The current study evaluated the diagnostic value of electroencephalographic recordings (EEG) in cats with epilepsy under special consideration of photic stimulation and hyperventilation. EEGs in six healthy cats were recorded under light (mean dose of 0.23 mg/kg/min) and deep (mean dose of 0.7 mg/kg/min) propofol anaesthesia, whereas EEGs in 13 diseased cats were recorded under a propofol anaesthesia which was kept as light as possible (mean dose of 0.39 mg/kg/min). Paroxysmal discharges were detected in six of 13 cats suffering from seizures (two cats with idiopathic epilepsy and four cats with symptomatic epilepsy). Activation techniques did not enhance the diagnostic value of the EEGs. Photic driving was detected in one of six healthy cats under light, in five of six healthy cats under deep propofol anaesthesia and in 11 of 13 cats with seizures. Systematic use of activation techniques does not seem to increase the diagnostic yield of the recorded EEGs and should not be used in a clinical setting until future studies indicate value. Further investigations into the origin of photic driving under propofol anaesthesia are needed and could lead to the development of a reliable animal model to research into drug effects on the EEG.
Statistics from Altmetric.com
SEIZURES in cats have a prevalence of 2.1 per cent (Schriefl and others 2008) and are less commonly observed than in dogs (Schwartz-Porsche 1994). In addition, idiopathic epilepsy seems to occur more frequently in dogs than in cats (Schwartz-Porsche 1994, March 1998, Rusbridge 2005). Furthermore, in cats symptomatic epilepsy is observed more frequently than reactive seizures or idiopathic epilepsy (Quesnel and others 1997, Barnes and others 2004, Rusbridge 2005, Schriefl and others 2008).
Due to the fact that idiopathic epilepsy in cats is less common, all diagnostic options have to be used to rule out all other causes of seizures to find a specific treatment for the presented animal (Fatzer and others 2000). These diagnostic investigations are in particular: thorough history-taking, physical and neurological examination, complete blood cell count, serum biochemistry, including liver function evaluation (ammonia and/or preprandial and postprandial bile acids), tests for feline leukaemia virus and feline immunodeficiency virus, radiographs of the thorax and abdomen, MRI of the brain and cerebrospinal fluid (CSF) analysis (Rusbridge 2005).
Another important diagnostic tool in seizure diagnostics, especially in human medicine, is the electroencephalographic (EEG) recording of the electrical brain activity (Mendez and Brenner 2006). Nowadays, in veterinary medicine, CT and MRI are widely available and commonly used to evaluate the cranial vault for structural and inflammatory brain disease as a potential reason for the underlying seizure origin (Podell 1996). Nevertheless, there are several advantages of EEG recording over MRI as it illustrates the electric activity function of the brain and may enable detection of a seizure focus and monitoring of treatment regimens (Jaggy and Bernadini 1998). Another benefit of EEG recording is its potential to identify epileptiform discharges and background abnormalities which may be helpful to differentiate between seizure and seizure-like phenomena (Penning and others 2009) or movement disorders (Kube and others 2006).
In the past, feline EEG research focused on healthy cats and different anaesthetic protocols (Gustafson Beaver and Klemm 1973, Wrzosek and others 2009) as well as individual descriptions of pathological patterns of the EEG activity (Croft 1962, Klemm 1968, Croft 1972). Therefore, the first aim of the current study was to compare EEGs of healthy cats and interictal recordings from those suffering from seizures, using the same anaesthetic drug for both groups.
The additional application of activation techniques is a common procedure in human medicine in order to increase the diagnostic value of the EEG (Mendez and Brenner 2006). In a large retrospective study on 1000 human EEGs, the authors found that one or more activation techniques were used in 85.5 per cent of the recorded EEGs and added useful information in about 11 per cent of all cases in which the routine resting EEG was normal or equivocal (Angus-Leppan 2007). The second aim of the present study was to evaluate the diagnostic usefulness of two activation techniques, photic stimulation and hyperventilation, in interictal feline EEG recordings.
Materials and methods
Six clinically healthy domestic shorthair cats from a cat colony of the University of Veterinary Medicine Hannover were investigated. The median age and weight were 30 months (range 22 to 106 months) and 4.95 kg (range 3.7 to 5.3 kg), respectively. Physical and neurological examination as well as complete blood count and serum biochemistry measurements before anaesthesia did not reveal any abnormalities. All procedures fulfilled the requirements of the German Animal Welfare Act and were approved by the Federal State Office for Consumer Production and Food Safety of Lower Saxony, Germany (AZ 09/1792).
Cats with seizures
In addition to the examination of healthy cats, EEGs of 13 client-owned cats suffering from recurrent seizures were also recorded. Animals underwent physical and neurological examination as well as routine blood work and thoracic and abdominal radiographs before anaesthesia. After EEG recording, cats were further investigated by MRI (n=11) of the brain and/or CSF tap examination (n=12). Final or presumptive diagnosis was established based on these examination results and in one cat on postmortem examination of the euthanased animal.
Four domestic shorthair cats with a mean age and weight of 21 months (range 10 to 73 months) and 4.35 kg (range 3.3 to 5.3 kg) suffered from presumptive idiopathic epilepsy. One of these animals had been pretreated with phenobarbital. No abnormalities were found in physical and neurological examinations, as well as routine blood work including ammonia levels, thoracic and abdominal radiographs, MRIs of the brain (three cats) and/or CSF taps (four cats; Table 1).
Nine cats (eight domestic short hairs and one Turkish van) with a mean age of 34 months (range 14 to 199 months) and weight of 4.5 kg (range 1.25 to 6.0 kg) displayed abnormalities on physical examination, neurological examination, complete blood count, serum biochemistry, MRI (eight cats) and/or CSF examination (eight cats) which were indicative for symptomatic epilepsy (Table 1). Two animals with symptomatic epilepsy had been pretreated with phenobarbital, three with diazepam and one with phenobarbital and diazepam.
In the healthy animals, anaesthesia was induced with propofol. A mean dose of 6.85 mg/kg (range 5.67 to 8.11 mg/kg) was needed before endotracheal intubation was possible. General anaesthesia was maintained with propofol constant rate infusion (CRI) during EEG recording. In three of the six healthy cats, anaesthesia was first held at a light plane until the first EEG recording period of 18 to 24 minutes (mean 20 minutes) was completed. Propofol CRI was then increased to 0.7 mg/kg/min. Thirty minutes later, another EEG recording period (mean 17 minutes) was started. In the other three healthy cats, propofol CRI was first held at high propofol rates of 0.7 mg/kg/min. After the first EEG recording period of 17 to 19 minutes (mean 18 minutes), propofol CRI was stopped. It was restarted at a lower level 30 minutes after cessation and another EEG recording period (mean 17 minutes) followed. It was randomised whether animals started with low or high propofol CRI. Mean doses for light and deep propofol anaesthesia were 0.23 mg/kg/min (range 0.13 to 0.33 mg/kg/min) and 0.7 mg/kg/min, respectively. In order to avoid muscle artefacts in the EEG, 0.6 mg/kg rocuronium bromide was administered before the first EEG recording period (Brauer and others 2011a). During anaesthesia, lactated Ringer's solution was administered intravenously at a rate of 5 ml/kg/h. The cats' lungs were ventilated by intermittent positive pressure ventilation. The oxygen flow rate was 1.0 l/min delivered via a small animal rebreathing system. Peripheral oxygen saturation of haemoglobin, pulse rate and end-tidal carbon dioxide (EtCO2) tension were constantly measured and monitored on a multiparameter monitore. During phases without hyperventilation, the carbon dioxide tension was kept at levels between 35 and 45 mm Hg.
Cats suffering from seizures were also anaesthetised with propofol for EEG recording, but general anaesthesia was kept as light as possible (based on clinical assessment of reflex activity, for example, swallowing reflex, palpebral reflex before administering rocuronium bromide). The animals underwent one EEG recording period of 14 to 21 minutes (mean 17 minutes) before further investigation with MRI and/or CSF examination. Mean induction and maintenance doses of propofol were 8.87 mg/kg (range 5.77 to 15.15 mg/kg) and 0.39 mg/kg/min (range 0.3 to 0.61 mg/kg/min), respectively. All but one of the cats received rocuronium bromide at a dose of 0.4 mg/kg intravenously for preventing muscle artefacts.
EEGs were recorded with a mobile electroencephalograph. Cats were placed in sternal recumbency and five subdermal needle electrodes (F3, F4, Cz, O1 and O2) were placed over the calvarium in order to record the EEG (Fig 1; Redding 1978). Acquisition parameters for EEG recording were sensitivity =70 µV/cm; time constant =0.3 seconds, high-frequency filter =70 Hz; 50 Hz notch filter inserted; impedance of all electrodes <10 kΩ; reference electrode on the bridge of the nose; ground electrode caudal to the external occipital protuberantia. Another two subdermal needle electrodes were used to record a lead II electrocardiogram.
The EEG recording procedure was supplemented by two activation techniques. First, an intermittent light stimulation was carried out in a normally lightened room with a photic stimulator placed about 20 cm in front of the opened, not ventrally rotated eyes. Flash frequency started at 5 Hz, increased in 5 Hz steps until 50 Hz was reached and decreased again down to 5 Hz. Each frequency was applied for 8 seconds (Stöhr and Kraus 2002). After a stimulation-free period of about three minutes, hyperventilation was started in order to reduce the constantly measured EtCO2 tension to a mean value of 25 mm Hg (range 18 to 34 mm Hg) within a mean duration of 211 seconds (range 180 to 373 seconds). Another three minutes were recorded after cessation of hyperventilation (posthyperventilation period; Flink and others 2002).
EEGs were examined visually in referential and bipolar montages. Paroxysmal discharges (eg, spikes, spike-wave discharges) as well as possible artefacts were identified and marked. Background activity was analysed for any visually detectable changes (eg, photic driving, focal slowing) during the recording procedure (Mendez and Brenner 2006).
None of the healthy cats showed any paroxysmal discharges in the EEG. Visually, background activity under light propofol anaesthesia was dominated by theta and delta activities with superimposed beta activity (Fig 2a). In deep propofol anaesthesia, a burst suppression pattern occurred in all six animals (Fig 2b). Background activity did not change under hyperventilation, neither in light nor in deep anaesthesia. Photic driving occurred in one cat under light anaesthesia and in five animals in deep anaesthesia (Fig 3). It was present in all these animals at lower flash frequencies of 5, 10 and 15 Hz and only in one cat also at higher frequencies of up to 45 Hz (Fig 4) regardless if the frequency was increasing or decreasing. This special rhythmic activity occurred in all montages and disappeared when the eyes were covered with a cardboard.
Interictal spikes were detected by visual analysis in EEGs of two cats (50 per cent) with presumptive idiopathic epilepsy and in four of nine cats (44 per cent) with symptomatic epilepsy (Figs 5, 6 and 7). There was no sudden onset of paroxysmal activity during the use of activation techniques. The EEGs of three cats with symptomatic epilepsy were marked by burst suppression patterns. All other EEGs of cats suffering from seizures were dominated by delta and theta rhythms which were superimposed by alpha and beta activities. Background activity did not visually change during hyperventilation of these cats but photic stimulation induced photic driving in all but two epileptic cats (85 per cent, 11/13 cats). Whereas cats suffering from idiopathic epilepsy showed photic driving only at lower flash frequencies of 5, 10 and 15 Hz, in animals suffering from symptomatic epilepsy photic driving occurred also at higher frequencies of 20 Hz (four cats), 25 Hz (three cats) up to 45 Hz (two cats) and 50 Hz (one cat); (Fig 4).
Only a few studies on EEG recordings in cats exist in the literature. Gustafson Beaver and Klemm (1973) described the disadvantages of manual restraint or restraint with light sedation and presented EEGs in healthy cats under general anaesthesia with sodium thiopental and pentobarbital. About 35 years later, Wrzosek and others (2009) investigated the effect of medetomidine sedation on the EEG of healthy cats. Both studies provided methods which could be further used in routine clinical EEG investigations of cats suffering from intracranial diseases. In addition to these two publications, some authors included a few cats besides a large number of dogs in their EEG studies on animals with neurological diseases (Croft 1962, Klemm 1968, Croft 1972). In the study from Croft (1962), two of the three investigated cats suffered from seizures. In one of them, the EEG was not further described; in the other one the EEG was normal. In his study from 1972, Croft recognised abnormal EEG activity in all six investigated animals suffering from space-occupying lesions of the brain. In 1968, Klemm reported that the EEG of one cat suffering from seizures showed generalised spikes and sharp waves. To the authors' knowledge, the current study is the first study investigating EEGs in healthy cats and cats suffering from seizures using a similar anaesthetic protocol. In addition, it is the first report on the systematic use of two different activation techniques in this species.
In the current study, interictal paroxysmal discharges, which are indicative of epilepsy, could be detected in 50 per cent of the animals (two of four cats) with idiopathic and in 44 per cent (four of nine cats) of symptomatic epileptic cats. Although activation techniques did not enhance the diagnostic value of these short-time recordings, the phenomenon of photic driving was detected in one of six cats in light propofol anaesthesia, in five of six healthy cats in deep propofol anaesthesia and 11 of 13 diseased cats. Despite the occurrence of photic driving not contributing to classification of the underlying disease of recurrent seizures, this phenomenon warrants further consideration in research projects on epileptic cats and drug effects on their EEGs.
No paroxysmal discharges in the EEGs of healthy cats were found in the current study, which is consistent with the results of Wrzosek and others (2009). Despite the use of general anaesthesia for restraint, abnormal EEG activity was observed in six of 13 cats (46 per cent) suffering from seizures.
The occurrence of paroxysmal discharges in the EEGs of epileptic dogs varies among laboratories between 20 and 86 per cent (Holliday and others 1970, Jaggy and Bernadini 1998, Berendt and others 1999, Jeserevics and others 2007). In the current study, 54 per cent of all investigated cats did not show any abnormal activity in their EEGs although they had a history of generalised seizures as described by the owners. In human medicine, activation techniques are used to increase the diagnostic value of the EEG (Mendez and Brenner 2006) and were effective in 11 per cent of all patients with normal routine EEG. In the current study, we adapted two activation techniques which have been successfully used in human medicine to the EEGs in diseased cats in order to attempt to provoke paroxysmal discharges.
Although both techniques, photic stimulation and hyperventilation, did not induce any sudden onset of paroxysmal discharges in any of the investigated cats, 11 of 13 diseased cats displayed EEG recordings with photic driving at least at low flash frequencies of 5 and 10 Hz. Interestingly, four of six healthy cats developed this phenomenon only at deep levels of propofol anaesthesia, whereas just one cat showed also at a light plane of propofol anaesthesia. In contrast, in healthy beagle dogs and dogs suffering from seizures which were previously studied with the same anaesthetic and activation method protocol, none of the investigated dogs showed photic driving at any flash frequency (Brauer and others 2011b).
Abnormal reactions to photic stimulation are so-called photoparoxysmal responses which are marked by spike and wave activity or polyspikes and are only elicited in about 0.5 per cent of normal subjects and about 7 per cent in human epilepsy patients (Kasteleijn-Nolst Trenité 2005). Use of photic stimulation in the current study was not able to elicit any photoparoxysmal response neither in epileptic nor in healthy cats, which might be a result of the small number of investigated animals.
A normal reaction to photic stimulation can be a rhythmic activity in the EEG which is also called photic driving (Kasteleijn-Nolst Trenité 1999, Aminoff 2005) and is a response over the posterior regions of the head characterised by a rhythmic activity time-locked to the stimulus, having a frequency identical to that of the flickering light (Aminoff 2005). Not all normal human beings show such a pattern in their EEG, and sometimes it occurs at specific flash frequencies only (Aminoff 2005). In human beings, differences in amplitudes of the EEG waves during photic driving of more than 50 per cent have to occur simultaneously to other EEG abnormalities to make the EEG suspicious for underlying structural lesions (Aminoff 2005). On the other hand, an asymmetry in the development of driving response is often correlated with ipsilateral focal slowing and the presence of structural lesions (Coull and Pedley 1978).
Propofol has multiple mechanisms of action as it directly activates γ-aminobutyric acid A (GABAA)-receptors, inhibits the N-methyl-D-aspartate receptor and reduces extracellular glutamate levels through either inhibiting Na+-channel-dependent glutamate release or through enhancing the glutamate uptake (Trapani and others 2000, Kotani and others 2008). In addition, propofol decreases the firing rate and the burst activity of dopamine neurons in the substantia nigra, probably mediated via the activation of GABAB receptors (Schwieler and others 2003), resulting in a reduced dopamine terminal efflux (Nissbrandt and others 1994). Concerning the results of the current study, the frequent occurrence of photic driving in healthy cats in deep propofol anaesthesia may therefore be a result of GABAB-receptor activation and lower dopamine levels.
Propofol was chosen in the current study because it has been used successfully for EEG recording in dogs (Accatino and others 1997, Bergamasco and others 2003). Dose-dependently, it may have proconclusive and anticonvulsive properties in human beings and animals although its mechanism of action has not been fully understood so far (Baraka and Aouad 1997, Bergamasco and others 2003, Löscher 2009). Regarding the cats showing photic driving at lower propofol levels, too, this phenomenon might be a function of disturbed dopamine efflux in these animals due to earlier GABAB-receptor activation.
Photic driving also occurred in 11 of 13 diseased cats when anaesthesia with propofol was kept as light as possible with a mean propofol CRI dose of 0.39 mg/kg/min. The development of photic driving in these animals might be due to the slightly higher propofol dose (0.39 mg/kg/min compared with 0.23 mg/kg/min in the healthy cats) or due to changes of neurotransmitters in context with the underlying diseases.
The mechanism by which hyperventilation provokes changes in the EEG is, even in human medicine, still unclear and several theories have been discussed (Patel and Maulsby 1987). It can lead to spike-wave discharges in human beings suffering from generalised epilepsies and to enhanced focal abnormalities in those with partial epilepsy (Aminoff 2005). In a study on 580 human EEGs with hyperventilation as activation procedure, epileptiform activity was seen in 72 patients who are epileptic (Angus-Leppan 2007). Generally, hyperventilation seems to be more effective in generalised epilepsy than in focal epilepsy (Mendez and Brenner 2006). In the present study, hyperventilation did not contribute to diagnostic outcome in all investigated cats although they were all suffering from generalised seizures. This may be due to the low number of animals or due to the fact that they were under general anaesthesia.
The major limitation of this study is the small number of investigated animals. For statistic evaluation, a larger number of cats with seizures due to idiopathic and symptomatic epilepsy would be needed. In addition, the investigated groups were very heterogeneous regarding the type of epilepsy and the corresponding lesions, any prior anticonvulsive treatment and age. Especially different prior anticonvulsive treatment modalities might have an impact on the occurrence of paroxysmal discharges in the EEG as has been shown before in dogs (Jaggy and Bernadini 1998). Since rhythmic activity after photic stimulation during propofol anaesthesia occurred in healthy and diseased cats, an extension of the current study including more seizuring animals and using the described activation techniques seemed not to be useful to enhance the information for clinical diagnostic workup.
In the current study, the effect of photic driving could be described for the first time in healthy and epileptic cats under propofol anaesthesia. The diagnostic value of EEG recording in epileptic cats using this anaesthetic protocol is comparable with the current data in epileptic dogs. Recording EEGs in cats suffering from seizures can add unique information to the diagnosis at least in a part of the patients. EEG recordings might be of particular importance in animals with unclear forms of seizures. Furthermore, feline EEG in combination with photic stimulation under propofol anaesthesia may be a promising animal model to investigate drug effects on the EEG, but EEG is not recommended in the routine veterinary diagnostic workup.
Christina Brauer received a Doctoral Scholarship from the Ministry for Science and Culture of Lower Saxony, Germany.
The authors thank Britta Bösing, Sonja Steinmetz and Jonathan Raue for their kind assistance during the study of the healthy cats and Prof. Dr. Wolfgang Löscher from the Department of Pharmacology, Toxicology and Pharmacy, University of Veterinary Medicine Hannover, Germany, and Dr. Alois Ebner from the Epilepsie-Zentrum Bethel, Mara Krankenhaus, Bielefeld, Germany, for proof-reading this manuscript.
Dr C. Brauer, Dr S. B. R. Kästner and Dr A. Tipoldis also at the Center for Systems Neuroscience Hannover, University of Veterinary Medicine Hannover, Bünteweg 17, D - 30559 Hannover, Germany
Provenance not commissioned; externally peer reviewed
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.