprived state, performing EEG within 24 hours of a seizure event, or repeating EEG on multiple occasions. Notably, a small proportion of the population may have EEG abnormalities of no clinical significance, and a number of medications can also cause EEG abnormalities. Therefore, the absence of epileptiform discharges does not “exclude” epilepsy and their presence does not “confirm” epilepsy. Seizures and epilepsy are clinical diagnoses and EEG findings must be interpreted in light of the clinical history. If a patient has had paroxysmal events and there is a strong clinical suspicion that the patient has had seizures, a normal EEG should not necessarily dissuade the clinician from treating these as seizures. Similarly, if the clinician has a strong suspicion that the events are not seizures (e.g., syncope, migraine), an epileptiform finding on routine EEG should not necessarily dissuade the clinician from that impression.