Impact of Hyperventilation and Sleep Deprivation Upon Visual Evoked Potentials in Patients with Epilepsy
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.266246
Source of Support: None, Conflict of Interest: None
Keywords: Activation tests (hyperventilation, deprivation of sleep), epilepsy, visual evoked potentials
The diagnosis of epilepsy is often difficult because of the diversity of epileptic etiology, the morphology of seizures, and the types of bioelectrical abnormalities. The intricate and constantly modified classification of epilepsy frequently adds to these problems. The lack of seizure activity in an electroencephalography (EEG) record does not exclude the diagnosis of epilepsy; during the first resting-state EEG, epileptiform activity is only found in 30–50% of the patients.,
In addition to standard EEG and videometric records, the evoked potentials have also been investigated for patients with epilepsy. The cortical evoked potentials are generated in response to a specific stimuli of particular modalities. Their analysis enables the assessment of the bioelectrical activity of the brain, which is especially helpful in subclinical damage to the central nervous system (CNS), in the absence of the structural changes detected by neuroimaging, and in equivocal cases. In case of patients with epilepsy, some abnormalities have been found in the visual evoked potentials (VEP) and the brainstem auditory evoked potentials (BAEP). Farnarier et al. described an increase in the amplitude of the N1/P1 components of the VEP evoked by the stimulation with a reversible pattern of the black-and-white checkerboard and flash in a group of children with idiopathic partial epilepsy; however, in case of children with symptomatic epilepsy and partial seizures, the amplitude of N1/P1 was reduced. The authors suggested that the increase in the VEP amplitude in idiopathic partial epilepsy is associated with pathological modulation of the synaptic transmission, leading to excessive cortical synchronization.
Most previous studies related to the evoked potentials in epilepsy have been based on the analysis of electrophysiological tests carried out on small groups of a dozen or so patients treated or untreated with antiepileptic preparations. Moreover, the published studies are primarily related to the changes of the evoked potentials under the influence of antiepileptic drugs.,,,,, The variety of clinical material and the methods used did not provide the opportunity to conduct a thorough comparative analysis of those results, which have often been inconsistent. Thus, the authors attempted to analyze changes in the VEP in patients with newly diagnosed epilepsy, who have not been treated with antiepileptic drugs yet, including the impact of the standard conditions and additional activation methods upon the electrophysiological parameters. In addition, the relationships between the parameters of the evoked potentials and the type of seizures as well as the seizure activity in EEG were investigated.
Eighty-one participants were enrolled in this study, including 36 females aged 18–72 years (average of 41.5 years) and 45 males aged 18–73 years (average of 38.4 years), with a recent diagnosis of idiopathic epilepsy with focal or generalized seizures who had not yet started antiepileptic treatment. All the patients were under the care of the Department of Neurology, Medical University of Wroclaw during the years 2004–2007. The diagnosis of epilepsy was established on the basis of a reliable medical history and EEG, both at rest and after deprivation of sleep. The exclusion criteria included symptomatic epilepsy, previous cerebral lesions not associated with seizures, and positive electromyographic tetany tests. The control group consisted of 42 healthy volunteers, including 19 females aged 19–71 years (average of 38.2 years) and 23 males aged 18–65 years (on average 37.5 years) with normal EEG recordings and negative electromyographic tetany tests. All the subjects included gave their written consent to participate in the study. The design of the study was approved by the Bioethics Committee at the Medical University of Wroclaw.
The patients underwent the neurological examination, EEG, and VEP procedures. The types of seizures and concomitant medical histories were established on the basis of the medical records. The EEG was performed using a 20–channel Elmiko DigiTrack device, with the standard placement of the electrodes according to the international 10–20 system. Standard recording included hyperventilation (3 min) and photic stimulation (intermittent light stimulus with variable frequency for 3 min). EEG was repeated after 24 h of sleep deprivation. A positive response to the hyperventilation induces, especially in the frontal regions, spikes, sharp waves, spike waves, and sharp wave–slow wave groups. Furthermore, this is also defined as the lateralization of the interictal epileptiform activity or the focal slow activity. The positive response to photic stimulation displays spikes, polyspikes, spike waves, or intermittent slow waves. The positive response may be restricted to posterior visual regions (i.e., local) or spread to anterior nonvisual cortical areas (i.e., with propagation). The epileptiform EEG activity following sleep deprivation was evaluated for wake, drowsiness, and each sleep stage.
The VEP were recorded using a Nicolet CA-1000 device. The surface, cup-shaped electrode (Ag/AgCl) was fixed to the occipital midline, above the external occipital tuberosity. The visual stimulation was performed using a full-field black-and-white checkerboard pattern (the angle size of the individual squares was 1.1°, the entire field was 18° × 22°) and a uniform flash of light (FL) emitted from a Nicolet television monitor, NIC-1005 model, with the frequency of 1.88 Hz and 15 Hz. After 3 min of hyperventilation, the VEP recording was repeated using FL only at the frequency of 1.88 Hz. The latency of the P100 component and P100/N145 amplitude (“peak to peak”) were evaluated. Furthermore, the VEP recordings were repeated after a sleepless night using all the types of stimulation described above (checkerboard pattern and FL at two frequencies).
The EEG and the VEP, according to the described scheme, were also performed on the healthy controls. All the subjects were asked not to take any medication during the 12 h preceding the electrophysiological tests.
Variables in the analyzed groups had normal (or close to normal) distributions, which were evaluated with the Kolmogorov–Smirnov test. Therefore, statistical analysis using parametric tests was possible. Using the Student's t-test, the VEP parameters were compared between the healthy controls and the patients, as well as the healthy controls and the subgroups of patients differing in terms of the type of seizures and abnormalities in EEG. The comparison of the VEP parameters recorded before and after deprivation of sleep was performed using two-way analysis of variance (ANOVA) with repetition. In addition, the one-way ANOVA and Tukey's least significant difference tests were used to compare the VEP parameters between the subgroups of patients differing in terms of the type of seizures and abnormalities in EEG. The statistical significance was accepted at the level of P = 0.05.
Clinical and electroencephalogram findings in persons with epilepsy
Within the group of patients, 42 (51.8%) had primary generalized tonic–clonic seizures, 17 (21%) had absence seizures, 19 (23.5%) had focal seizures with impairment of consciousness or awareness, and 3 (23.5%) had focal seizures evolving into bilateral seizures, previously called secondary generalized seizures [Table 1].
The activation tests in EEG (intermittent photic stimulation (IPS) and hyperventilation test) were used in all 81 patients, whereas EEG after deprivation of sleep was performed on 75 patients. The epileptiform activity in the resting-state EEG was observed in 46 out of the 81 patients (56.8%). In case of patients with very common seizures and the epileptiform of EEG patterns, it was decided not to perform additional sleep deprivation tests. In such cases, there was no doubt about the diagnosis of epilepsy.
Positive responses to the hyperventilation tests in EEG were recorded in 39 patients (48.1%). A group of 35 patients (43.2%) had positive responses to IPS in EEG recordings. Deprivation of sleep evoked or enhanced epileptiform activity in EEG in case of 35 patients (46.7%).
In the initial VEP recording using checkerboard stimulation, no statistically significant differences in P100 latency or amplitude were observed between the patients with epilepsy and controls [Table 2]. After FL stimulation at the frequency of 1.88 Hz P100, latency was significantly longer (P = 0.045) in the patients than in the controls [Table 2]. With the increased frequency of the stimulus (15 Hz), a highly significant shortening of P100 latency was noted for patients when compared to the control group (P < 0.001) [Table 2].
On analysis of the VEP performed after hyperventilation, using the FL at the stimulus frequency of 1.88 Hz, P100 the latency was significantly prolonged (P = 0.006) in patients with epilepsy when compared to the control group [Table 2].
On analysis of the VEP performed after deprivation of sleep, the patients showed significantly prolonged P100 latency at an FL frequency of 1.88 Hz (P = 0.015) and significantly decreased P100 latency at an FL frequency of 15 Hz (P < 0.001) in comparison with the controls [Table 2]. No differences were observed between the patients and controls in terms of the VEP parameters obtained with the checkerboard stimulation.
No significant differences in VEP parameters were found between the patients either with or without epileptiform activity in EEG. The patients with positive or negative hyperventilation tests in EEG did not differ significantly in terms of any of the VEP parameters. Comparing the patients with positive and negative IPS responses in EEG, P100 latency in the former subgroup was found to be significantly prolonged both at rest and when using stimulation with a uniform FL at the frequency of 1.88 Hz (P = 0.04). No such differences were found for other P100 parameters.
There were no significant differences in the mean latencies or amplitudes of the VEP between the subgroups of patients with different types of seizures [Table 3].
In the current study, the cortical evoked potentials were investigated with regards to their contribution to the diagnosis of epilepsy. The majority of the previously reported studies on this subject ,,,, have included relatively small groups of patients with a wide variety of clinical symptoms and treatments. The results of these studies are inconsistent, but methodological differences prevent their analyses from being reliably compared. Our study comprised a group of 81 patients with recent diagnoses of epilepsy, where electrophysiological studies were performed before the antiepileptic treatment was introduced, which encouraged us to expect more relevant results.
No statistically significant differences were found in patients' VEP parameters obtained with binocular stimulation with a checkerboard pattern when compared with the control group. Martinović et al. pointed out that patients with epilepsy with a short duration of the disease present with normal P100 latency, with the VEP obtained using the same type of stimulation. These authors also showed a positive correlation between the P100 latency values and the disease duration and frequency of seizures. Mervaala et al. found a prolongation of the VEP latency in patients with a longer duration of epilepsy and its more severe course. This may suggest that the VEP latency in patients with epilepsy is affected by the disturbances in inhibition processes and the synaptic transmission involved in epileptogenesis, however, the influence of the antiepileptic drugs upon the electrophysiological parameters also has to be considered.
Following the analysis of the VEP obtained with a uniform FL stimulation in patients suffering from epilepsy in comparison to controls, we found that P100 latency was characterized by a high variability, which was dependent on the frequency of FL stimulus. The activation tests (IPS, hyperventilation, sleep deprivation) further increased this variability. Similar findings were also obtained in a subgroup of patients with positive responses to IPS EEG. This differential response to the two modes of visual stimulation, i.e., checkerboard and FL stimulation, can be explained with the differences in the stimulation of different regions of the retina and visual pathway elements and cortex regions. The structural stimulus (reversible checkerboard pattern) activates the ganglion cells located in the center of the retina, and then the stimulation is transmitted via the lateral geniculate body to the striated cortex of the brain. The structural pattern generates the action potential, transferred through small-cell part of the visual pathway, which is characterized by slow conduction and tonic response to a visual stimulus. A uniform FL stimulus activates the large-cell part of the visual pathway, which is characterized by rapid conduction and phase in response to a visual stimulus.,,,,
Prolonged P100 latency in patients with epilepsy may be associated with an altered balance between the excitation and the inhibition processes as a result of disturbed functioning of the neuronal network consisting of excitatory glutaminergic interneurons and inhibitory GABAergic ones. The influence of the GABAergic system on the visual pathways of rats was studied by Hetzler et al. The authors reported prolonged latency of a majority of components of the visual response after the administration of a GABAA agonist, which may suggest the role of the GABAA receptor in modulating the N1-P2 cortical responses.
The hyperventilation test caused significantly prolonged P100 latency in case of our patients with epilepsy, but not in case of the controls. This might be associated with the adverse effects of focal seizures on the autoregulation of cerebral blood flow and an increase in the sensitivity of neurons to ischemia.,,,,, The increase in the intracellular calcium concentration in the vascular smooth muscle cells during hyperventilation leads to vasospasm with a subsequent reduction of regional cerebral blood flow and an increase in the anaerobic metabolism. Under such conditions, glutamate is released, which exhibits a neurotoxic effect of neuronal damage through the N-methyl-D-aspartic acid (NMDA) receptor.,,, In case of healthy volunteers, Gavriysky  found significantly prolonged VEP of the P100 latency wave in the third minute of HV, and a slight shortening in the first and fifth minute after the HV stimulation with the reversible checkerboard pattern. These results could be associated with a temporary prolongation of synaptic transmission in the visual pathways due to hyperventilation. The author also takes into account the inhibitory effect of hypocapnia on the ascending reticular activating part. Hyperventilation has various effects on neuronal excitability and synaptic transmission, perhaps also an increased dysrhythmia in the brainstem-thalamo-cortical network.,
The significantly prolonged P100 latency after sleep deprivation in our patients with epilepsy might be explained by the predominance of the inhibition processes within the visual pathway. Such a hypothesis is supported by the findings of McDermott et al. who demonstrated a significant reduction in the excitability of field CA1 pyramidal cells of the hippocampus in rats after 72 h of sleep deprivation, along with the inhibition of synaptic conductivity and an increase in the latency of VEP.
Visual stimulation with FL at the frequency of 15 Hz, evoking a “steady-state potential” type response, induced an advantage of excitation processes over inhibition, which presumably resulted in a significant shortening of P100 latency in patients with epilepsy, both at rest and after sleep deprivation. The increase in the frequency of the stimulus > 4.5 Hz results in a permanent state of arousal of the visual system, recorded in the form of a sine wave with the maximum amplitude of oscillation at the stimulus frequency of 8–10 Hz. This type of response is defined as steady-state evoked potentials. They are useful for determining the amplitude of the visual responses. The latency of steady-state potentials is not equivalent to a transient VEP P100 latency. In their experimental studies on rats, Bale et al. demonstrated the important role of excitatory receptors (NMDA and nicotinic) and the inhibitory receptor (GABA) in generating the visual response of steady-state potentials. The authors found a significant increase in the VEP amplitude as a result of stimulation of the NMDA and nicotinic receptors and a decrease in amplitude after the stimulation of GABAergic receptors. Although we observed no significant effect of high frequency stimuli on the VEP amplitude in our patients with epilepsy, the suggested mechanism may be associated with the background of another of our findings: the shortening of P100 latency at a higher FL frequency. On the basis of the abovementioned observations, one might hypothesize that the steady-state arousal of the visual system enhances the activity of glutamate receptors in patients with epilepsy.
We found no significant relationships between the parameters of VEP and EEG. In the single available report on this subject, Lücking et al. ascertained a greater stability of regularly repeated VEP in individuals with epilepsy and normal EEG than in those with paroxysmal activity. We also did not demonstrate significant differences in VEP parameters among a subgroup of patients with particular types of seizures. Similar findings were reported by Donáth  who studied VEP obtained with monocular stimulation with a reversible checkerboard pattern in 72 patients with various types of seizures and treated with antiepileptic medications. However, Rysz and Gajkowski  found the longest P100 latency and the highest N75/P100 amplitude in a group of patients with primary generalized seizures, even though their values remained within normal limits.
Further analysis of the VEP in our study revealed a significant relationship between the positive IPS response during EEG and the prolongation of P100 latency with the FL stimulation at the frequency of 1.88 Hz. According to many authors, for patients with the positive EEG response to photic stimulation, visual stimulus induces hyperactivity of neurons in the occipital cortex with the spreading of excitation through the corticospinal and the brainstem connections. This leads to increased activity of brainstem neurons, which in turn suppresses the cortical ones.,,,,, Contrasting results were obtained by Faught and Lee  and Bülent  who performed a VEP study using a structural stimulus (reversible checkerboard). They showed shortened P100 latency in patients with photoparoxysmal response in EEG, and suggested that the hyperexcitability of the visual cortex in patients with epilepsy is genetically determined.
In this study, VEP were investigated in persons with epilepsy using the structural and FL stimuli and additional activation methods, lowering the threshold of cortical excitability. Changes in the VEP latency were shown depending on the type of stimulation and the frequency of stimulus. These findings suggest the important role of an altered balance between the glutamatergic and the GABAergic systems in the excitability of visual neural networks in patients with epilepsy, especially during the visual stimulation typical for steady-state potentials.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2], [Table 3]