42
CHAPTER
Epilepsy and Sleep
Erick N. Viorritto and Sujay M. Kansagra
Even before the advent of EEG in the early twentieth century, clinical observation had borne out a connection between epilepsy and sleep. It was observed that nocturnal seizures were more likely to occur at specific times during sleep, and that daytime seizures may disrupt sleep the following night (1). Once EEG allowed for a more rigorous study of both epilepsy and sleep physiology, the complexity of that relationship became evident.
Epilepsy and sleep are interrelated in a variety of ways. Seizures disrupt the normal architecture and physiology of sleep, and the manifestations of many epilepsy syndromes display specific patterns related to the patient’s sleep–wake cycle, as summarized in Table 42.1. Furthermore, sleep disorders are often comorbid with epilepsy, which may exacerbate the patient’s epilepsy and complicate treatment. Finally, therapies for epilepsy (both pharmacologic and nonpharmacologic) may impact an individual’s normal sleep physiology, and therapies for sleep disorders may affect seizure control in the patient with epilepsy.
THE EFFECT OF SLEEP ON EPILEPSY
Sleep consists of two easily distinguishable states: rapid eye movement (REM) sleep and nonrapid eye movement (NREM) sleep. NREM sleep is further divided into three stages—N1, N2, and N3 sleep. There is increasing neuronal synchronicity on EEG as one goes from light N1 sleep to N3 sleep (termed “slow-wave sleep”). Clinically, there are two peaks during the night in which seizures are most common. These occur between 9 p.m. and 11 p.m., and then later in the night between 3 a.m. and 5 a.m (1). When examined in relation to the EEG, most sleep-related seizures occur during NREM sleep, where synchronous neuronal activity may serve to facilitate the generation and propagation of seizures. This interpretation, however, does not explain why sleep-related seizures are more common during stage N2 sleep (61%–68%) compared to the more synchronous N3 stage (9%–14%) (2). It is possible that the thalamocortical hypersynchrony that is evident during stage N2 sleep, as manifest by the generation of sleep spindles and K-complexes on the EEG, may promote interictal spikes and epileptic activity (3). REM sleep appears to be relatively protective against seizures, with most studies showing that less than 1% of sleep-related seizures occur during this stage. This may be due to the relatively desynchronized activity that is seen during REM sleep (4). While seizures in REM sleep are rare, epileptiform discharges during REM sleep may be the most accurate in helping to localize an epileptogenic focus for epilepsy surgery.
Sleep deprivation is a fairly standard activating procedure used to increase the yield for capturing interictal abnormalities and seizures on EEG. Sleep deprivation activates interictal discharges in one-third of patients with epilepsy, and the yield in patients with sleep-related epilepsies may be much higher (1). Sleep deprivation also likely has direct effects on cortical excitability, serving to lower the seizure threshold (2).
Focal Epilepsies
It has been observed since the nineteenth century that there is a substantial cohort of patients with epilepsy who only have seizures during sleep. According to one study, such “pure sleep epilepsies” occurs in about 6% of patients with epilepsy. A larger group, making up approximately 10% of epilepsy patients, has seizures predominantly during sleep with only occasional daytime seizures (2). About 80% of the pure sleep epilepsies are focal, with idiopathic focal epilepsy more likely to have seizures restricted to sleep than to lesional focal epilepsy (2).
Idiopathic focal epilepsies, such as benign epilepsy of childhood with centrotemporal spikes (BECTS), display focal-onset seizures occurring predominantly during sleep. In 70% to 80% of cases, the seizures are exclusive to sleep (2). A neurophysiologic feature of this syndrome is the marked activation of interictal discharges during drowsiness and light NREM sleep (2). Studies have shown that in up to one-third of patients with BECTS, the interictal discharges themselves are seen only during sleep (1).
Early-onset childhood epilepsy with occipital spikes (Panayiotopoulos syndrome) also presents with focal-onset seizures predominantly occurring during sleep, with prominent autonomic fluctuations, behavioral disturbances, and significant emesis. Two-thirds of the seizures begin during sleep, and they can last over an hour in a quarter of cases (5). As in BECTS, a hallmark of this epilepsy is the emergence of interictal discharges during NREM sleep (2).
Most patients with autosomal dominant nocturnal frontal lobe epilepsy syndrome (ADNFLE) likewise present with seizures that occur exclusively during sleep (2). This syndrome is characterized by nocturnal clusters of brief, stereotyped motor seizures during sleep. There is typically a maintenance of consciousness during the seizures and minimal or absent postictal confusion. The foci for these seizures may be deep, and thus they may lack interictal and ictal EEG findings, leading to their misdiagnosis as a parasomnia or sleep disorder (6).
Symptomatic focal epilepsies from the frontal lobe and temporal lobe can be sleep-related and produce seizures that occur predominantly out of sleep. The most commonly encountered of these is temporal lobe epilepsy, owing to the relative frequency of this epilepsy type; frontal lobe foci however have the greatest likelihood of displaying a sleep-related pattern. Approximately 61% of frontal lobe seizures arise from sleep, compared to 11% of temporal lobe seizures (7). If over 90% of a patient’s seizures arise out of sleep, the patient is considered to have nocturnal frontal lobe epilepsy (NFLE) or, less commonly, nocturnal temporal lobe epilepsy (NTLE), depending on the location of the foci (2). Evidence suggests that NTLE may carry a better prognosis following surgical resection than temporal lobe epilepsies in which seizures are not restricted to sleep (7).
Interictal discharges in symptomatic focal epilepsies are facilitated by NREM sleep. Discharges can also be seen during REM sleep with less frequency. These interictals tend to show a more restricted field, and studies suggest that these REM-related interictal discharges may be a more reliable indicator of the epileptic focus than interictal discharges seen during wakefulness or NREM sleep (2,4).
Generalized Epilepsies
Most idiopathic generalized epilepsies present with seizures primarily during wakefulness, although a pattern of seizures related to sleep still exists. This category includes the “awakening epilepsies,” in which seizures predominantly occur within the first two hours after the patient wakes up (8). This is seen in the cases of juvenile myoclonic epilepsy (JME) and epilepsy with generalized tonic–clonic seizures on awakening. Sleep deprivation tends to elicit seizures in patients with these syndromes. Interictal generalized discharges are seen most frequently during NREM sleep and least frequently during REM sleep, and may be seen in association with K-complexes on the EEG (2). Fewer than 10% of patients with idiopathic generalized epilepsy present with a pure sleep epilepsy (9). Symptomatic generalized epilepsies may present with seizures during both wakefulness and sleep, although again a clear sleep-related pattern can be seen. In Lennox-Gastaut syndrome, multiple seizure types can be seen during wakefulness, including generalized tonic–clonic, tonic, atonic, myoclonic, and atypical absence seizures. Tonic seizures during NREM sleep can be seen in over 90% of patients, and are accompanied by paroxysmal fast activity on EEG. Interictal spike-wave discharges also become more prominent during NREM sleep where runs of generalized polyspike discharges and rhythmic bursts of fast activity may be seen (1). In contrast, only 2% to 5% of the infantile spasms associated with West syndrome occur during sleep, instead occurring in clusters upon awakening. The interictal hypsarrythmia EEG pattern in West syndrome, however, is most prominent during early NREM sleep, and in some patients may only be evident during NREM sleep (1,2).
Epileptic Encephalopathies
Clear sleep-related EEG changes are seen in both continuous spike-wave during slow-wave sleep (CSWS) and Landau-Kleffner syndrome (LKS). In both disorders, sleep is accompanied by near-continuous, epileptiform activity on the EEG. In the waking state, both disorders are marked by deterioration in neurocognitive function. CSWS is associated with nocturnal focal motor seizures and occasionally absence seizures during wakefulness, while LKS may or may not be accompanied by clinical seizures (2).
EFFECT OF EPILEPSY AND SEIZURES ON SLEEP
There is a reciprocal relationship between sleep and epilepsy: while it is clear that sleep plays a role in seizure frequency, so too do seizures and epilepsy affect an individual’s sleep. These changes are likely due to multiple factors. There may be an effect from the underlying pathological mechanism of the epilepsy on sleep, independent of whether or not seizures are occurring. There are also changes in nocturnal sleep that can be seen as a result of either nocturnal seizures or diurnal seizures in the period preceding sleep. Finally, many antiepileptic drugs (AEDs) and therapies also affect sleep.
The Effect of Epilepsy and Seizures on Sleep Continuity and Architecture
Epilepsy appears to have an impact on sleep continuity and sleep architecture independent of the presence of seizures on a given night. This appears to predominantly involve an increase in wake time after sleep onset (WASO) and an increased number of awakenings on nights in which seizures did not occur (1). This has been hypothesized to be due to aberrant GABA release. It also appears that this effect is more prominent in temporal lobe epilepsy compared to primary generalized epilepsy (10). In temporal lobe epilepsy, nocturnal seizures are also associated with independent changes in sleep architecture, with increased WASO, decreased sleep efficiency, decreased REM sleep, decreased N2 sleep, decreased N3 sleep, increased N1 sleep, and increased next-day drowsiness. On the night following a diurnal temporal lobe seizure, decreased REM sleep is also seen, without the changes in WASO, sleep efficiency, and amounts N1, N2, and N3 sleep seen after nocturnal seizures (11).
