Elsevier

Epilepsy & Behavior

Volume 14, Issue 4, April 2009, Pages 617-621
Epilepsy & Behavior

Diurnal rhythms in seizures detected by intracranial electrocorticographic monitoring: An observational study

https://doi.org/10.1016/j.yebeh.2009.01.020Get rights and content

Abstract

Few studies have evaluated human seizure occurrence over the 24-hour day, and only one group has employed intracranial electrocorticography monitoring to record seizures. Circadian patterns in seizures may have important implications in diagnosis and therapy and provide opportunities in research. We have analyzed spontaneous seizures in 33 consecutive patients with long-term intracranial EEG and video monitoring. Several aspects of seizures were noted, including time of day, origin, type, and behavioral state (sleeping/awake). We recorded 450 seizures that showed an uneven distribution over the day, depending on lobe of origin: temporal lobe seizures occurred preferentially between 1100 and 1700 hours, frontal seizures between 2300 and 0500 hours, and parietal seizures between 1700 and 2300 hours. In the awake state, larger proportions of clinical seizures were seen from 0500 to 1100 hours and from 1700 to 2300 hours. During sleep, larger proportions occurred from 1100 to 1700 hours and from 2300 to 0500 hours. Our results suggest that seizures from different brain regions have a strong tendency to occur in different diurnal patterns.

Introduction

The unpredictability of epileptic seizures is an important factor in the disabling character of the disease. It has been shown that not all seizures occur randomly [1], [2], [3], [4], [5]. Animal studies have provided some answers, as clear diurnal patterns of seizures have been observed in various epilepsy models. For instance, in studies of rodents with limbic epilepsy, it was observed that seizure latency was shorter and more spontaneous seizures occurred during exposure to light than during darkness [6], [7], [8], [9], [10], [11], [12].

In humans, patterns of seizure occurrence have been studied, but most authors have assessed only the random or nonrandom character of these patterns [2], [3], [4]. Data on precise temporal distribution of seizures, however, are particularly scarce. Four studies have provided details on the temporal distribution of seizures using long-term EEG monitoring, three of these four studies used scalp EEG monitoring [5], [8], [13]. In one study, the pattern of partial seizure occurrence in adults was compared with that of seizures in an epileptic rat model [8]. It was found that in patients with mesial temporal lobe epilepsy (MTLE), seizures did not occur randomly, but in a sinusoidal daily distribution, comparable to the occurrence of seizures in rats. In another study, clear patterns in both temporal lobe seizures and extra-temporal seizures were demonstrated [13]. In a recent study, we evaluated the temporal distribution of different types of clinical seizures also using scalp EEG. We observed significantly more seizures from 1100 to 1700 hours and significantly fewer seizures from 2300 to 0500 hours [5].

These three studies have provided valuable information; however, they all used scalp EEG monitoring to record seizures. Scalp EEG may not detect all seizures, especially more localized or subtle seizures and seizures arising from deep structures that cannot be detected on the scalp. In addition, muscle artifacts may obscure a fast seizure EEG rhythm. At present, intracranial electrocorticography (ECoG) monitoring (IEM) is the gold standard for recording seizures and delineating epileptic foci without these limitations. To our knowledge, only one recent study used IEM to evaluate the temporal distribution of seizures [14]. These authors found seizures from the parietal, occipital, mesial temporal, and neocortical temporal lobes to be distributed nonuniformly, with peak occurrences at different times over the 24-hour day. In that study, however, only seizures in adults were included and the behavioral state during seizures (sleep or awake) was not addressed.

To understand the pathophysiology of epilepsy, further details on circadian patterns in seizure occurrence need to be ascertained, preferably using IEM for accurate seizure detection. This knowledge may have significant clinical implications for diagnostic procedures and timing of therapy in epilepsy. Therefore, we performed a retrospective analysis of all children and adults with various types of seizures who recently underwent long-term intracranial ECoG monitoring in the nationwide epilepsy surgery program in The Netherlands.

Section snippets

Intracranial ECoG monitoring and video recording systems

Long-term IEM was performed in the Intensive Epilepsy Monitoring Unit at the University Medical Centre in Utrecht. Subdural strips and grids were used for IEM. Strips were eight-electrode arrays made of 5-mm-diameter platinum disks spaced 10 mm between centers. Grids consisted of a multiple of eight 5-mm-diameter platinum disks (maximum of 64) in a rectangular array with 10-mm center-to-center distance (Ad-Tech, Racine, WI, USA). In one patient, placement of subdural strips through burr holes

Subjects

From 1999 to 2008, the seizures of 33 patients were analyzed. The population consisted of 26 adults (9 women: mean age 29.7 years, range 17–45; mean duration of IEM 4.6 days, range 2–8) and 7 children (4 girls: mean age 11.6 years, range 5–15; mean duration of IEM 4.0 days, range 2–6).

Imaging of the brain with MRI revealed dysplasia in 13 patients, a tumor in 5, mesiotemporal sclerosis (MTS) in 3, and multiple causes (cavernomas and MTS) in one. In another five patients various lesions were found

Discussion

In our patients undergoing IEM, seizures of mesial temporal, neocortical temporal, frontal, and parietal origin occurred in a nonrandom distribution over the 24-hour day. Depending on the lobe of origin, the peak prevalence during the 24-hour day differed.

Temporal distribution of seizures in humans has been studied previously, although knowledge is scarce. In three studies in which seizure diaries were used, seizure occurrence was not specified with respect to clock time. When all three studies

Acknowledgments

We thank Professor Ley Sander for reading the article and his knowledgeable suggestions. All patients were discussed in the Dutch National Task Force on Epilepsy Surgery by neurologists from the epilepsy centers in Heeze, Oosterhout, Heemstede, and Zwolle. This work was supported by the Christelijke Vereniging voor de Verpleging van Lijders aan Epilepsie.

References (18)

There are more references available in the full text version of this article.

Cited by (0)

1

On behalf of the Dutch National Task Force on Epilepsy Surgery.

View full text