dogs starting to have recurrent seizures have a worse outcome. As an example, Labrador retrievers had a better outcome when they developed epilepsy later in life (Heynold et al., 1997). However, a recent study from the UK could not find that the onset of seizures before the age of 1 year had any influence on survival outcome (Arrol et al., 2012). Interestingly, the same group could not find that the presence of focal seizures influenced the overall outcome, which is different to what is seen in human medicine (Kwan et al., 2011).
Pseudoresistance
Once a clinician encounters pharmacoresistant epilepsy, the first question to be asked should be about the possibility of an identifiable, underlying disease process. If seizures persist because the underlying disease has not been identified and treated incorrectly, it is termed pseudoresistance. Conditions mimicking epilepsy (see Chapter 9) need to be also considered; these include cardiac associated syncope, transient vestibular disorders, movement disorders and episodic pain. Sometimes AED treatment can aggravate such mimics (Penning et al., 2009) and so each patient presenting with assumed pharmacoresistant epilepsy needs to be thoroughly investigated with an open mind. One may also encounter failure of AED therapy if owner compliance is poor or pharmacokinetic and -dynamic properties were not considered adequately.
Mechanisms of Pharmacoresistant Epilepsy
There are many possible causes of pharmacoresistant epilepsy. The mechanisms of drug resistance are likely to be variable and multi-factorial (Regesta and Tanganelli, 1999; Kwan et al., 2011). Genetic factors, such as the aforementioned polymorphisms, may be relevant and help to explain why two dogs of the same breed and with the same epilepsy characteristics differ in their response to AED therapy. Disease-related factors are also of importance, and these include the aetiology and pathophysiology underlying the seizure disorder, the clinical course of the disease, changes in drug targets or binding sites, drug uptake into the brain and changes of the epilepsy circuit. Finally, drug-related factors such as ineffective pharmacodynamics or -kinetic properties and/or development of tolerance can all play a role in the formation of pharmacoresistance.
Most pharmacoresistant patients will often not respond to multiple AEDs despite the individual differences in their mechanisms of action (Regesta and Tanganelli, 1999; Löscher and Potschka, 2002). Patients that do not respond to the first standard AED will only have a 3% chance to respond to the subsequent chosen AEDs, which needs especially to be considered during AED trials (Kwan and Brodie, 2000). This argues against epilepsy-induced changes in specific drug targets as a major cause of drug-resistant epilepsy, and makes it more likely that nonspecific mechanisms are responsible.
There are three proposed major theories for AED resistance:
1. The drug-target hypothesis: reduced drug-target sensitivity in epileptogenic brain tissue.
2. The multidrug transporter hypothesis: clearance of anti-epileptic drugs from the epileptogenic tissue through over-expression of multidrug transporters.
3. Change in the neuronal network properties.
Drug-target hypothesis
Based on the drug target hypothesis, reduced sensitivity of drug targets such as receptors or ion channels to AEDs is a key cellular mechanism that may cause drug resistance. This hypothesis is based on findings that show that in the hippocampus of human patients with pharmacoresistant temporal lobe epilepsy, the use-dependent inhibition of sodium channels by carbamazepine is lost. This finding did not extend to lamotrigine, which has a pharmacologic action similar to that of carbamazepine (Remy et al., 2003). Polymorphisms in the sodium channel encoding gene SCN2A were found in humans to be responsible for AEDs acting on the sodium channel, but also for other non-sodium channel targeted AEDs (Kwan et al., 2008). As aforementioned, Kennerly et al. (2009) showed that PB non-responders also had changes in the SCN2A gene when compared to PB responders. In addition, two other genes encoding ion channels (KCNQ3 and GABRA2) were affected (Armijo et al., 2005). In a rodent model for pharmacoresistant epilepsy it was demonstrated that there was a shift from GABAA diazepam-sensitive to GABAA diazepam-insensitive receptors in the hippocampus of PB non-responders (Volk et al., 2006). In the same model a significant loss of neurons in the CA1, CA3c/CA4 and dentate hilus of non-responders was found. This could lead to altered network properties, which could be responsible for refractoriness. In human medicine, altered network properties are well recognized in patients with hippocampal sclerosis. Hippocampal sclerosis has been associated with refractoriness to AEDs (Kwan et al., 2011). In dogs, the hippocampus is also involved in seizure propagation, which can result in MRI changes (Kuwabara et al., 2010). Such MRI abnormalities have not been associated with changes in seizure frequency or length in dogs but hippocampal changes visible on MRI are often associated with AED therapy failure in cats (Fatzer et al., 2000; Brini et al., 2004; Schmied et al., 2008).
A wealth of literature on human epilepsy cites evidence of autoantibodies to ion channels (GABAB receptors, Lancaster et al., 2010; NMDA-receptors, Dalmau et al., 2008; calcium channels, McKnight et al., 2005; voltage-gated potassium channels, McKnight et al., 2005). Interestingly, it also appears that cats with hippocampal changes develop autoanti-bodies to voltage-gated potassium channels (Pakozdy et al., 2013). These patients rarely respond to standard AED and in human medicine immunomodulatory treatment has been trialled with variable results (Vincent et al., 2010). Other proposed cellular patho-mechanisms of pharmacoresistant epilepsy that have been suggested include electrical coupling via gap junctions in neurons and glial cells (Voss et al., 2009) and mitochondrial oxidative stress and dysfunction (Waldbaum and Patel, 2010).
Multidrug transporter hypothesis
A drug can only enter the brain via two routes, either traversing the BBB or via the ventricular system and cerebrospinal fluid (CSF). The BBB restricts the entrance of any drug or other xenobiotic substance in order to protect the CNS from toxicity. The endothelial cells in the BBB are connected by tight junctions and surrounded by a basement membrane with astrocytic foot processes covering 95% of the endothelial lining. The BBB lacks transendothelial pathways such as transcellular channels or fenestrations (Löscher and Potschka, 2002). The functional significance is that the BBB resembles continuous phospholipid membranes, which stops the diffusion of hydrophilic, large or protein-bound drugs. Lipid-soluble drugs were thought on the other hand to diffuse easily through the BBB. Apart from the passive transport mechanism of lipophilic compounds, the BBB hosts a carrier-mediated transport system (Pardridge, 1999). In the last decade, multiple multidrug transporters of the ATP-binding cassette superfamily, especially P-gp and multidrug resistance-associated protein (MRP), have been shown to be expressed physiologically on the luminal side of the endothelial cells of the BBB. P-gp and MRP are also expressed in the choroid plexus epithelial cells limiting brain entrance even further (Rao et al., 1999). These transporters appear to act as an active defence mechanism against the penetration of potential CNS toxic lipophilic compounds, therefore limiting the penetration of lipophilic drugs, such as AEDs (Fromm, 2000; Spector, 2000). AEDs which are transported by multidrug transporters include valproate, gabapentin, topiramate, phenytoin, carbamazepine, phenobarbitone, felbamate and lamotrigine (Löscher and Potschka, 2002).
In contrast to the aforementioned hypotheses of drug-refractoriness, the multidrug transporter hypothesis is based on the assumption that it is not the brain target itself but the reduced AED concentration at the target site that causes pharmacoresistance. Indeed, an increased expression of multidrug transporters, such as P-gp and MRP in brain capillary
