Epilepsy: A Comprehensive Textbook
2nd Edition

Chapter 187
Chronobiology
Margaret N. Shouse
Mark S. Quigg
Introduction
The term chronobiology refers to the clocking of biologic events. There are many biologic clocks that control behavioral and physiological processes.54 Epilepsy is affected by biologic clocks, a phenomenon that has been documented for more than a century.22,24,33,34,35,57,82,83
This chapter focuses on three kinds of clocks or periodicities as they pertain to epilepsy: (a) circadian rhythms, which recur at about 24-hour intervals, (b) ultradian rhythms, which recur at <24-hour intervals, and (c) infradian rhythms, which recur at >24-hour intervals (e.g., monthly or seasonal).
More than one periodicity can exist. For example, some patients have 24-hour seizure peaks, whereas most have multiple peaks within 24 hours, sometimes at about 90-minute intervals defined by the basic rest–activity cycle (BRAC). Many patients with circadian and ultradian seizures also exhibit strong infradian patterns, as evidenced by multiple cyclic variations in hormonal secretions (Chapters 194,195,196,197,198). The way in which these periodic rhythms affect epilepsy depends on several interrelated variables that define epileptic syndromes,15 including (a) seizure type, (b) etiology, (c) age-dependent course, and (d) prognosis.
Circadian (24-Hour) Rhythms/Variations and Seizure Events
A circadian rhythm in the strict sense refers to any periodicity (hormones, temperature, mobility, etc.) that fluctuates over the course of approximately a 24-hour day and does not necessarily refer to the sleep–wake cycle. The sleep–wake cycle is a circadian rhythm in many mammals, such as rats and humans, but it recurs several times a day in cats. The “time of day” difference does not alter the fact that interictal discharges (IIDs) and clinically evident seizures exhibit similar distributions during different sleep and waking states in all these species. Accordingly, most studies of epilepsy have focused on variables that exhibit variation in the sleep–wake cycle without examining the role of circadian “time of day” rhythms per se.
There are several reasons for the paucity of clear-cut data on circadian rhythms. Perhaps the most relevant is that, in contrast to the sleep–wake cycle, circadian periodicity is not directly measurable because it is not simply a matter of clock time. It requires robust circadian markers or reference points, the use of various mathematical constructs, and data collection over prolonged sampling intervals. True, endogenous circadian rhythms recur on a “free-running” basis regardless of exogenous cues, such as the light–dark cycle, or of any other endogenous factors, such as linkage to sleep versus waking states.
Findings on biologic events that are clearly under circadian rhythm versus sleep–waking state control are difficult to interpret. Masking effects, or perturbation of circadian markers, can complicate interpretation of observations on circadian seizure modulation. For example, cortisol secretion has a circadian periodicity, but seizures themselves elevate cortisol (e.g., Quigg83). Cortisol secretion is also modulated by stress, which is itself a seizure precipitant (e.g., Frucht et al.27). Melatonin is intimately involved in light-related circadian organization of the sleep–wake cycle and, by definition, is susceptible to masking effects by exogenous clues. Exogenous melatonin administration can improve seizure control in experimental animals90 and in clinical trials.23 However, the ameliorative effects of melatonin cannot be dissociated from improvements in sleep, which is disordered in epilepsy (see Chapter 188),44 or from seizure-related factors (e.g., Bazil5).
These issues have been extensively reviewed by Quigg.82,83,84 In this section, we review the evidence on circadian variation in seizure activity as function of the sleep–wake cycle vis-à-vis time of day (clock time) and show examples of endogenous circadian rhythm modulation, when available.
Clinical Findings
Three epilepsy categories are classified according to the timing of generalized myoclonic or tonic–clonic seizures in the sleep–wake cycle. Timing categories are referred to as awakening (diurnal), sleep (nocturnal), and diffuse epilepsies, the latter including epileptic syndromes in which seizures occur randomly in sleep or waking states as well as in the day or night.44,46 These three epilepsy groups are also classified as a function of the different epileptic syndromes with which they are affiliated.15,45,46,75,77
Epilepsies Characterized by Seizures on Awakening (Diurnal): Primary Generalized Epilepsies
Epilepsies characterized by seizures on awakening are a special class of diurnal seizures, which are considered more entrained to the sleep–wake cycle than to endogenous circadian rhythms.31 They are most often primary generalized epilepsies in which the etiology is assumed to be genetic. These include patients with typical absence, juvenile myoclonic epilepsy (JME), and generalized tonic–clonic seizures.2,45,46,47,77
Patients with absence and JME often have generalized tonic–clonic seizures.29,30,48 Greater than 90% of these patients have generalized tonic–clonic convulsions (GTCs) only on arousal from sleep,45,46 most frequently during prolonged drowsy periods occurring between 10 minutes and 2 hours after morning awakening.8,45,46 Myoclonic and absence seizures are also common at this time. Table 1 shows the prevalence of generalized myoclonic seizures after awakening from nocturnal sleep.119
Table 1 Distribution of 51 primary generalized myoclonic seizures in relation to the time of predilection for occurrence of seizures and to the efficacy of treatment in 33 patients
Time of predilection for occurrence Total (n = 33 patients) Satisfactory control (n = 18) Unsatisfactory control (n = 15)
Morning awakening 26 17 9
Evening period of relaxation 6 5 1
Sleep onset 3 2 1
Sleep 6 2 4
Nocturnal wakening 10 9 1
Adapted from Touchon J. Effect of awakening on epileptic activity in primary generalized myoclonic epilepsy. In: Sterman MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. New York: Academic Press; 1982:239–248; with permission.
FIGURE 1. Electroencephalogram (EEG) recorded after morning awakening. Eye closure elicits a series of polyspike-and-wave (PSW) discharges associated with electromyographic (EMG) activity. Subsequent eye opening inhibited such EEG manifestations, which resumed again on closing the eyes. e.c., eyes closed; e.o., eyes opened; EOG, electrooculogram; L. Delt., left deltoid; R. Delt., right deltoid. (From Gigli GL, Calia E, Luciani L, et al. Eye closure sensitivity without photosensitivity in juvenile myoclonic epilepsy: polysomnographic study of electroencephalographic epileptiform discharge rates. Epilepsia. 1992;32:677–683; with permission.)
Many patients also exhibit more frequent and prolonged interictal epileptiform discharges after awakening.37,51,52,77,78 Polyspike-and-wave complexes of JME often occur after awakening from nocturnal sleep, as depicted in FIGURE 1.31 FIGURE 2 shows that the duration of 3/sec spike–wave complexes
P.1962

associated with absence seizure disorders can be longer after awakening than at other times in the sleep–wake cycle.52
Ictal events are more entrained to awakening than are IIDs. Frequent polyspike-and-wave and spike-and-wave complexes also occur during non–rapid-eye-movement (NREM) sleep (e.g., Fig. 2), even when myoclonic and absence seizures occur only on awakening.
Epilepsies characterized by seizures on awakening typically exhibit an age-dependent clinical course.15,70,73,97,98,99 Onset is usually between 4 and 15 years of age. Spontaneous remission, reduction of seizures, or complete medication control is common between puberty and 20 to 25 years of age, although some patients can be more drug refractory.70,78,98
These relatively benign epileptic syndromes are associated with normal developmental and neurologic functions. The presumed pathophysiology32 is a disturbance in the electrophysiologic and neurochemical response of neocortical cell populations to synchronous synaptic inputs resulting from inhibition of arousal cells located in the hypothalamus and in the ascending reticular activating system (ARAS).4,74,103,104
FIGURE 2. Circadian distribution of 3/sec spike-and-wave activity per 15 minutes. Discharge duration is longest at awakening and so is more likely to represent a clinical seizure. Discharge rate is also high at the beginning and end of sleep. Time of day along the abscissa is in terms of the 24-hour clock. The bar along the abscissa indicates the sleep or waking state: Unfilled, awake; filled, non–rapid-eye-movement (NREM) sleep; slanted lines, rapid-eye-movement (REM) sleep. (From Kellaway P, Frost JD Jr, Crawley JW. Time modulation of spike-and-wave activity in generalized epilepsy. Ann Neurol. 1980;8:491–500; with permission.)
FIGURE 3. Spike density distribution at 5-minute intervals in benign epilepsy with centrotemporal spikes (BECT) showing a decremental pattern. An abrupt increase in spike activity is associated with sleep onset and is followed by an overall decline during subsequent non–rapid-eye-movement (NREM) sleep and particularly on awakening. Periodic troughs occur throughout sleep in relation to REM onset. SWS, slow-wave sleep. (From Kellaway P, Frost JD Jr. Biorhythmic modulation of epileptic events. In: Pedley TA, Meldrum BS, eds. Recent Advances in Epilepsy, Vol. 1. London: Churchill Livingstone; 1983:139–154; with permission.)
Epilepsies Characterized by Seizures During Sleep (Nocturnal): Localization-Related Epilepsies
Circadian ictal and interictal discharge patterns occur during the “subjective night” and/or during sleep in epilepsies arising from a focal region of dysfunction (e.g., Frost et al.26). These may be relatively benign or not. For example, some frontal lobe syndromes, notably nocturnal paroxysmal dystonia62,63,94 and autosomal-dominant nocturnal frontal lobe epilepsy (see Chapter 251), display bizarre partial motor seizure manifestations during NREM sleep. The seizures do not secondarily generalize in NREM sleep and are readily controlled by antiepileptic medication. Other “benign” localization epilepsies have an age-dependent course with a specific time frame for onset and spontaneous remission. The prognosis for spontaneous remission of seizure manifestations, is usually good.2,6,7,15,70,78,98,99 These patients also display frequent IID and partial seizures with or without secondary generalization during NREM sleep. Examples are benign partial epilepsy with centrotemporal spikes (BECT, also called benign rolandic epilepsy),58 benign epilepsies with occipital spikes,6,7 Landau-Kleffner syndrome (acquired epileptic aphasia), and patients manifesting electrical status epilepticus during slow sleep (ESES).113 FIGURE 3 shows an example of BECT in which the interictal spike discharge peaks at sleep onset.51
Simple partial or complex partial seizures, particularly those accompanied by secondary generalized tonic–clonic seizures, have long been thought to occur more frequently in sleep than in waking (e.g., Janz45,46). Greater than 50% of patients with temporal or frontal lobe foci reported secondary generalized tonic–clonic seizures only in sleep, whereas complex partial seizures without secondary generalization were reported to occur more often during waking.45,46 Because localization-related epilepsies are more likely to be symptomatic than are primary generalized epilepsies,15,45,46,78 it is not surprising that the prognosis is not always benign. For example, complex
P.1963

partial seizures of temporal or frontal lobe origin do not have an age-dependent course. Onset frequently occurs before the age of 20 years, spontaneous remission is rare, and seizures are often drug refractory.15,45,46,78
Recent findings using continuous electroencephalogram (EEG)-video monitoring for presurgical evaluation have revealed some interesting new developments in drug-refractory localization-related epilepsies.17,39,48,81 Different timing patterns emerge when patients with drug-refractory complex partial seizures are evaluated as a function of the presence or absence of secondary generalized seizures and as a function of foci in frontal, temporal parietal, or occipital lobes (Fig. 4).40 Partial seizures occur more frequently in waking than in sleep, regardless of the location of the seizure focus (Fig. 4A). Complex partial seizures with secondary generalization occur significantly more often in NREM than in waking at all seizure foci except in frontal lobe (Fig. 4B). It was concluded that different mechanisms govern the timing of seizure initiation versus propagation.
Quigg proposed that circadian diurnal rhythms govern the initiation of symptomatic partial seizures with limbic foci, whereas nocturnal sleep-related mechanisms trigger onset of focal seizures in symptomatic focal seizure disorders with nonlimbic foci. Several findings have been assembled to address this hypothesis:
  • The timing of symptomatic limbic seizures (mesial temporal lobe sclerosis) can be differentiated from that of nonlimbic (extralimbic) seizures with respect to linkage with circadian versus predominantly vigilance-related seizures. Table 2 summarizes this difference on the basis of recent studies dating from 1998 to 2004.84
  • Figure 583 illustrates the 24-hour distribution of seizures in two patients with left hippocampal sclerosis and temporal lobe onset of seizures (Fig. 5A, B) and in a single patient with dual pathology (Fig. 5C, D). The three patients with symptomatic mesial temporal lobe epilepsy (MLTE) (Fig. 5A–C) all showed complex partial seizures during the subjective day, whereas the timing of symptomatic, parietal lobe seizures occurred nocturnally and during sleep (Fig. 5D).
  • Figure 6 contrasts the absence of a circadian rhythm in extratemporal focal seizures (Fig. 6A) with the presence of a circadian seizure distribution in an animal model and in patients with MTLE.83 Findings are plotted as a function of clock time in humans (Fig. 6A, B) and with reference to the circadian temperature cycle in the 12-hour/12-hour light/dark cycle in animals (Fig. 6C). A comparison of findings in Figures 6B and 6C supports the conclusion that MLTE in both species is regulated by a circadian rhythm.82,84,87 This circadian pattern was also maintained in rats during constant darkness when plotted in relation to the free-running circadian temperature cycle (Fig. 6D).87
The interpretation of these findings is complicated by the fact that rats are nocturnally active and humans are diurnally active. Thus, seizure activity seems dissociated from circadian
P.1964

rest–activity rhythm regulation. In addition, circadian entrainment disappeared when the data from free-running rhythms were plotted with reference to circadian rest–activity patterns (not shown in Fig. 6). Both discrepancies could suggest an underlying circadian modulation of limbic seizure occurrence that does not involve rest–activity or sleep–wakefulness patterns.
FIGURE 4. A: Percentage of partial seizures arising from various brain regions during waking (gray bars) or sleep (black bars). Partial seizures with or without secondary generalization of seizures (n = 579 of 613 partial seizures in 133 patients) were included. B: Percentage of partial seizures with secondary generalization (n = 396 of 613 partial seizures in 133 patients) during waking (gray bars) or sleep (black bars). In both histograms, data are presented for each location assessed. Overall, partial seizures were more likely to occur during wakefulness regardless of seizure onset site (A), whereas partial seizures with secondary generalization were significantly more likely to occur in sleep at all onset sites except fron-tal lobe (B). Sleep refers to non–rapid-eye-movement sleep because no seizures were registered during rapid-eye-movement sleep. Data were obtained during in-patient monitoring of patients who were candidates for epilepsy surgery due to refractory complex partial seizures. FLE, frontal lobe; MLTE; mesial temporal lobe; NTE, neocortical temporal lobe; OLE, occipital or parietal lobes; TLE, temporal lobe. *p <0.05; **p <0.001; ***p <0.0001. (From Herman ST, Walczah TS, Bazil CW. Distribution of partial seizures during the sleep–wake cycle: differences by seizure onset. Neurology. 2001;56:1453–1459; with permission.)
Table 2 Distribution of focal seizures by location of seizure
Study Location Light (%) Dark (%)
Quigg et al., 199887 MTLE 60 40
  XTLE 54 46
Herman et al., 200140 TLE 74 26
  FLE 43 57
Crespel et al., 199817 MTLE 84 16
  FLE 39 61
Pavlova et al., 200481 TLE 81 19
  XTLE 59 41
FLE, frontal lobe epilepsy; MTLE, mesial temporal lobe epilepsy; TLE, temporal lobe epilepsy; XTLE, nontemporal lobe neocortical epilepsy. Adapted from Quigg M. Chronobiology, sleep, and seizures (2004 Profiles in Seizure Management, Case 2). Princeton Media Associates; 2004(3);1–11. Available: http://www.princetonhcm.com/public/index.php?program=2004-79-6&rid=226; with permission.
Epilepsies in Which Seizures Occur Randomly (Diurnal and Nocturnal): Symptomatic Generalized Epilepsies
In these epileptic syndromes, ictal and interictal discharges occur in all sleep and waking states.45,46,67 This random seizure distribution is often associated with diffuse, severe cerebral dysfunction.43,52,79,97 Three well-known examples are West,14,123 Lennox-Gastaut,30,79 and progressive myoclonus36 syndromes.
FIGURE 5. Daily distribution of seizures revealed by detailed, long-term diaries maintained by three subjects with medically refractory epilepsy and their spouses. Localization of seizures was later confirmed during epilepsy surgery evaluation. Curves provide the best-fit estimate of time of peak occurrence (95th confidence interval). A, B: Patients with left hippocampal sclerosis and left temporal lobe seizures. C, D: Patient with a pattern of seizures with different symptomatology and a dual pathology: Right temporal lobe seizures associated with ipsilateral hippocampal sclerosis occurred diurnally (C), and parietal lobe seizures associated with a cortical malformation occurred nocturnally and out of phase with temporal seizures (D). (Composite from Quigg83 as follows: Panels A, B from Quigg M. Seizures and circadian rhythms. In: Bazil CW, Malow BA, Sammaritano MR, eds. Sleep and Epilepsy: The Clinical Spectrum. Amsterdam: Elsevier; 2002:127–142; panels C, D adapted from Quigg M, Straume M. Dual epileptic foci in a single patient express distinct temporal patterns dependent on limbic versus nonlimbic brain location. Ann Neurol. 2000;48:117–120; with permission.)
FIGURE 6. Distribution of spontaneous epileptic seizures in humans and in the self-sustained electrical status epilepticus rat model of limbic epilepsy. The times of seizure occurrence in humans were determined using continuous video-electroencephalogram (EEG) recordings from either scalp or intracranial electrodes. In rats, continuous EEG and hippocampal depth electrodes provided times of seizures. Where appropriate, data were fitted using cosinor-nonlinear least squares analysis. A: A biphasic distribution of seizures in patients with extratemporal lobe partial epilepsy (XTLE). When the observed distribution is compared with the expected uniform rate of seizure occurrence using chi-square analysis, the distribution is statistically random. In contrast, limbic seizures in humans with mesial temporal lobe epilepsy (MTLE) (B) and in rats (C, D) occur in similar patterns. Seizures of limbic origin are accurately modeled by a cosine function with the calculated time of peak occurrence reported with the 95% confidence limit. Both species have diurnally predominant seizures despite the fact that rats are nocturnal and humans are diurnal in activity. The limbic seizures of rats occur in similar distributions when entrained to a 12-hour/12-hour light/dark cycle (C) as compared to free-running circadian rhythms in constant darkness or a 12-hour/12-hour dark/dark cycle (D). Note that the times of seizure occurrence are normalized to a 24-hour circadian clock provided by each animal’s circadian rhythm of temperature because, once in constant darkness, the period of circadian rhythms may run slightly shorter or longer than exactly 24 hours. The plot in panel D confirms that the occurrence of limbic seizures is modulated in an endogenous, circadian fashion. (Composite from Quigg M. Seizures and circadian rhythms. In: Bazil CW, Malow BA, Sammaritano MR, eds. Sleep and Epilepsy: The Clinical Spectrum. Amsterdam: Elsevier; 2002:127–142; as adapted from Quigg M, Clayburn H, Straume M, et al. Epilepsia. 2000;41:502–509; and Quigg M, Straume M. Dual epileptic foci in a single patient express distinct temporal patterns dependent on limbic versus nonlimbic brain location. Ann Neurol. 2000;48:117–120; with permission.)
Onset can be early in life, but average age at onset in this category of seizure disorders is otherwise evenly distributed across the age spectrum.45,70 Seizures tend to be medically refractory, and the prognosis is poor.6,15,30,77,78,123 In some cases, morbidity and even death is expected (e.g., Lafora, essential hereditary myoclonus, or Unverricht-Lundborg syndrome).36,78 In other cases, conversion to other serious seizure disorders is common (e.g., West syndrome, which is also called infantile spasms, to Lennox-Gastaut syndrome79). Spontaneous remission does not occur. Cases with extreme neurologic complications display substantial disruption of sleep states and of the sleep–wake cycle (see Chapter 188). Extreme neurodegenerative disorders also provide the only example in which endogenous circadian rhythm disorders are clearly manifested.39,44
Basic Mechanisms
Strong rhythms of circadian ictal and to some extent interictal discharge patterns arise in two groups of epilepsy.26,33,34,35,36,37,41,42,43,45,46,47 In primary generalized epilepsies, ictal and interictal discharges often appear on awakening, although interictal discharge is common in NREM sleep as well. Localization-related epilepsies frequently exhibit generalized ictal and interictal discharges events during sleep, often at sleep onset or at the end of sleep. The same ictal and interictal patterns seen in humans have also been reported in experimental animal models of sleep versus awakening epilepsy (Fig. 7A vs. Fig. 7B).99
What is known about basic chronobiology mechanisms that might explain these differences? To address this question, it is necessary to consider two main hypothalamic regulatory systems: (a) the suprachiasmatic nuclei of the anterior hypothalamus, which is the “master” circadian clocking mechanism, and
P.1965

(b) the anterior/preoptic, lateral, and posterior hypothalamic cell populations, which generate the sleep–wake cycle.
The suprachiasmatic nucleus (SCN) of the hypothalamus regulates nearly all circadian rhythms, including the subsystems linked to the sleep–wake cycle.13,54,82,83 Sleep-dependent processes include NREM-related changes such as plasma growth hormone secretion and urinary calcium excretion. Lesions of either the SCN or its afferent retinal pathway eliminate most physiologic and behavioral circadian rhythms, including the sleep–waking circadian cycle.21,73,102 Cells of the SCN continue to exhibit circadian discharge patterns in isolation122 and also restore normal circadian patterns when transplanted into animals with genetically abbreviated periodicities.88 These findings confirm that the SCN is the primary circadian clock. Various afferent and efferent anatomic interconnections have been identified to account for modulation of component circadian rhythms, such as the adjacent cell populations generating the sleep–wake cycle. Circadian oscillation in the SCN is genetically determined.1,88 Several genes encode clocking proteins to create an autoregulatory feedback loop using transcription and translation.61 γ-Aminobutyric acid (GABA) is the main transmitter.69 GABA release is well known to have antiepileptic effects.
Considering the undisputed, pervasive influence of the SCN on circadian rhythms, it is difficult to understand why there is so little evidence linking this structure or the circadian light–dark cycle to ictal or interictal seizure events and vice versa. A number of reasons have been suggested (e.g., Quigg83). Examples are masking effects, which can occur as a result of extraneous environmental or internal factors, and the only fairly recent availability of prolonged polysomnographic and video recordings needed to capture sufficient seizure events for “time of day” analysis.
Independent oscillators also exist, even in structures normally entrained to discharge patterns of cells in the SCN. Destruction of the SCN abolishes circadian “time of day” rhythms, but the usual result is a replacement with other rhythms. FIGURE 8 shows the persistence of sleep–wake behaviors and unchanged overall percentages of sleep and wake time after SCN lesions. The circadian rhythm is replaced by an ultradian rhythm. It is curious that no one has attempted to differentiate “time of day” effects from sleep–waking-state effects on seizures before and after SCN lesions.
Another complication is that cells of the SCN discharge in the light phase of the light–dark cycle and are silent during the dark phase, regardless of whether mammals exhibit sleep nocturnally or diurnally. Taken alone, this fact is not disturbing because many circadian rhythms are totally out of phase with cellular discharge patterns of the SCN (e.g., melatonin release). The troublesome aspect is that rats, which are nocturnally active, still exhibit IIDs during diurnal sleep (e.g., Ascapone and Penry2 and Kostopoulos53). In contrast, humans frequently exhibit seizures on awakening (primary generalized epilepsy [PGE]) or in sleep (most localization-related epilepsies), even though humans are typically awake diurnally and asleep nocturnally. Thus, with few exceptions, circadian effects on seizures and of seizures on circadian rhythms are so far primarily entrained to sleep and waking states.
The only way in which to eradicate or substantially diminish sleep or waking states is to destroy the generators of these states. Well-documented clinical (e.g., Von Economo121) and experimental findings, including transection, lesion, unit recording, microdialysis, c-fos expression, and/or regional blood flow studies, localize control of the sleep–wake cycle to the hypothalamus.4,13,69,76,120 Histaminic cells in the tuberomammillary nucleus of posterior hypothalamus, orexin (hypocretin)-containing cells of the lateral hypothalamus, and cholinergic cells in the magnocellular region of the anterior hypothalamus generate awakening and maintain arousal, whereas GABAergic and galanin-containing cells of the anterior/preoptic hypothalamic areas (ventral lateral preoptic [VLPO]) generate sleep onset and maintenance.60,69,88,95,110,111,112 These hypothalamic regions are reciprocally connected60,95,96 and dominate all the EEG, behavioral, autonomic, and, conceivably, hormonal correlates of the sleep–wake cycle.
In the intact animal, these forebrain regions are also reciprocally interconnected and functionally interact with brainstem cholinergic, noradrenergic, and/or serotonergic cells originating in the pontine tegmentum. These brainstem cells also discharge during onset and maintenance of wakefulness and, with the exception of occasional bursts of cholinergic cell discharge (see Chapter 188), are silent during NREM
P.1966

sleep.4,10,49,69,76,120,121 For this reason, the coverage of basic mechanisms focuses on these particular regions.
FIGURE 7. Timing of spontaneous seizures during the sleep–wake cycle in (A) 9 amygdala-kindled kittens and (B) 12 cats with systemic penicillin epilepsy (300,000–400,000 IU/kg). Temporal lobe kindling is a model of localization-related epilepsies, whereas systemic penicillin epilepsy is a model of primary generalized epilepsy. Slow-wave sleep (SWS) in cats is equivalent to non–rapid-eye-movement (NREM) sleep in humans. AD, afterdischarge. (From Shouse MN, King A, Langer J, et al. Basic mechanisms underlying seizure-prone and seizure-resistant sleep and awakening states in feline kindled and penicillin epilepsy. In Wada JA, ed. Kindling 4. New York: Plenum Press; 1990:313–327; with permission.)
FIGURE 8. Percentage of polygraphically recorded wake time in each 30-minute period over 48 consecutive hours in two adult male rats recorded in constant dim light. A: An intact control rat, showing a typical 24-hour circadian rhythm of sleep–wake time. B: A rat with complete ablation of the suprachiasmatic nuclei (SCN). The 24-hour rhythm of sleep–wake time was completely lost. This rat shows prominent ultradian rhythms in the 3- to 4-hour range. (Modified from Mistlberger HE, Bergmann BM, Rechtschaffen A. Relationships among wake episode lengths, contiguous sleep episode lengths, and electroencephalographic delta waves in rats with suprachiasmatic nuclei lesions. Sleep. 1987;10:12–24; with permission.)
FIGURE 9. A: Spontaneous unit activity in a posterior lateral hypothalamic (PLH) neuron during slow-wave sleep (SWS), which here refers to non–rapid-eye-movement (NREM) sleep, and at awakening (arrow) before and during a subconvulsive dose of penicillin (200,000 IU/kg) in a cat. B: Evoked orthodromic response in a PLH neuron induced by stimulation of the external capsule (100 pulses, 0.8 mA), before and after penicillin (200,000 IU/kg) in a cat. Stimulus onset was at time 0. Note the increased amplitude of evoked excitation and the reduced duration of postexcitatory inhibition after penicillin when compared to the pre-penicillin record. (From Shouse MN, King A, Langer J, et al. Basic mechanisms underlying seizure-prone and seizure-resistant sleep and awakening states in feline kindled and penicillin epilepsy. EEG, electroencephalogram; EOG, electrooculogram. In: Wada JA, ed. Kindling 4. New York: Plenum Press; 1990:313–327; with permission.)
Epilepsies Characterized by Seizures on Awakening (Diurnal): Primary Generalized Epilepsies
The morphology of interictal discharges, including 3/sec spike-and-wave and polyspike-and-wave complexes,15,29,77 is compatible with stimulus-evoked or recruited cortical EEG patterns resembling drowsiness. This may explain why interictal discharges arise during drowsy wakefulness.32,37,77,78 Conversion to an ictal discharge pattern associated with myoclonic or tonic–clonic seizures could be precipitated by corticopetal influences originating in arousal cells of the hypothalamus, the brainstem reticular formation, or both, as follows.
Arousal cells in the magnocellular basal forebrain and in the posterior hypothalamus are also called waking-active neurons. These cells have direct projections to the entire neocortical mantle.49,92,93 Sudden bursts of “excitatory” input from these arousal cells might exacerbate the diffuse, relatively mild cortical hyperexcitability thought to underlie this group of seizure disorders (e.g., Gloor32). One hypothesized mechanism is a direct (monosynaptic) effect mediated by normal or abnormal secretion of several transmitters such as acetylcholine (Ach) and/or histamine (e.g., Jones49).
Figure 9 shows spontaneous (Fig. 9A) and evoked (Fig. 9B) unit discharge of a waking-active neuron in the posterior hypothalamus during slow-wave sleep and awakening before and 1 hour after a subconvulsive dose of penicillin in a cat.99 Note that spontaneous cellular discharge increases on awakening and is exacerbated at this time by penicillin (Fig. 9A). The
P.1967

peristimulus histogram shows increased amplitude of evoked excitation and reduced duration of postexcitatory inhibition in the penicillin model of PGE on awakening (Fig. 9B; also see Fig. 7B).
Alternatively, indirect (polysynaptic) effects could be responsible. For example, waking-active neurons of the posterior hypothalamus and orexin-containing cells of the lateral hypothalamus also project to and synergize actions of the arousal cells in the ARAS,74 including norepinephrine (NE), serotonin (5-HT), and/or Ach. These brainstem arousal cells in turn have widespread terminals in thalamus and cortex.49,68,92,93,104 An impressive body of data supports the hypothesis that inhibition of arousal cells promotes synchronized thalamocortical discharges associated with NREM sleep and also the sleep-related interictal discharges in the penicillin and genetic models of awakening epilepsy.3,28,53,103,104
Sustained activation of the pontine arousal cells is thought to have tonic antiepileptic effects (e.g., Corcoran16; see Chapter 188), but the abrupt increase in discharge of these cells on awakening may be epileptogenic. For example, NE is thought to promote the ability to focus neuronal attention in wakefulness. The outcome is a coordinated sensorimotor act.103 A surge of NE release on awakening might simply focus neuronal attention on and activate the epileptic neocortical cell populations in this group of seizure disorders. A similar effect could accompany the release of 5-HT and/or Ach.
Epilepsies Characterized by Seizures During Sleep (Nocturnal): Localization-Related Epilepsies
Seizures and interictal discharges arising during sleep could be provoked by monosynaptic and polysynaptic projections from sleep-active neurons of the anterior/preoptic hypothalamus to cortical or subcortical seizure foci. Hypothalamic sleep-active neurons suppress waking-active neurons located in the ARAS,4,44,69 notably, the Ach, NE, and/or 5-HT cells of the pontine tegmentum. These cells in turn have widespread monosynaptic and polysynaptic projections to the entire neocortical mantle as well as to archicortical and subcortical limbic sites49 and could thus precipitate seizure discharge during NREM sleep. The role of the thalamocortical system and its regulation of epileptogenic phasic events such as sleep spindles also cannot be overlooked (see Chapter 188).
Epilepsies in Which Seizures Occur Randomly (Day or Night): Symptomatic Generalized Epilepsies
Circadian patterns of ictal and interictal discharges are not prominent in these epileptic syndromes (e.g., Horita41,42). The extensive cerebral pathology associated with these syndromes could encompass the SCN, the hypothalamic generators of the sleep–wake cycle, and other brainstem and forebrain regions that ultimately express circadian rhythms.
P.1968

Ultradian (<24-Hour) Rhythms and Seizure Events
Historically, the ultradian rhythm most often linked to epilepsy is called the basic rest–activity cycle (BRAC).12,105,107 The BRAC pervades the circadian sleep–wake cycle. During sleep, the synchronized EEG discharges of NREM sleep represent “rest” periods and alternate with the asynchronous EEG discharges of REM sleep, which represent “activity” periods. During waking, rest periods (drowsiness) alternate with activity periods (alert waking). In humans, this periodicity occurs at roughly 90-minute intervals. There is great individual variability, and 100- to 120-minute intervals have often been detected (e.g., Kellaway et al.52).
A second, more recently described ultradian rhythm refers to periodic microarousals, defined as a cyclic alternating pattern (CAP) of EEG desynchronization occurring on a background of prolonged, homogeneous EEG synchronization (noncyclic alternating pattern [NCAP]). CAPs have two components or phases.114,115 Phase A corresponds to the initial or peak level of EEG desynchronization with sleep EEG transients, whereas phase B follows phase A and reflects a return to the stable synchronized-EEG NCAP state. Three different subtypes of CAP A have been identified and related to various seizure disorders. CAPs have been studied during NREM sleep, in which brief EEG microarousals are most readily detected (e.g., Parrino et al.80).
Clinical Findings
Ultradian BRAC seizure patterns are most evident in localization-related epilepsies of temporal or frontal lobe origin. In contrast to ictal events, IIDs with BRAC periodicities can occur in all seizure disorders. Ultradian CAP rhythms have been studied mostly in idiopathic epilepsies,31,38,115,116 although studies have been performed in three types of localization-related epilepsies.80,119 The observations are depicted in epileptic syndromes characterized by awakening, sleep, and random seizure patterns as follows.
FIGURE 10. Timing of ictal and interictal discharges at 90-minute intervals in a patient with temporal lobe epilepsy. A 24-hour spike count per 4 minutes was recorded from the right temporal electrode of a patient with six seizures recorded during this period. Seizures (arrows) occurred at intervals or multiples of intervals of 1.5 hours. No distinct rapid-eye-movement sleep was recorded in the patient. (From Stevens JR, Lonsbury B and Goel SL. Seizure occurrence and interspike interval. Arch Neurol. 1980;26:409–411; with permission.)
Epilepsies Characterized by Seizures on Awakening (Diurnal): Primary Generalized Epilepsy
Seizures occur primarily on arousal from nocturnal sleep, but there are less prominent, secondary peaks on arousal from daytime naps and sometimes at sleep onset, as seen in Table 1.33,57,119 The appearance of more than one peak constitutes an ultradian rhythm, but it does not reflect a BRAC pattern per se. Interictal discharges, on the other hand, can show strong ultradian rhythms corresponding to a BRAC pattern.51,65,66,105,106,107,108
Ultradian CAP patterns of IIDs have been studied most thoroughly in JME.21 However, Terzano et al.,114 who developed the CAP concept and methodology, studied a small and heterogeneous population of patients with various idiopathic (presumed genetic) seizure disorders115,116 and found that the rate of spike-and-wave or polyspike-and-wave complexes is highest in the arousal phase (CAP A) and lowest in the sleep recovery phase (CAP B) when compared to NCAP. Seizure discharge is most often affiliated with k-complexes and bursts of slow waves, which are thought to represent phasic arousal events during CAP A.
Epilepsies Characterized by Seizures During Sleep (Nocturnal): Localization-Related Epilepsies
This group of epilepsies, especially those of temporal or frontal lobe origin, is known to be predisposed to ultradian seizures related to the BRAC.11,106,107,108 Many authors have reported peaks in the timing of complex partial seizures and of interictal discharges, notably spikes and sharp waves, at about 90-minute intervals throughout the sleep–wake cycle,11,106,107,108 as depicted in FIGURE 10.107 Similar findings have been detected in animal models. An example is the prevalence of spontaneous convulsions emanating from kindled temporal lobe foci during the REM transition in FIGURE 7A.99
A CAP factor similar to that seen in awakening epilepsies also applies to certain sleep epilepsies, notably those of temporal or frontotemporal origin.80 CAP modulation of IIDs in benign epilepsy with rolandic spikes (BERS) has not been detected.118 The basis for this negative outcome is speculative. Amygdala-kindled kittens shows EEG microarousals
P.1969

associated with subclinical and possibly clinical seizures, which occur throughout the sleep–wake cycle at much shorter time intervals than the ∼20-minute BRAC in cats and might reflect a CAP cycle.100
Epilepsies in Which Seizures Occur Randomly (Waking and Sleep): Symptomatic Generalized Epilepsies
These patients show weak ultradian seizure patterns. Some, such as those with West syndrome, can have demonstrable ultradian BRAC periodicities when a discernible REM cycle exists.43,51,52,79
Basic Mechanisms
The ultradian BRAC has been localized to the brainstem. This conclusion is based on its persistence in the brainstem but not the forebrain following transection at the midcollicular level (cerveau isolé preparation).10,76,91,120 Transection findings are confirmed by other experimental methods such as selective lesion, extracellular unit, microdialysis, c-fos, and immunohistochemical studies (e.g., Baghdoyan and Lydic4 and Beaussart7). The same brainstem neurotransmitters released during the awakening phase of the circadian sleep–wake cycle are also thought to serve as the brainstem generators of the BRAC. Discharge of NE and 5-HT cells in the medial and medial-lateral pontine tegmentum declines in the transition into REM, whereas Ach discharge in the dorsolateral pedunculopontine tegmentum increases and generates the phasic “arousal” events of the transition. These phasic arousal events are proposed to precipitate seizure discharge and its propagation at this time (Fig. 7A vs. Fig. 7B) (see Chapter 188).37,99 The tonic discharge of Ach cells during REM is proposed to have antiepileptic effects in sustained periods of alert waking and REM (see Chapter 188). The combination of reduced cellular discharge and transmitter release of NE and 5-HT cells and increased Ach activity is mediated by local GABAergic interneurons (e.g., Baghdoyan and Lydic4 and Lin et al.60).
Reduced NE and 5-HT release plus the phasic discharge bursts and chemical release of Ach cells at these times may promote ultradian ictal or IID manifestations or both. Because each cell type has widespread projections,49,68 epileptogenic effects could be mediated by direct or indirect projections to epileptic cells anywhere in the brain. On the other hand, the periodic EEG microarousals associated with the arousal phase of CAP (i.e., CAP A) might reflect transient activation of the same brainstem arousal mechanisms regulating the BRAC or the forebrain hypothalamic mechanisms generating matutinal awakening.
Infradian (>24-Hour) Rhythms and Seizure Events
The existence of infradian or long-running rhythms, such as seasonal patterns, has been known for well more than a century.22,24,34,35,56 Here, we provide two better-known examples relevant to seizure disorders.
Primary generalized epilepsies, which are characterized by seizures on awakening, have an age-dependent course paralleled by changes in the amplitude of EEG transients, such as the spike component of k-complexes. The k-complexes are associated with aborted arousals in sleep and with IIDs during sleep in this group of seizure disorders.37,77,78 The clinical course also corresponds to the age-dependent secretory patterns of hormones and/or hormone-releasing factors such as melatonin, sex, and steroid-releasing factors. All of these chemicals have been implicated in arousal as well as in the ictal and interictal discharges in these epileptic syndromes.83,91,125 Common genetic variables potentially related to an arousal dysfunction have been proposed to underlie these correlated age-dependent changes.19,20,71
FIGURE 11. A decremental epileptic discharge pattern in which the duration and number of 3/sec spike-and-wave discharges per unit time peaks at the onset of non–rapid-eye-movement (NREM) sleep (here called slow-wave sleep [SWS]) and declines in successive NREM sleep cycles over the night. The suppression of spike-and-wave activity that occurs during REM sleep is comparable over successive REM cycles. Time of day on the abscissa is according to the 24-hour clock. The bar along the abscissa indicates the sleep or waking state: Unfilled, awake; filled, SWS; slanted lines, REM sleep. (From Kellaway P, Frost JD Jr, Crawley JW. Time modulation of spike-and-wave activity in generalized epilepsy. Ann Neurol. 1980;8:491–500; with permission.)
FIGURE 12. An incremental–decremental epileptic discharge pattern in the duration and number of 3/sec spike-and-wave discharges per 15 minutes. The augmenting effect of non–rapid-eye-movement (NREM) sleep (here called slow-wave sleep [SWS]) on 3/sec spike-and-wave discharges at first increases and then declines toward the end of sleep. An equivalent reduction in spike-and-wave activity occurs in each REM period. Time of day along the abscissa is according to the 24-hour clock. The bar along the abscissa indicates the sleep or waking state: Unfilled, awake; filled, SWS; slanted lines, REM sleep. (From Kellaway P, Frost JD Jr, Crawley JW. Time modulation of spike-and-wave activity in generalized epilepsy. Ann Neurol. 1980;8:491–500; with permission.)
Hormonal regulation of various seizure disorders is detailed elsewhere (see Chapters 194,195,196,197,198), including catamenial epilepsy. Hormonal changes associated with catamenial epilepsy also affect monoamines, as also covered elsewhere. Once menses develop, a monthly or near-monthly periodicity in seizures is exhibited before or during menstruation in 30% to
P.1970

70% of women with epilepsy (e.g., Foldvary-Schafer et al.25). Once menopause occurs, a monthly cycle disappears.
Interactions Between Periodicities
Most ictal and interictal discharge patterns show admixtures of peaks and troughs related to circadian and ultradian variables as well as to different sleep stages (e.g., Kellaway and Frost51 and Kellaway et al.52), sometimes regardless of epileptic syndrome. For example, circadian peaks are seen at sleep onset in centrotemporal spikes (Fig. 3) and in 3/sec spike-and-wave discharges (Fig. 11).52 During subsequent ultradian NREM/REM cycles, a stepwise decline in seizure discharges may occur during successive NREM cycles (a so-called decremental pattern) and alternate with a near-complete suppression during REM epochs. FIGURE 12 illustrates a different pattern, in which spike-and-wave discharges at first increase and then decrease during successive NREM cycles (a so-called incremental–decremental pattern), again alternating with a suppression of spike-and-wave activity during REM sleep.52 With this diversity, a logical question is: How do different biologic rhythms interact with state-related variables to generate different temporal seizure patterns?
This is a complex issue that is partially addressed by the statistical model of sleep-related epileptic discharges illustrated in FIGURE 13.51 Here, the joint probabilities associated with hypothesized circadian (Fig. 13A) and ultradian (Fig. 13B) rhythms are calculated to predict the distribution of 3/sec spike-and-wave complexes during 8-hour sleep periods as a function of different sleep onset times (Fig. 13C–F). The model assumes a constant sinusoidal 24-hour circadian rhythm (Fig. 13A) and a sleep-related, sinusoidal 1.5-hour ultradian rhythm (Fig. 13B). The apexes of these curves correspond to peak circadian and ultradian effects. The 1.5-hour ultradian rhythm represents the NREM–REM cycle, which is shown only for one sleep onset time in FIGURE 13B but is assumed to be present throughout each hypothetical 8-hour sleep period depicted in FIGURE 13C–F.
FIGURE 13. A statistical model showing the circadian (A) and ultradian (B) periodicities and how they predict the distribution of spike-and-wave probabilities over 8-hour sleep cycles as a function of sleep onset time. C: Sleep onset at 30° from A results in an incremental–decremental pattern in which spike-and-wave incidence at first increases after sleep onset and then subsides at the end of the sleep cycle. D: A 120° sleep onset lag results in a decremental pattern in which peak spike-and-wave discharge occurs at sleep onset and declines thereafter. E: A 210° sleep-onset lag with a bimodal pattern in which spike-and-wave incidence peaks at the beginning and end of the sleep cycle. F: A 300° sleep-onset lag produces an incremental pattern in which spike-and-wave incidence increases with the duration of sleep. (From Kellaway P, Frost JD Jr. Biorhythmic modulation of epileptic events. In: Pedley TA, Meldrum BS, eds. Recent Advances in Epilepsy, vol. 1. London: Churchill Livingstone; 1983:139–154, with permission.)
Spike-and-wave discharge probabilities range from 0.0 to 1.0 on the ordinate. The highest joint probabilities are generated when the apexes of circadian and ultradian curves coincide. Conversely, the lowest joint probabilities are predicted when the troughs of the two curves coincide. Intermediate probabilities reflect interaction between the heights of the
P.1971

different curves at a given time. In FIGURE 13C–F, spike-and-wave probabilities are plotted as a function of different sleep onset times, expressed as phase angle θ from circadian time 0 on the abscissa.
Figure 13C shows that when sleep onset occurs at a 30° phase angle, the first two ultradian cycles occur on the ascending slope of the circadian curve, the middle cycle occurs at the apex of both the circadian and ultradian BRAC curves, and the last two ultradian cycles fall on the descending slope of the circadian curve. This interaction generates an incremental–decremental pattern like that depicted in FIGURE 12, in which spike-and-wave incidence at first increases in successive NREM sleep bouts, peaks in the middle, and declines toward the end of sleep.
Figure 13D shows that when sleep onset occurs at a 120° phase angle, the first ultradian cycle occurs at the apex of the circadian and ultradian curves, and the subsequent ultradian cycles fall on a progressively descending slope of the circadian curve. This generates a decremental probability pattern similar to those seen in Figures 3 and 11, in which epileptic discharge rates decline over successive NREM cycles.
Figure 13E shows that when sleep onset occurs at a 210° angle, the first two ultradian cycles occur on the descending slope of the circadian curve, the middle at the base of the circadian curve, and the last two on the ascending slope of the circadian curve. This generates a bimodal probability pattern similar to that seen in FIGURE 2, in which spike-and-wave discharge peaks at the beginning and end of sleep.
Figure 13F shows that when sleep onset occurs at a 300° angle, successive ultradian cycles fall on a progressively ascending circadian curve. This generates an incremental spike-and-wave probability distribution, which is the opposite of the pattern predicted in FIGURE 13D.
Notwithstanding limitations of this model (e.g., waking is ignored), it can predict epileptiform discharge patterns that are not explained by state-dependent variables such as NREM versus REM sleep stage alone. For example, sleep stage does not seem to explain distinctive patterns observed over the course of the sleep cycle, particularly during successive NREM sleep epochs. Effects of hypothesized phase advance and delay manipulations, as exemplified by irregularities in sleep habits in FIGURE 13, suggest how time-dependent factors associated with chronobiology, provide this modulation.51,52
Subsequent experimental findings did not support these hypotheses about sleep displacement effects.50 For example, acute sleep delay (6 hours) sufficient to produce a 90° phase shift did not significantly alter the distribution pattern of 3/sec spike-and-wave discharges over successive NREM epochs during a 12-hour sleep recording period when compared to the two, preceding 12 baseline polysomnograms. The authors concluded that the presence of sleep per se is the critical regulatory factor and that the timing of spike-and-wave discharge is secondarily modulated by ultradian REM-NREM cycles.
Summary and Conclusions
Clinical Findings
  • There are no studies of interictal discharges and few studies of seizures in which endogenous circadian “time of day” patterns have been clearly differentiated from the sleep–wake cycle. This gap needs to be filled.
  • The timing of IIDs in sleep and waking states sometimes corresponds to the timing of seizures, but there is no
    P.1972

    empirical basis for suggesting that IIDs routinely predict state-dependent seizure occurrence in humans or animals (e.g., Herman et al.,40 versus Marlow et al.64). This dissociation prompted Stevens et al.108 to state, “all that spikes isn’t fits,” meaning that an IID is a seizure that did not occur. The reason that IIDs do not generate seizures, especially in seizure-prone sleep and waking states, is still enigmatic, although some aspects are addressed in Chapter 188.
  • Circadian (24-hour) state-related ictal and to some extent interictal discharge rhythms are strong in epilepsies with awakening seizures, which are primary generalized epilepsies of hereditary or idiopathic origin (e.g., absence and benign juvenile myoclonic epilepsy with or without tonic–clonic convulsions). Circadian state-related seizure patterns are also strong in sleep epilepsies, which often originate from temporal or frontal seizure foci, but also include BERS, benign epilepsy with occipital spikes, Landau-Kleffner syndrome, and ESES. Circadian sleep–wake seizure patterns are rarely seen in epilepsies with extensive organic complications (e.g., West syndrome).
  • Ultradian (<24-hour) state-dependent ictal patterns are most frequently detected in epilepsies with awakening or sleep seizures. Ultradian interictal discharge patterns also frequently occur in various seizure disorders regardless of the timing of ictal events. Ultradian interictal discharge patterns reflect both 90- to 120-minute BRAC cycles and the briefer CAP cycles associated with transient EEG arousals, especially k-complexes and bursts of slow waves.
  • Infradian rhythms, as seen in catamenial epilepsy, are not contingent on seizure type, timing, or etiology. Other long-running rhythms can be selective. For example, primary generalized epilepsies of hereditary origin and good prognosis show an age-dependent course, here meaning a relatively specific age at onset and spontaneous remission.
  • Multiple periodicities seem to interact with other factors, such as sleep stage, to generate diverse but predictable temporal seizure discharge patterns.
Basic Mechanisms
  • The circadian sleep–wake cycle has been localized to the hypothalamus. The posterior histaminic, anterior cholinergic, and possibly the lateral hypocretin-secreting areas induce awakening, whereas the GABAergic and galaninergic cells of the anterior hypothalamic/preoptic basal forebrain area induce sleep. These arousal- or sleep-active cells could provoke seizures via direct or indirect innervation of the epileptic cells involved in the genesis of awakening or sleep epilepsies.
  • The ultradian BRAC has been localized to the brainstem, likely the pontine tegmentum. Activation of BRAC-related ictal or interictal discharges could result from increased electrochemical activity via direct or indirect projections from these reticular formation neurons to epileptic cells anywhere in the brain. Similarly, activation of ultradian CAP-related interictal discharges could reflect transient increases in electrochemical activity of cholinergic, noradrenergic, and/or serotonergic cells in the brainstem ARAS or of the forebrain hypothalamic generators of arousal.
  • Infradian rhythms seem governed by age-dependent genetic and hormonal influences.
Acknowledgments
This work was supported by the Department of Veterans Affairs.
References
1. Albrecht U, Eichele G. The mammalian circadian clock. Curr Opin Genet Dev. 2003;13:271–277.
2. Ascapone J, Penry JK. Some clinical aspects of benign juvenile myoclonic epilepsy. Epilepsia. 1984;25:108–114.
3. Avanzini G, Panzoca F, de Curtis P. The role of thalamus in vigilance and epileptogenic mechanisms. Clin Neurophysiol. 2000;111(Suppl 2):S19–S26.
4. Baghdoyan HA, Lydic R. Neurotransmitter and neuromodulators regulating sleep. In: Bazil CW, Malow BA, Sammaritano MR, eds. Sleep and Epilepsy: The Clinical Spectrum. Amsterdam: Elsevier; 2002:17–44.
5. Bazil CW, Shon D, Crispin D, et al. patients with intractable epilepsy have low melatonin, which increases following seizures. Neurology. 2000;55:1746–1748.
6. Beaumanoir A. Infantile epilepsy with occipital spikes and good prognosis. Eur Neurol. 1983;22:43–52.
7. Beaussart M. Benign epilepsy of children with rolandic (centro-temporal) paroxysmal foci. A clinical study of 221 cases. Epilepsia. 1972;13:795–811.
8. Beyer L, Javonovic UJ. Elektroencephalographische und klinische Korrelate bei Aufwachepileptikern mit besonderer Berucksichtigung der therapeutischen Probleme. Nervenzart. 1966;37:333–336.
9. Bergamasco B, Bergamini L, Mutani R. Spontaneous sleep abnormalities in the case of dyssynergia cerebellaris myoclonia. Epilepsia. 1967;8:271–281.
10. Bremer F. Cerveau isolé et physiologie du sommeil. C R Soc Biol (Paris). 1935;118:1235–1242.
11. Cadhillac J. Complex partial seizures and REM sleep. In: Sterman MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. New York: Academic Press; 1982:315–324.
12. Carskadon MA, Dement WC. Normal sleep and its variations In: Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 4th ed. Philadelphia: Elsevier/WB Saunders; 2004:13–23.
13. Chokroverty S. Physiological changes in sleep. In: Bazil CW, Malow BA, Sammaritano MR, eds. Sleep and Epilepsy: The Clinical Spectrum. Amsterdam: Elsevier; 2002:46–63.
14. Chugani H, Shields D, Newmon DA, et al. Infantile spasms: PET identities, focal cortical dysgenesis and cryptogenic cases for surgical treatment. Ann Neurol. 1990;27:406–413.
15. Commission Report. Proposal for revised classification of epilepsies and epileptic syndromes: Commission on Classification and Terminology of the International League Against Epilepsy. Epilepsia. 1989;30:389–399.
16. Corcoran ME. Characteristics of accelerated kindling after depletion of noradrenaline in adult rats. Neuropharmacology. 1988;27:1081–1084.
17. Crespel A, Baldy-Moulinier M, Coubes P. The relationship between sleep and epilepsy in frontal and temporal lobe epilepsy: practical and physiopathologic considerations. Epilepsia. 1998;30:150–157.
18. Dalla Bernadina B, Beghini G. Rolandic spikes in children with and without epilepsy (20 subjects polygraphically studied during sleep). Epilepsia. 1976;17:161–167.
19. Degen R, Degen HE. Some genetic aspects of rolandic epilepsy: waking and sleep EEGs in siblings. Epilepsia. 1990;31:795–801.
20. Delgado Escueta AV, Greenberg DA, Treiman L, et al. Mapping the gene for juvenile myoclonic epilepsy. Epilepsia. 1989;30(Suppl 4):S8–S18.
21. Eastman C, Mistleberger R, Rechtschaffen A. Suprachiasmatic nuclei lesions eliminate temperature and sleep rhythms in the rat. Physiol Behav. 1984;32:357–368.
22. Echeverria ME. De l’épilepsie nocturne. Ann Med Psych. 1879;6:177.
23. Fauteck JD, Schmidt H, Lerchl A. et al. Melatonin in epilepsy: first results of replacement therapy and first clinical results. Biol Signals Recept. 1999;8:105–110.
24. Fere C. Les Epilepsies et les Epileptiques. Paris: Alcan; 1890.
25. Foldvary-Schafer N, Falcon T. Catamenial epilepsy: pathophysiology, diagnosis, and management. Neurology. 2003;61(Suppl. 2):S1–S15.
26. Frost JD Jr, Hrachovy RA, Glaze DG, et al. Sleep modulation of interictal spike configuration in untreated children with partial seizures. Epilepsia. 1991;32:341–346.
27. Frucht MM, Quigg M, Schwaner C, et al. Distribution of seizure precipitants among epilepsy syndromes. Epilepsia. 2000;41:1534–1539.
28. Fuentealba P, Steriade M. The reticular nucleus revisited: intrinsic and network properties of a thalamic pacemaker. Prog Neurobiol. 2005;75:125–141.
29. Gastaut H. Clinical and electroencephalographic classification of epileptic seizures. Epilepsia. 1970;11:102–103.
30. Gastaut H, Roger J, Soulayrol R, et al. Childhood epileptic encephalopathy with diffuse slow spike-waves (otherwise known as “petit mal variant”) or Lennox syndrome. Epilepsia. 1966;7:139–179.
31. Gigli GL, Calia E, Luciani L, et al. Eye closure sensitivity without photosensitivity in juvenile myoclonic epilepsy: polysomnographic study of electroencephalographic epileptiform discharge rates. Epilepsia. 1991;32:677–683.
32. Gloor P. Generalized epilepsy with spike-wave discharge: A reinterpretation of its electroencephalographic and clinical manifestations. Epilepsia. 1979;20:571–588.
33. Gowers WR. Epilepsy and Other Chronic Convulsive Diseases: Their Causes, Symptoms and Treatment. Philadelphia: Blakiston’s; 1901.
P.1973

34. Griffiths GM, Tyler Fox J. Rhythm in epilepsy. Lancet. 1938;2:409–416.
35. Grigg-Damberger M, Damberger SJ. Historical aspects of sleep and epilepsy. In: Bazil CW, Malow BA, Sammaritano MR, eds. Sleep and Epilepsy: The Clinical Spectrum. Amsterdam: Elsevier; 2002:3–16.
36. Guiliani G, Marchesi GF, Quattrini A, et al. L’epilessia mioclona familiare progressiva (sindrome di Unverricht-Lundborg). Indagine genetica, clinica, poligrafica e istopatologica. Minerva Med. 1977;68:2923–2954.
37. Halâsz P. Generalized epilepsy with spike-wave pattern (GESW) and intermediate states of sleep. In: Sterman MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. New York: Academic Press; 1982:219–238.
38. Halâsz P, Terzano MG, Parrino L. Spike-wave discharge and the microstructure of sleep–wake continuum in idiopathic generalized epilepsy. Neurophysiol Clin. 2002;32:38–53.
39. Heikkla E, Hatonen T, Telakivi T, et al. Circadian studies in neuronal ceroid-lipofuscinosis (NCL). Am J Med Genet. 1995;5:229–234.
40. Herman ST, Walczah TS, Bazil CW. Distribution of partial seizures during the sleep–wake cycle: differences by seizure onset. Neurology. 2001;56:1453–1459.
41. Horita H. Epileptic seizures and sleep–wake rhythm. Psychiatry Clin Neurosci. 2001;55:171–172.
42. Horita H, Uchida E, Maekawa K. Circadian rhythm of regular spike-wave discharges in childhood absence epilepsy. Brain Dev. 1991;13:200–202.
43. Hrachovy RA, Frost JD, Kellaway P. Sleep characteristics in infantile spasms. Neurology. 1981;31:688–694.
44. Jan M. Melatonin for the treatment of handicapped children with severe sleep disorders. Pediatr Neurol. 2000;23:229–232.
45. Janz D. The grand mal epilepsies and the sleeping–waking cycle. Epilepsia. 1962;3:69–109.
46. Janz D. Epilepsy and the sleeping–waking cycle. In: Vincken PJ, Bruyn GW, eds. Handbook of Clinical Neurology, Vol. 15: The Epilepsies. Amsterdam: North-Holland; 1974:457–490.
47. Janz D. Epilepsy with grand mal on awakening and the sleep–waking cycle. Clin Neurophysiol. 2000;111(Suppl.2):S103–S110.
48. Jobst BC, Williamson PD, Neuschwader TB, et al. Secondarily generalized seizures in mesial temporal lobe epilepsy: clinical characteristics, lateralizing signs, and association with sleep–wake cycle. Epilepsia. 2001;42:1279–1287.
49. Jones BE. Basic mechanisms of sleep–wake states. In: Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 4th ed. Philadelphia: Elsevier/WB Saunders; 2004:136–153.
50. Kellaway P, Frost JD Jr., Burtin WD Jr. Circadian modulation of 3 Hz spike–wave activity. Epilepsia. 1996(Suppl. 5);37:S15.
51. Kellaway P, Frost JD Jr. Biorhythmic modulation of epileptic events. In: Pedley TA, Meldrum BS, eds. Recent Advances in Epilepsy, Vol. 1. London: Churchill Livingstone; 1983:139–154.
52. Kellaway P, Frost JD Jr, Crawley JW. Time modulation of spike-and-wave activity in generalized epilepsy. Ann Neurol. 1980;8:491–500.
53. Kostopoulos GK. Spike-and-wave discharges of absence seizure as a transformation of sleep spindles: the continuing development of a hypothesis. Clin Neurophysiol. 2000;111(Suppl 2):S27–38.
54. Kryger MH, Roth T, Carskadon M. Circadian rhythms in humans. In: Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 2nd ed. Philadelphia: WB Saunders; 1994:301–308.
55. Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 2nd ed., Section V. Philadelphia: WB Saunders; 1994:227–321.
56. Laidlaw J. Catamenial epilepsy. Lancet. 1956;2:1235–1237.
57. Langdon-Down M, Brain WR. Time of day in relation to convulsions in epilepsy. Lancet. 1929;2:1029–1032.
58. Lerman P. Benign partial epilepsy with centro-temporal spikes. In: Roger J, Bureau M, Dravet CH, et al., eds. Epileptic Syndromes in Infancy, Childhood and Adolescence, 2nd ed. London: John Libbey; 1992:189–200.
59. Lin JS, Sakai K, Jouvet M. Hypothalamo-preoptic histaminergic projections in sleep–wake control in the cat. Eur J Neurosci. 1994;6:618–625.
60. Lin JS, Sakai K, Vanni-Mercier G, et al. A critical role of the posterior hypothalamus in the mechanisms of wakefulness determined by microinjection of muscimol in freely moving cats. Brain Res. 1989;479:225–240.
61. Lowery PL, Shimomura K, Antoch MP, et al. Positional synthetic cloning and functional characterization of the mammalian circadian mutation TAU. Science. 2000;288:483–492.
62. Lugaresi E, Cirgnotta F. Hypnogenic paroxysmal dystonia: epileptic seizures or a new syndrome? Sleep. 1981;4:129–138.
63. Lugaresi E, Cirgnotta F, Montagna P. Nocturnal paroxysmal dystonia. J Neurol Neurosurg Psychiatry. 1986;49:375–380.
64. Malow BA, Lin X, Kushwara R, et al. Interictal spiking increases with sleep depth in temporal lobe epilepsy. Epilepsia. 1998;39:1309–1316.
65. Martins da Silva A, Aarts JHP, Binnie CD, et al. The circadian distribution of interictal epileptiform EEG activity. Electroencephalogr Clin Neurophysiol. 1984;58:1–13.
66. Martins da Silva A, Binnie CD. Ultradian variation of epileptiform EEG activity. In: Martins da Silva A, Binnie CD, Meinardi H, eds. Biorhythms and Epilepsy. New York: Raven Press; 1985:69–77.
67. Marzec ML, Malow BA. Approaches to staging sleep in polysomnographic studies with epileptic activity. Sleep Med. 2003;4:409–417.
68. McBride RI, Sutin J. Projection of the locus ceruleus and adjacent pontine tegmentum of the cat. J Comp Neurol. 1976;165:265–284.
69. McGinty DJ, Szymusiak R. Sleep promoting mechanisms in mammals. In: Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 4th ed. Philadelphia: Elsevier/WB Saunders; 2004:169–184.
70. Mendez M, Radtke RA. Interaction between sleep and epilepsy. J Clin Neurophysiol. 2001;18:106–127.
71. Metrakos K, Metrakos JD. Genetics of convulsive disorders II: Genetic and electroencephalographic studies in centrencephalic epilepsy. Neurology. 1961;11:474–483.
72. Mistlberger HE, Bergmann BM, Rechtschaffen A. Relationships among wake episode lengths, contiguous sleep episode lengths, and electroencephalographic delta waves in rats with suprachiasmatic nuclei lesions. Sleep. 1987;10:12–24.
73. Moore RY, Eicher VB. Loss of a circadian adrenal corticosterone rhythm following suprachiasmatic lesions in the rat. Brain Res. 1972;42:201–206.
74. Moruzzi G, Magoun HW. Brain stem reticular formation and activation of the EEG. Electroencephalogr Clin Neurophysiol. 1949;1:455–473.
75. Nagao H, Morimoto T, Takahashi M, et al. The circadian rhythm of typical absence seizures—the frequency and duration of paroxysmal discharges. Neuropediatrics. 1990;21;79–82.
76. Nauta WJH. Hypothalamic regulation of sleep in rats: an experimental study. J Neurophysiol. 1946;9:285–316.
77. Niedermeyer E. Abnormal EEG patterns (epileptic and paroxysmal). In: Niedermeyer E, Lopes da Silva F, eds. Electroencephalography. Baltimore: Urban & Schwarzenberg; 1982:155–178.
78. Niedermeyer E. Epileptic seizure disorders. In: Niedermeyer E, Lopes da Silva F, eds. Electroencephalography. Baltimore: Urban & Schwarzenberg; 1982:339–428.
79. Ohtahara S, Yamatogi Y, Ohtsuka Y. Prognosis of Lennox syndrome. Long-term clinical and electroencephalographic follow-up study, with special reference to West syndrome. Folia Psychiatr Neurol Jpn. 1976;30:275–287.
80. Parrino L, Smerieri A, Spaggiari MC, et al. Cyclic alternating pattern (CAP) and epilepsy during sleep: how a physiological rhythm modulates a pathological event. Clin Neurophysiol. 2000;111(Suppl 2):S39–S46.
81. Pavlova MK, Shea SA, Bromfield EB. Day/night patterns of focal seizures. Epilepsy Behav. 2004;5:44–49.
82. Quigg M. Circadian rhythms: interaction with seizures and epilepsy. Epilepsy Res. 2000;42:43–55.
83. Quigg M. Seizures and circadian rhythms. In: Bazil CW, Malow BA, Sammaritano MR, eds. Sleep and Epilepsy: The Clinical Spectrum. Amsterdam: Elsevier; 2002:127–142.
84. Quigg M. Chronobiology, sleep, and seizures (2004 Profiles in Seizure Management, Case 2). Princeton Media Associates; 2004(3):1–11. Available: http://www.princetonhcm.com/public/index.php?program=2004-79-6&rid=226.
85. Quigg M, Clayburn H, Straume M, et al. Effects of circadian regulation and rest-activity state on spontaneous seizures in a rat model of limbic epilepsy. Epilepsia. 2000;41:502–509.
86. Quigg M, Straum M, Menaker M, et al. Dual epileptic foci in a single patient express distinct temporal patterns dependent on limbic versus nonlimbic brain location. Ann Neurol. 2000;48:117–120.
87. Quigg M, Straume M, Menaker M, et al. Temporal distribution of partial seizures: comparison of an animal model with human partial epilepsy. Ann Neurol. 1998;43:748–755.
88. Ralph MR, Foster RG, Davis FC, et al. Transplanted suprachiasmatic nucleus determines circadian period. Science. 1990;247:975–978.
89. Ralph MR, Menaker M. Effects of diazepam on circadian phase advances and delays. Brain Res. 1986;372:405–408.
90. Rudeen PK, Philo RC, Symmes SK. Antiepileptic effects of melatonin in the pinealized Mongolian gerbil. Epilepsia. 1980;21:149–154.
91. Sandyk R. Melatonin and petit-mal epilepsy. Int J Neurosci. 1992;65:383–390.
92. Saper CB. Organization of cerebral afferent system in the rat, II: Magnocellular basal nucleus. J Comp Neurol. 1984;222:313–342.
93. Saper CB. Organization of cerebral afferent system in the rat, II: Hypothalomocortical projection. J Comp Neurol. 1985;237:21–46.
94. Sforza E, Montagna P, Rinaldi P, et al. Paroxysmal period motor attacks during sleep: clinical and polygraphic features. Electroencephalogr Clin Neurophysiol. 1993;86:161–166.
95. Sherin JE, Elmquist, Torrealba F, et al. Innervation of histaminic tuberomammillary neurons by GABAergic and galaninergic neurons in the ventrolateral preoptic nucleus of the rat. J Neurosci. 1998;18:4705–4721.
96. Sherin JE, Shiromani PJ, McCarley RW, et al. Activation of preoptic neurons during sleep. Science. 1996;271:216–219.
97. Shouse MN. Epileptic seizure manifestations during sleep. In: Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 2nd ed. Philadelphia: WB Saunders; 1994:801–814.
98. Shouse MN. Seizures. In: Lee-Chiong T, ed. Encyclopedia of Sleep Medicine. Hoboken, NJ: Wiley; 2006:793–803.
99. Shouse MN, King A, Langer J, et al. Basic mechanisms underlying seizure-prone and seizure-resistant sleep and awakening states in feline kindled and penicillin epilepsy. In: Wada JA, ed. Kindling 4. New York: Plenum Press; 1990:313–327.
100. Shouse MN, Langer J, King A, et al. Paroxysmal microarousals in amygdala-kindled kittens: could they be subclinical seizures? Epilepsia. 1995;36:290–300.
P.1974

101. Siegel JM. Brain stem mechanisms generating REM sleep. In: Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 4th ed. Philadelphia: WB Saunders; 2004:120–135.
102. Stephan F, Zucker I. Circadian rhythms in drinking behavior and locomotor activity of rats are eliminated by hypothalamic lesion. Proc Natl Acad Sci USA. 1972;69:1583–1586.
103. Steriade M. Brain electrical activity and sensory processing in waking and sleep. In: Kryger MH, Roth T, Dement WC, eds. Principles and Practice of Sleep Medicine, 2nd ed. Philadelphia: WB Saunders; 1994:105–124.
104. Steriade M, Contreras D, Amzica F, et al. Cortical and thalamic cellular events during transition from sleep patterns to paroxysmal activity. J Neurosci. 1995;15:623–642.
105. Sterman MB, Shouse MN, Passouant P. Sleep and Epilepsy. Academic Press: New York; 1982.
106. Stevens JR. Sleep is for seizures: a new interpretation of the role of phasic ocular events in sleep and wakefulness. In: Sterman MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. New York: Academic Press; 1982:249–264.
107. Stevens JR, Kodama H, Lonsbury B, et al. Ultradian characteristics of spontaneous seizure discharges recorded by radio telemetry in man. Electroencephalogr Clin Neurophysiol. 1971;31:313–325.
108. Stevens JR, Lonsbury BL, Goel SL. Seizure occurrence and interspike interval. Arch Neurol. 1972;26:409–411.
109. Szuba M, Baxter LJ. Electroconvulsive therapy increases circadian amplitude and lowers core body temperature in depressed patients. Biol Psychiatry. 1997;42:1130–1137.
110. Szymusiak R, Alam MN, McGinty DJ. Discharge patterns of neurons in cholinergic regions of the basal forebrain during waking and sleep. Brain Res. 2000;115:171–182.
111. Szymusiak R, Alam MN, Steininger TL, et al. Sleep–wake discharge patterns of ventral preoptic/anterior hypothalamic neurons in rats. Brain Res. 1998;803:178–188.
112. Szymusiak R, McGinty DJ. Sleep related neuronal discharge in the basal forebrain of cats. Brain Res. 1986;370:82–86.
113. Tassinari CA, Bureau M, Dravet C, et al. Electrical status epilepticus during sleep in children (ESES). In: Sterman MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. New York: Academic Press; 1982:481–494.
114. Terzano MG, Mancia D, Salati MR, et al. The cyclic alternating pattern as a physiologic component of normal NREM sleep. Sleep. 1985;8: 137–145.
115. Terzano MG, Parrino L, Anelli S, et al. Modulation of generalized spike-and-wave discharges during sleep by cyclic alternating pattern. Epilepsia. 1989;30:772–781.
116. Terzano MG, Parrino L, Anelli S, et al. Effects of generalized interictal EEG discharges on sleep stability: assessment of by means of cyclic alternating pattern. Epilepsia. 1992;33:317–326.
117. Terzano MG, Parrino L, Spaggiari MC. The cyclic alternating pattern in the dynamic organization of sleep. Electroencephalogr Clin Neurophysiol. 1988;69:437–447.
118. Terzano MG, Parrino L, Spaggiari MC, et al. Discriminatory effect of cyclic alternating pattern in focal lesional and benign rolandic interictal spikes during sleep. Epilepsia. 1991;32:616–628.
119. Touchon J. Effect of awakening on epileptic activity in primary generalized myoclonic epilepsy. In: Sterman MB, Shouse MN, Passouant P, eds. Sleep and Epilepsy. New York: Academic Press; 1982:239–248.
120. Villablanca J. The electroencephalogram in the chronic cerveau isolé cat. Electroencephalogr Clin Neurophysiol. 1965;19:576–586.
121. Von Economo C. Encephalitis Lethargica: Its Sequelae and Treatment. London: Oxford University Press; 1931.
122. Welsh DK, Logothetis DE, Meister M, et al. Individual neurons dissociated from the suprachiasmatic nucleus express independently phased circadian firing rhythms. Neurons. 1995;14:697–706.
123. West WJ. On a peculiar form of infantile convulsion. Lancet. 1841;1:724–725.