Canine and Feline Epilepsy. Luisa De Risio

      Fig. 5.4. MRI of a 17-month-old female spayed Staffordshire bull terrier with subacute progressive onset of right-sided forebrain signs. Transverse T2W (a), T2* GRE (b), T1W (c), T1WC (d) and FLAIR (e and f) images show an intraparenchymal area of signal change in the right parietal lobe. The lesion is hypointense on T2* GRE (b), has a hypointense centre on T2W (a), T1W (c) and FLAIR (e and f), peripheral hyperintensity on T2W (a) and FLAIR (e and f), and mild peripheral contrast enhancement on T1WC (d). These signal changes were suggestive of intraparenchymal haemorrhage in the right parietal lobe. FLAIR images (e and f) show marked hyperintensity (most likely marked secondary vasogenic oedema) within the corona radiata. The dog had positive faecal culture for Angiostrongylus vasorum.

      CVA can recur and relapses are most frequent in dogs where an underlying cause is identified but it is difficult to treat (Garosi et al., 2005a).

       Post-stroke seizures and epilepsy

      Seizures can occur secondary to CVAs and TIAs and can be classified as early and late depending on time of occurrence (less than 7 days or more than 7 days, respectively) following stroke (or TIA). Two or more recurrent late post-stroke seizures (PSS) are referred to as post-stroke epilepsy (PSE) (Slapo et al., 2006), although some have defined PSE also as first unprovoked seizure caused by a previous stroke (Jungehulsing et al., 2013). PSS can be focal or generalized, and status epilepticus can occur. PSS have a negative effect on outcome in patients with CVAs. PSS may exacerbate secondary cerebral injury by inducing glutamate excitotoxicity, and enhancing the mismatch between energy supply and demand under ischaemic conditions, leading to breakdown of ion gradients, mitochondrial damage, and eventually an irreversible state of injury (Menon and Shorvon, 2009). Experimental studies suggest that repeated seizure-like activity in the context of cerebral ischaemia significantly increases stroke size and can impair functional recovery. In people, the incidence of early and late PSS ranges from 2% to 16%, depending on study population, stroke subtype, follow-up duration and how the authors have defined early and late PSS and PSE (Arntz et al., 2013; Conrad et al., 2013; Jungehulsing et al., 2013). Patients with intracerebral haemorrhage have the highest incidence of PSS, followed by patients with ischaemic stroke and patients with a transient ischaemic attack (Arntz et al., 2013). The overall incidence of PSE is 2–6.4% in people with CVA (Arntz et al., 2013; Graham et al., 2013). In one study, the incidence of PSE was estimated as 1.2%, 3.5%, 9.0% and 12.4% at 3 months, 1, 5 and 10 years post-CVA, respectively (Graham et al., 2013). Stroke severity is a major risk factor for the developement of PSS and PSE (Arntz et al., 2013; Conrad et al., 2013; Graham et al., 2013; Jungehulsing et al., 2013).

      Data on incidence of PSS and PSE in veterinary medicine are limited. In one study including 33 dogs with brain infarction, two (6%) dogs with forebrain ischaemic infarcts developed recurrent generalized seizures at 10 and 31 weeks after the diagnosis of brain infarction (Garosi et al., 2005a). In another study including 27 dogs with a clinical diagnosis of cerebral ischaemic stoke (Gredal et al., 2013), seizures were reported as part of the acute symptomatology in 15 dogs (56%). Seven of these 15 dogs developed PSE. PSE has also been reported in four of five dogs (Paul et al., 2010), 1 of 16 dogs (Gonçalves et al., 2011) with ischaemic or haemorrhagic strokes affecting the prosencephalon. In addition, recurrent seizures have been reported in 20% (15/75) of dogs with an MRI diagnosis of intracranial haemorrhage (Lowrie et al., 2012).

      PSE develops more often in people with initial late PSS than in those with initial early PSS. This may be due to different pathophysiology of early and late PSS. Early PSS may be due to acute cellular biochemical disturbances either in the brain or systemically, such as altered electrolyte and acid–base balance, brain oedema, and release of excitatory neurotransmitters secondary to cerebral hypoxia or metabolic changes, whereas late PSS and PSE may result from gliotic scarring causing persistent changes in the cell networks (Slapo et al., 2006; Menon and Shorvon, 2009).

      Acute ischaemia has been shown to lead to increased extracellular concentrations of glutamate and reduced GABA-ergic function, and also to functional or structural impairment of GABA-ergic interneurons. The ischaemic penumbra of a stroke (Fig. 5.5) can contain electrically irritable tissue that provides a focus for seizure activity. The area has been shown to exhibit enhanced release of excitotoxic glutamate, ionic imbalances, breakdown of membrane phospholipids and release of free fatty acids. Epileptogenesis may result from selective neuronal cell death and apoptosis, changes in cellular membrane properties, mitochondrial changes, receptor changes (e.g. loss of GABA-ergic receptors), deafferentation and collateral sprouting (both at the site of ischaemia as well as in remote areas) and inflammatory changes. Experimental data also suggest that epileptogenesis is enhanced by hyperglycaemia at the time of ischaemia (Menon and Shorvon, 2009). Further studies are needed to clarify the pathophysiology of PSS and PSE. PSS may also occur due to recurrent strokes.

      Fig. 5.5. Illustration of the core and penumbra of an ischaemic infarct in the brain. In the core of the ischaemic infarct, hypoperfusion is severe and results in necrosis rapidly. The degree of ischaemia and subsequent cellular damage is less severe and potentially reversible in the penumbra (which is the area surrounding the core). The brain tissue within the penumbra may recover normal cellular function if perfusion is restored promptly or may become permanently damaged if ischaemia persists.

      There are no evidence-based guidelines for the treatment of PSS and PSE in people and animals. In general, early PSS and particularly status epilepticus are treated aggressively (see Chapter 24). In people, recurrent early seizures are commonly treated with AEM for 3–6 months only, whereas PSE treatment is prolonged similarly to other causes of structural epilepsy (Menon and Shorvon, 2009). The choice of the AEM is influenced by the presence of concurrent disorders (e.g. renal or hepatic dysfunction), pharmacokinetic interactions with other treatments, tolerability and potential adverse effects. To date, no AEM has been identified to be clearly superior in the treatment of PSS and PSE. In people, levetiracetam is considered both safe and effective against post-stroke seizures, and may have neuroprotective effect in brain ischaemia (Belcastro et al., 2011). The benefits of neuroprotective and prophylactic antiepileptic treatment for PSE require further investigations.

       Diagnostic investigations

      Imaging studies of the brain such as computed tomography (CT) and magnetic resonance imaging (MRI) are necessary to support the diagnosis of CVA, to differentiate between ischaemic and haemorrhagic CVA, and to determine the location and extent of the lesion. CT is very sensitive at detecting acute haemorrhage which appears hyperdense, but it may not detect acute ischaemia in the brain (Garosi, 2010). Conventional MRI can help detecting both ischaemic and haemorrhagic CVA, however differentiation between CVA and other intracranial diseases may be challenging in some cases (Cervera et al., 2011; Wolff et al., 2012). Sensitivity and specificity of routine (not including T2* gradient echo sequences and diffusion weighted images) high-field MRI (with or without provision of clinical data) in overall lesion detection and differentiation of CVAs from neoplastic and inflammatory brain disorders in dogs are 39% and 98%, respectively. Sensitivity and specificity of routine high-field MRI (with knowledge of clinical data) are 33% and 89%, respectively, in the diagnosis of haemorragic CVAs, and 67% and 100%, respectively, in the diagnosis of ischaemic CVAs (Wolff et al., 2012). MRI pulse sequences such as T2* gradient echo (for haemorrhagic CVAs), diffusion and