A rapid and rational clinical reasoning approach will not only help to identify the possible underlying pathology of an epileptic dog, but will also help to give the likely prognosis and allow treatment to be initiated swiftly to ensure the best possible outcome for the patient. This article will provide a step-by-step approach for clinical reasoning to improve clinical efficiency and to maximise patient care.
- Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://group.bmj.com/group/rights-licensing/permissions
Statistics from Altmetric.com
RECURRENT seizures are one of the most frequent neurological presentations with an estimated prevalence of 0.6 to 2 per cent in first-opinion practice (Schwartz-Porsche 1986, Schriefl and others 2008, Kearsley-Fleet and others 2013). Epilepsy appears to be more common in pure-breed dogs (Bellumori and others 2013) with a prevalence of up to 33 per cent within single breeds (Berendt and others 2008, Ekenstedt and Oberbauer 2013). Dogs with epilepsy have an increased risk of developing behavioural changes, reduced quality of life and early death (Chang and others 2006, Berendt and others 2007, Shihab and others 2011, Wessmann and others 2012). In human medicine, the best improvement in quality of life for epilepsy patients is achieved when treatment leads to freedom from seizures (Birbeck and others 2002, Poochikian-Sarkissian and others 2008, Kwan and others 2010), and this should be the aim for the canine patient.
Importance of classification systems
The brain, despite its complexity, has only certain ways of expressing dysfunction. A seizure is a clinical sign and not one single disease. A plethora of intrinsic and extrinsic causes can lead to seizures.
To try and classify any complex disease with a simplified classification system (Box 1) will result in an academic debate, as ‘not one shoe will fit all feet, despite having the correct shape’, eg, a genetic mutation, such as the one in the Epm2b gene, causes a metabolic disorder (Lafora body storage disease) that causes recurrent seizures (Lohi and others 2005). Nevertheless, practically speaking as a clinician when presented with a seizuring patient, it is more important to think pathophysiologically and determine how different pathologies can either alter brain function directly or indirectly. Therefore, most neurologists continue to divide the work-up in to an intracranial and an extracranial approach (eg, Chandler and Volk 2008). Extracranial causes of seizures (metabolic or toxic) can often be determined by a detailed history, blood and urine tests and imaging modalities accessible in first-opinion practice, such as ultrasonography. On the other hand, the suspicion of an intracranial lesion usually requires advanced imaging techniques (CT or ideally MRI) and cerebrospinal fluid (CSF) analysis to help differentiate the various aetiologies (Chandler and Volk 2008). MRI, especially using high-field 1.5 or 3T MRI scanners, is the better modality for identifying intracranial lesions affecting brain structure, but it can be cost-prohibitive, if available at all (Bush and others 2002). Up to 50 per cent of all dogs that present with seizures can have brain lesions identified on MRI (Bush 2002). Brain neoplasms, which cause seizures, are most commonly highly vascularised, cause mass effect and are located in the frontal lobe (Schwartz and others 2011). But a dog younger than six years old with generalised seizures, a normal extracranial work-up and without interictal neurological deficits has less than a 3 per cent chance of being diagnosed with an intracranial pathology (Smith and others 2008).
In human medicine, the International League Against Epilepsy (ILAE) has embarked on the endeavour to classify the different epilepsies based on their underlying aetiologies:
▪ Genetic or primary;
▪ Structural/metabolic, including symptomatic; and
▪ Unknown, if no underlying condition can be identified (Berg and others 2010).
This revised classification system continues to be discussed and a new classification system is currently under review. The human classification system can be mirrored for canine epilepsy:
▪ Primary/genetic epilepsy. Formerly most often termed as idiopathic epilepsy, which remains the most common epilepsy type in the dog;
▪ Structural epilepsy. ‘Symptomatic’ or secondary epilepsies being caused by a structural disease process;
▪ Reactive seizures. Formerly termed by some as symptomatic or secondary epilepsies being caused by metabolic disturbances or intoxication, which when corrected can cease seizure activity; and
▪ Unknown (Ekenstedt and Oberbauer 2013).
Seizure semiology and lateralisation
In human medicine, seizure semiology, using the symptoms and clinical signs caused by a seizure to determine its origin, is thought of as a simple, clinically efficient and cost effective tool (Tufenkjian and Luders 2012). The location of the symptomatogenic zone usually overlaps or is in close spatial relationship to the epileptogenic zone and therefore indicates the origin of the seizure. In veterinary medicine, seizure semiology is less useful, as electroencephalography (EEG) recordings are more challenging to perform in veterinary patients and therefore the origin of the electrical altered epileptic focus is difficult to establish. However, there is a more than three times chance that a dog in which the seizure onset is asymmetric, for example, initially only one side of the face is twitching, will have intracranial structural brain disease (structural epilepsy) (Fig 1, Table 1) (Armasu and others 2012). Depending on the epileptic focus, motor, sensory and vegetative changes and automatisms can be seen, which can also help the clinician to determine which parts of the brain might be affected by seizure activity (seizure onset and propagation).
How can the clinician determine if seizures are due to brain disease or not? As with all clinical conditions, a step-by-step methodical approach will guide the clinician on the route of success to achieve a likely diagnosis (Fig 2):
▪ Define the problem;
▪ Define the system;
▪ Define the location;
▪ Define the lesion.
Define the problem
History – what is the presenting complaint?
Most patients with a seizure disorder appear healthy at the time of presentation. As a clinician, one usually does not observe the seizure and therefore we have to rely on the owner describing the episode. Most owners will nowadays have the possibility of getting a video of the event, which can help to differentiate the various episodic disorders (Videos 1, 2, 3). Common episodic events that need to be differentiated from seizures are syncope, narcolepsy/cataplexy, pain, vestibular attacks, movement disorders and neuromuscular weakness (Chandler and Volk 2008).
Episodes of syncope are usually characterised by a sudden, short, transient loss of consciousness and postural tone. The animals appear flaccid during the episode, but can have a brief myoclonus or jerk before the animal collapses. Especially in cats with third-degree atrioventricular block these episodes can be confused with brief focal seizures (Penning and others 2009). Usually there are no pre- or postepisodic signs, the episode is associated with exercise rather than out of rest, the recovery is usually instant, multiple episodes can occur shortly after each other and they do not respond to antiepileptic drugs. Actually, antiepileptic drugs can cause a reduction in cardiorespiratory function and therefore these episodes can get worse with antiepileptic drug treatment (Penning and others 2009).
The rather rare sleeping disorder narcolepsy can cause cataplectic attacks that can resemble seizure activity. Cataplectic attacks can be triggered by food, excitement, stress and pharmacologically with physostigmine. Following the trigger the affected animal will become flaccid and collapse. Narcolepsy is a sleeping disorder, which is characterised by an altered sleep-wake cycle. Familial (eg, dobermann and labrador retriever) and sporadic forms exist (Toth and Bhargava 2013). The pathophysiology was first described in the dog. The most common forms are caused by a defect in the hypocretin receptor 2 gene or are secondary to reduced CSF hypocretin levels.
Animals that experience episodic pain can show episodic behavioural changes resembling seizures, for example, cervical nerve roots can get irritated by a lateral disc protrusion/extrusion resulting in freezing or jerking/neck spasms/myoclonus type movements.
Compulsive behavioural changes
Compulsive behavioural changes can also resemble sensory seizures, such as episodes of aggression or compulsive-tail-chasing. As with idiopathic epilepsy, animals are usually normal in between episodes. Compulsive behavioural changes, however, are not associated with changes in muscle tone or in the level of consciousness and usually a behavioural trigger can be identified. A detailed history, and a thorough physical and neurological examination can help differentiate these episodes from seizures.
Transient vestibular episodes
Transient vestibular episodes can be difficult to diagnosis, especially if they are secondary to presumed periodic blood perfusion issues. The animals show the common clinical signs of head tilting, nystagmus and ataxia episodically. Most of the patients will have no altered consciousness during an episode. These episodes will not respond to standard antiepileptic drug treatment.
Episodic movement disorders
In the past couple of years, episodic movement disorders are becoming better characterised and recognised. Some of these disorders were considered in the past to be seizure disorders. Canine epileptoid cramping syndrome or ‘Spikes disease’ in border terriers was formerly considered by some neurologists as a focal seizure disorder, and now it is thought to resemble a paroxysmal dyskinesia reported in people (paroxysmal dystonic choreoathetosis). This movement disorder is characterised by episodic mild tremors, dystonia and difficulties walking (Black and others 2013).
Genetics has helped further in characterising these disorders. Recently, two independent research groups found a deletion in BCAN, which is associated with episodic falling in Cavalier King Charles spaniels, and developed a gene test (Forman and others 2012, Gill and others 2012). Episodic falling is characterised by episodes of tetany, hypertonicity and ‘deer-stalking’. Most of the movement disorders are initiated or signs deteriorate when the animal is excited or stressed. They usually are associated with movement, rarely occur at rest or out of sleep, are episodic and involve an increase in muscle tone (dystonia), do not affect consciousness, do not respond to certain antiepileptic drugs and are usually shorter than a seizure in duration (Chandler and Volk 2008). Movement disorders can affect only specific body parts, such as the head in idiopathic head tremors, which has been reported in dobermanns and English bulldogs (Wolf and others 2011, Guevar and others 2014). Idiopathic head tremors can resolve in 50 per cent of the cases spontaneously. If you as a clinician are presented with a pure breed dog, which has a paroxysmal episode that does not look like a generalised tonic-clonic seizure and does not respond as well as expected to antiepileptic drugs, a search of the relevant web databases for a breed-specific movement disorder should be considered.
Most of these paroxysmal disorders can be differentiated from focal seizures by getting a clear description from the owner of the character of the episodes and/or ideally have a video of the episode to phenotype the event (see Boxes 2 and 3).
Characteristics of paroxysmal disorders other than seizures
In general, ‘non-seizure’ paroxysmal episodes lack:
▪ An identifiable precipitating event like an aura – sensory seizure activity, such as behavioural change (attention seeking, sniffing, starring), lasting a couple of minutes just before the motor seizure activity;
▪ Vegetative signs, for example, hypersalivation;
▪ Generalisation of motor activity, for example, generalised tonic or tonic-clonic seizure; and
▪ An impairment of consciousness. Usually dogs with an impaired consciousness will not be able to look in the owner's eyes during the event, so this is a good question to ask the owner and is better than asking them if they think their animal can hear them, most owners will say that they can.
Characteristics of seizures
Common characteristics of seizures:
▪ Increased muscle tone is far more likely in seizures. The most common recognised seizure is a generalised tonic-clonic seizure. The animal gets stiff first (tonic phase), loses proprioception and collapses into lateral recumbency, then the tonic-clonic phase starts followed often by running movements. Atonic seizures are rare and a ‘floppy’ collapse should guide the clinician to consider syncope or cataplexy as a more likely explanation.
▪ Rhythmic alternating muscle contractions are common in both focal and generalised seizures.
▪ Seizures often first involve facial muscles (eye or facial muscle twitching).
▪ Stereotypical. Most animals will have only one type of seizure onset that is generalised or can be secondary generalised. Seizures typically originate from the same epileptic focus and propagate through similar pathways.
▪ The ictus (seizure itself) normally lasts one to two minutes.
▪ Most seizures exhibit several stages – preictal behavioural changes, (prodrome [hours to days] and/or aura [minutes]), ictus, postictal behavioural or neurological deficits (hours to days). Apart from the seizure, the postictal changes are most likely to be recognised by the owner.
▪ Seizures often, but not always, occur at rest or out of sleep.
▪ Most of the seizure disorders will at least initially respond to antiepileptic treatment.
▪ The golden standard to differentiate a seizure from other paroxysmal events is an ictal electroencephalography recording; however, this is rarely possible, although interictal recordings can also be useful.
Video recordings of a cat and two dogs with seizures. These videos can be viewed at inpractice.bmj.com/content/supplemental
Define the system
As the presenting complaint is seizure(s), the system that needs to be examined closer is the central nervous system, which can be either directly or indirectly affected.
Define the location
Seizures are always a sign of forebrain dysfunction. Normal cerebral activity is regulated by feedback loops, a network of inhibitory interneurons. If there is an imbalance between excitation (too much) and inhibition (too little), hypersynchronisation of neural networks occurs, which lead to seizures. These changes can either occur focally (epileptic focus – focal seizure onset) and then potentially spread throughout the brain (second generalisation) or involve the whole brain at once (generalised seizure onset). The neurological examination will therefore need to focus on the evaluation of forebrain function; however, it is important not to ignore the rest of the neurological examination, as the identification of multifocal or widespread neurological disease will alter your clinical decision making.
The forebrain might be indirectly or directly affected; seizures occur secondary to extracranial causes and can be accompanied with symmetrical clinical signs and neurological deficits (Figs 1, 3) or intracranial causes (Fig 4). Intracranial causes can be further subdivided into functional disorders: idiopathic epilepsy, no visible structural changes of the brain on MRI, or gross pathological examination and therefore unremarkable interictal neurological examination; and structural diseases: presence of gross structural changes of the brain causing asymmetrical neurological deficits, for example, neoplasms, inflammatory/infectious causes, vascular accidents or cerebral anomalies. Transient cerebral dysfunction postictally can alter the neurological examination and can last hours to days. Reported postictal dysfunctions are behavioural changes such as fear, aggression and confusion; blindness, usually with normal pupillary light reflexes consistent with ‘central’ blindness; menace response deficits; miosis contralateral to the lesion, due to disinhibition of the oculomotor nucleus; gait abnormalities especially ataxia and ‘conscious’ proprioceptive (paw position) deficits (Chandler and Volk 2008). The clinician should therefore repeat the neurological examination if neurological deficits are found in close proximity to the last seizure event.
Possible neurological deficits
Dogs with structural forebrain disease can have the following deficits:
▪ Changes in mentation (quality and level). Behavioural changes such as compulsive pacing (towards the lesion), head turning or head pressing and obtundation.
▪ The gait can be relatively normal, even in relatively severe structural cerebral disease.
▪ Mild hemiparesis and/or proprioceptive deficits, such as decreased or absent paw positioning, contralateral to the lesion.
▪ Menace deficits contralateral to the lesion.
▪ Decreased or absent menace response, with normal pupillary light reflex contralateral to the lesion.
▪ Hypoalgesia (facial) contralateral to the lesion.
▪ Decreased or absent response to nasal septum stimulation.
▪ Seizures themselves can be the first and only initial sign of structural brain disease, such as cerebral neoplasia in the frontal or olfactory lobe, which are brain structures that might not cause deficits recognised with the neurological examination (see the article on neoplasia on pp 24-29).
Some animals might have neurological deficits affecting both sides equally. Dogs and cats with seizures secondary to extracranial causes can have similar neurological deficits, but they are usually symmetrical in presentation. There may be waxing and waning in levels of consciousness.
Define the lesion
Finding asymmetrical neurological deficits on the interictal neurological examination can increase the likelihood of finding intracranial structural brain disease significantly by 25 times, and if symmetrical deficits are found by 10 times (Fig 1, Table 1) (Armasu and others 2012). These authors have also reported that cluster seizures are more common with structural brain disease. Taking age and breed into account can further narrow down the differential list (Table 2). Motor seizures can also affect one side of the body more than the other (asymmetrical seizures). Focal onset (usually asymmetrical) seizures are more often reported in dogs with structural brain disease, such as inflammatory/infectious causes, neoplasms, cerebral anomalies and vascular accidents. Generalised onset (symmetrical) seizures are more common with idiopathic epilepsy, metabolic, toxic and degenerative causes, and with hydrocephalus. Since metabolic and toxic disease tends to have diffuse, symmetrical effects on the brain, seizures tend to be generalised and symmetrical in onset (Chandler and Volk 2008). A normal interictal neurological examination is one of the most important predictors for the diagnosis of idiopathic epilepsy (Smith and others 2008, Armasu and others 2012). When also considering age (six months to six years) and if the breed is at a higher risk makes the diagnosis of idiopathic epilepsy highly likely. Many breeds have a predisposition towards epilepsy. A hereditary and familial basis for idiopathic epilepsy has been proposed in a number of breeds, including the golden retriever, labrador retriever, Australian, German and Belgian shepherd (Tervueren) dogs, Bernese mountain dog, beagle, Irish wolfhound, English springer spaniel, keeshond, Hungarian vizsla, standard poodle, border collie and Lagotto Romagnolo (Ekenstedt and Oberbauer 2013). The clinician should also always ask if there is a familial history of epilepsy in the animal presented with seizures.
Certain breeds have also been reported to have a predisposition, genetic susceptibility or causative mutation, such as a mutation in Epm2b causing Lafora body storage disease and progressive myoclonic epilepsy (Lohi and others 2005) described in wire-haired dachshunds, beagles and basset hounds; L2-hydroxyglutaric aciduria in Staffordshire bull terriers; neuronal ceroid lipofuscinosis in English setters, border collies, American bulldogs, dachshunds, American Staffordshire terriers, Australian shepherd dogs and Tibetan terriers (Ekenstedt and Oberbauer 2013); gliomas in boxer dogs; and inflammatory brain diseases in certain terrier breeds.
To define the lesion further the clinician also needs to take into account general historical points and physical examination findings indicating extracranial disease. Historical signs more common in seizures caused by metabolic disease include waxing and waning clinical signs, temporal relationship of clinical signs to feeding, gastrointestinal disturbances, increased or decreased appetite, pica and hypersalivation. Owners of dogs presenting with acute seizures should also be asked if exposure to a toxin is likely. Rarely the toxin can be identified and therefore a good history is pivotal. Common toxins to be considered are lead, ethylene glycol, organophosphates and metaldehyde (Chandler and Volk 2008).
After the initial characterising of the lesion and defining the location of the lesion within the system, diagnostics can be performed to further define and refine the clinical problem until you reach the most likely diagnosis (Figs 1 to 4, Table 1 to 3). Table 3 summarises initially blood and urine tests.
Abdominal ultrasound and thoracic radiographs
Body cavity imaging is not required in every patient presenting with seizures and should be seen as supporting diagnostic tests for certain clinical findings (eg, suspected hepatopathy or for staging of neoplasms). Abdominal imaging can establish the presence and type of a portosystemic shunt or other acquired liver disease. Another example for which abdominal ultrasonography is useful is to check for bilaterally enlarged adrenal glands in pituitary dependent Cushing's syndrome when seizures are caused by a macroadenoma compressing the forebrain.
Advanced imaging should be considered if a structural intracranial lesion is suspected or cannot be ruled out and when seizure control remains poor despite treatment with adequate antiepileptic drugs reaching therapeutic serum concentration. Seizures may cause transient postictal changes visible on brain MRI, most commonly symmetrical signal changes in the temporal and piriform lobes (Mellema and others 1999). Not all changes visible on MR brain images are causative for seizures and can be incidental findings, for example, certain subarachnoid diverticula (Matiasek and others 2007) or lateral ventricle asymmetry (Pivetta and others 2013). MRI in veterinary patients may also not be sensitive enough to detect subtle changes such as cortical dysplasias.
CT does not provide the same level of detail as an MRI scan, but it can be a useful technique if skull fractures (trauma) or intracranial haemorrhage is suspected.
Cerebrospinal fluid analysis
CSF analysis should be considered for any dog or cat presenting with multifocal disease on examination and/or when the clinician has the suspicion of an inflammatory/infectious disease. When taking CSF it needs to be considered that seizures themselves can cause a transient mild increase in the CSF cell count (Goncalves and others 2010). A contraindication for CSF collection is when a bleeding disorder is suspected or the animals have clinical signs suggestive of raised intracranial pressure.
EEG is the functional test to verify that the paroxysmal event is a seizure and helps to localise the origin of the epileptiform electric activity. EEG has been used in veterinary medicine from the 1950s but only a few centres use it as a diagnostic test (Chandler and Volk 2008). Dogs differ greatly in skull morphology and masticatory muscle mass, which not only make it difficult to standardise but means that they are also prone to artefacts. Artefacts can be reduced by sedating the patient, but this also alters the EEG. Indications apart from defining seizure activity is to monitor the success of antiepileptic drug treatment in status epilepticus (Serrano and others 2006).
Seizures can be caused by many pathologies, metabolic derangements and toxins. A logical decision making process taking into account history, signalment, seizure semiology, clinical findings and neurological deficits can collapse the differential list to only a few, for example, dogs with a normal physical and interictal neurological examination, generalised onset seizures, unremarkable routine blood and urine tests, and an age of seizure onset between six months and six years old have a more than 97 per cent chance of having idiopathic epilepsy (Smith and others 2008).
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.