SEARCH

SEARCH BY CITATION

Keywords:

  • Drug resistance;
  • Animal model;
  • Antiepileptic drug;
  • Epilepsy;
  • Seizure;
  • Genetic

Summary

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Dogs with spontaneous diseases can exhibit a striking similarity in etiology, clinical manifestation, and disease course when compared to human patients. Therefore, dogs are intensely discussed as a translational model of human disease. In particular, genetic studies in selected dog breeds serve as an excellent tool to identify epilepsy disease genes. In addition, canine epilepsy is discussed as a translational platform for drug testing. On one hand, epileptic dogs might serve as an interesting model by allowing the evaluation of drug efficacy and potency under clinical conditions with a focus on chronic seizures resistant to standard medication, preventive strategies, or status epilepticus. On the other hand, several limitations need to be considered including owner-based seizure monitoring, species differences in pharmacokinetics and drug interactions, as well as cost-intensiveness. The review gives an overview on the current state of knowledge regarding the etiology, clinical manifestation, pathology, and drug response of canine epilepsy, also pointing out the urgent need for further research on specific aspects. Moreover, the putative advantages, the disadvantages, and limitations of antiepileptic drug testing in canine epilepsy are critically discussed.

Given that humans and dogs share many diseases with striking similarities in clinical manifestations, studies in canine patients have been suggested to fill the gap in translating basic research and preclinical findings to human patients (Karlsson & Lindblad-Toh, 2008; Nowend et al., 2011; Rowell et al., 2011). Parallels in pathophysiologic mechanisms, disease onset, and progression as well as therapeutic responses have been reported for various human and canine diseases including epilepsy (Loscher et al., 1985; Loscher, 1997). Epidemiologic studies in epileptic dogs have revealed that canine epilepsies reflect a broad range of causes with various gene mutations suspected in idiopathic (genetic) epilepsies (Ekenstedt et al., 2012) as well as different brain insults such as trauma, infections, or tumors resulting in symptomatic (structural/metabolic) epilepsies (Pakozdy et al., 2008; Schwartz et al., 2011). Please note that the old classification (idiopathic and symptomatic) will be used throughout the review as several studies that have been based on the old nomenclature are cited.

Dog breeds with a high prevalence of epilepsy have attracted a considerable amount of attention over the last decade (Ekenstedt et al., 2012). Geneticists have pointed out that studies focusing on a selected dog breed might serve as an excellent tool for the identification of disease genes for several reasons: (1) within one dog breed genetic variation is reduced as compared to among humans (Rowell et al., 2011), (2) the dog genome is less diverged from the human genome than the mouse genome (Karlsson & Lindblad-Toh, 2008), and (3) dogs have approximately the same number of genes as humans (Karlsson & Lindblad-Toh, 2008). Therefore, identification of causative loci in dogs might help to elucidate novel genes and pathways in human epilepsies. For a comprehensive review, readers are referred to a recent article (Ekenstedt et al., 2012).

As already suggested in 1985 (Loscher et al., 1985), epileptic dogs might also serve as a translational model in antiepileptic drug development. Since then, the progress made in veterinary neurology diagnostics and the broadened clinical experience with add-on medications in canine epilepsy has rendered an improved basis for the evaluation of efficacy and potency of antiepileptic drug candidates. The fact that epilepsy seems to be difficult to treat in a relevant percentage of canine patients (see below) indicates that testing in dogs might also help to confirm whether an antiepileptic drug candidate will indeed help to overcome drug resistance. Moreover, Leppik et al. (2011) have recently pointed out that naturally occurring canine status epilepticus might serve as a translational platform for evaluating compounds for use in human trials. On the other hand, several limitations need to be considered with drug testing in dogs regarding pharmacokinetic issues, owner-based seizure monitoring, and lingering problems in diagnostic assessment.

The current review aims to (1) discuss putative promises and advantages of drug testing in canine patients; (2) critically compare the epidemiology, the clinical features, and the responsiveness to antiepileptic drugs of canine epilepsies versus human epilepsies, (3) point out the limitations and emphasize which aspects require further assessment to improve our knowledge regarding the constructive/etiologic, face, and predictive validity (Fig. 1) of canine epilepsy as a translational model.

image

Figure 1. Animals models are generally evaluated in terms of three validity criteria: (1) Constructive validity, indicating whether the cause of the disease is mimicking clinical etiology, taking into consideration that epilepsy is not induced but occurs spontaneously in canine patients we replaced the term “constructive validity” with “etiological validity.” (2) Face validity, indicating whether the clinical symptoms and the pathologic alterations mimic those in human patients. (3) Predictive validity, indicating whether the responsiveness to therapy is comparable to that in human patients. The figure gives a summary of the current state of knowledge regarding different factors and aspects of etiologic, face, and predictive validity of epilepsy in canine patients.

Download figure to PowerPoint

Need for a Translational Model?

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

The fact that drug resistance rates in humans with epilepsy have declined only minimally during the last decades despite ongoing development of novel antiepileptic drugs has restimulated the discussion about the preclinical screening and assessment of antiepileptic drug candidates (Bialer & White, 2010; Loscher & Schmidt, 2011).

Agreement exists that future antiepileptic drug development needs to cope with the challenges of drug-resistant epilepsy, which is regarded as a multifactorial problem. Considering that intrinsic severity is a major determinant of drug refractoriness (Rogawski & Johnson, 2008) and considering different disease-associated alterations in targets or pharmacokinetics (Loscher, 2005; Beck, 2007; Potschka, 2010), future drug candidates need to be selected based on their potential for disease modification and/or their potential to counteract or overcome specific resistance mechanisms. In this context a dilemma occurs in the screening decision trees for antiepileptic drug evaluation (Loscher & Schmidt, 2011). Early assessment needs to match the requirements of high-throughput screening. On the other hand, there is evidence that complex disease-associated alterations might affect the therapeutic response. These alterations are probably more intensely reflected by more elaborate chronic epilepsy models. Future developmental concepts might be based on rational mechanistic considerations regarding factors contributing to intrinsic severity and drug resistance followed by evaluation in chronic models of difficult-to-treat or drug-resistant epilepsy.

Chronic rodent epilepsy models are frequently used to assess the efficacy of antiepileptic drug candidates (Loscher & Schmidt, 2011). Although these models might have a high predictive validity (Loscher & Schmidt, 2011), they might not render valid data regarding the extent to which a compound will prove of value under clinical conditions. This is particularly true when one considers the fact that rodent models exhibit a low level of complexity regarding interindividual differences in the pathophysiologic cellular and molecular alterations and in the genetic background. Following preclinical assessment in chronic rodent models, epileptic dogs with spontaneous recurrent seizures might render a translational model, in which the long-term efficacy and tolerability of novel therapeutic approaches can be studied under clinical conditions. Considering interindividual differences, testing in dogs might also allow further development and validation of personalized therapeutic concepts. However, as described in detail below, several limitations need to be taken into account when considering studies in canine patients.

Epileptic dogs might also be considered for testing of new diagnostic devices like ambulatory electroencephalography (EEG), which is difficult to evaluate in small laboratory animal models. In a recent publication, Davis et al. (2011) described the testing of a novel implanted device to wirelessly record and analyze continuous intracranial canine EEG. The authors concluded that the striking similarity of intracranial seizure onset patterns with those in human partial onset epilepsy suggests canine epilepsy as an appropriate model for testing human antiepileptic devices and new approaches to epilepsy surgery and the new EEG technology as a versatile platform for evaluating seizures and response to therapy in the natural, ambulatory setting.

Etiology, Clinical Features, and Pathology

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Dogs are recognized as a species with a high risk to develop epileptic seizures. Canine epilepsies can be of similar origin as those in human patients. The prevalence of epileptic seizures in dogs is commonly cited to range between 0.5% and 5.7% (Chandler, 2006). It is well recognized that many dog breeds carry an increased risk for epilepsy with high breed-specific prevalence rates, for example, reaching rates of 9% or higher, sometimes even 18% in certain breeds (Casal et al., 2006; Berendt et al., 2009; Gullov et al., 2011). Therefore, the prevalence of epilepsy in purebred dogs appears even higher than the prevalence of epilepsy in people in Europe, commonly cited as four to eight cases per 1,000 (Pugliatti et al., 2007). However, in this context one needs to be cautious regarding the actual numbers, as there has been a tendency in the past to include reactive seizures, for example, due to intoxication and peripheral metabolic disturbances, in the diagnosis of canine epilepsy. This fact renders a bias to epidemiologic evaluations and probably resulted in an overestimation of the prevalence in several studies. In a recent study focusing on the etiology of juvenile-onset seizures in 136 dogs, 9 dogs (7%) were identified with reactive seizures (Arrol et al., 2012). Steinmetz et al. (2013) identified 32% reactive seizures in comparison to the proportion of 25% dogs with symptomatic epilepsy and 42% dogs with idiopathic or cryptogenic in their retrospective analysis of hospital records from dogs with seizures.

Human medicine terminology is commonly applied to categorize seizures and epilepsies in dogs. In general, seizure semiology in dogs mirrors descriptions made of people with various seizure types (Fig. 1). Furthermore, status epilepticus can occur in dogs and has also been proposed as a translational model for drug development (Leppik et al., 2011). In addition to parallels in seizure types, behavioral comorbidities have been reported in epileptic dogs, which seem to mimic psychiatric comorbidities in human patients (Shihab et al., 2011).

Regarding etiology, canine epilepsy researchers share the experience that idiopathic epilepsy (presumed genetic) might be the most common form of epilepsy in dogs younger than 5 and up to 6 years of age at the onset of epilepsy, with a prevalence up to 75% (Podell et al., 1995; Bush et al., 2002; Chandler, 2006; Thomas, 2010; Arrol et al., 2012). As in human epilepsy patients, advanced magnetic resonance brain imaging (MRI) including postcontrast T1, T2, and fluid-attenuated inversion recovery (FLAIR) sequences combined with cerebrospinal fluid analysis is frequently done in dogs to exclude structural brain disease. Imaging analyses in epileptic dogs without interictal neurologic deficits revealed significant lesions in 2.2% of dogs younger than 6 years and 26.7% of dogs older than 6 years of age (Smith et al., 2008). Another study reported lesions in 23% of the dogs with normal interictal neurologic examination (Bush et al., 2002). Among the brain insults causing symptomatic epilepsy in dogs, intracranial neoplasia and encephalitis proved to be most frequent in a study from Austria (Pakozdy et al., 2008). Dogs with histopathologically confirmed primary or secondary intracranial neoplasia proved to show a high risk for seizures, occurring in 62% of the patients (Schwartz et al., 2011). Recently, head trauma has been analyzed as another frequent cause of epilepsy, with 15.5% of dogs presenting with seizures in a hospital having a history of head trauma (Steinmetz et al., 2013).

The variance in the percentage of canine epilepsy patients with structural brain lesions might be influenced by the selection of the study population, which can always be affected by the referral conventions in the environment of a study and even more importantly by the imaging equipment and expertise at the study site. It needs to be emphasized that lack of respective diagnostic equipment probably results in an overestimation of idiopathic epilepsy as also demonstrated by Berend and Gram (1999). Therefore, it is of course critical that only clinical centers with specialized veterinary neurology units and respective equipment are involved in clinical studies. Thereby inclusion and exclusion criteria have to demand MRI as an important tool and basis for accurate classification. In a recent study analyzing the etiology of juvenile-onset seizures based on a thorough diagnostic procedure, idiopathic epilepsy was diagnosed in 75%, symptomatic epilepsy in 17%, and reactive, nonepileptic seizures in 7% (Arrol et al., 2012). In the same study, 2 of 136 dogs were classified with probable symptomatic epilepsy, reflecting an abnormal interictal neurologic examination, but normal MRI and cerebrospinal fluid analysis. The study confirms that the common practice of nonspecialized veterinarians to suspect idiopathic epilepsy in dogs with juvenile seizure onset is biased by a relevant subgroup of dogs with brain lesions as well as dogs with nonepileptic seizures.

During the last decade, relevant progress has been made in the diagnostics of canine epilepsy and the differentiation between idiopathic and symptomatic epilepsies. As already mentioned, MRI has become available in clinical centers for dogs (Wolff et al., 2012). In addition, cerebrospinal fluid analysis is used routinely to distinguish encephalitis from other central nervous system diseases (Mariscoli & Jaggy, 1997). Related to the anatomy of the dogs' skull which is, to large extents, covered by strong muscles, optimization of EEG procedures face specific challenges related to muscle artifacts, for example, the frequent need for sedation (medetomidine, propofol, acepromazine) and use of muscle relaxants, the short recording protocols, and the limited number of electrodes without routine use of temporalis electrodes (Brauer et al., 2011). Consequently, it will be unlikely that EEG in epileptic dogs will be useful to document seizure foci unless novel implantable EEG devices and continuous EEG monitoring are used (Poma et al., 2010; Davis et al., 2011).

Whereas epidemiologic data have been rather limited in the past, due to diagnostic restrictions, for symptomatic epilepsies with a structural basis, several studies have already addressed the epidemiology as well as the genetics of breed-specific idiopathic epilepsy syndromes and of progressive myoclonus epilepsy in dogs. Therefore canine idiopathic epilepsies with a specific focus on breed-specific aspects as well as canine progressive myoclonus epilepsy are discussed more intensely.

Canine idiopathic epilepsies

The existence of genetically different breed-specific idiopathic epilepsy syndromes is suggested by the fact that predominant seizure types, age of onset, occurrence of cluster seizures, and status epilepticus as well as pharmacosensitivity may vary between different dog breeds. Comparisons have been made with human epilepsy syndromes: for example, a genetically defined juvenile idiopathic epilepsy syndrome with a self-remitting clinical course in the Lagotto Romagnolo dog breed was compared to juvenile idiopathic epilepsy in human patients (Jokinen et al., 2007); an idiopathic epilepsy in Belgian shepherds was compared to familial partial epilepsy with variable foci in human patients (Berendt et al., 2009); and idiopathic epilepsy in Border Collies, Australian Shepherd, and Shetland Sheepdogs often characterized by a poor pharmacoresponse seems to resemble frontal lobe epilepsy in people (Morita et al., 2002; Hulsmeyer et al., 2010; Weissl et al., 2012).

Due to high prevalence of idiopathic epilepsy in specific dog breeds, a strong genetic background has been recognized either as a founder effect or as a variety of genetic inheritance models proposed by segregation analysis. A recent candidate gene study in four dog breeds examined 52 genes mostly encoding ion channels and neurotransmitters already known to be involved in human or murine idiopathic epilepsy, but no major associations or linkages to idiopathic epilepsy were found in the four breeds studied in detail (Ekenstedt et al., 2012). Consequently, current research attempts focus on identification of novel canine breed-specific idiopathic epilepsy loci and genes with possible relevance for the human patient population utilizing genome-wide association studies and considering the unique architecture of the dog's genome. A promising discovery is the recent identification of a truncating mutation in the canine LGI2 gene, also defined as an ortholog of the human epilepsy gene LGI1, as the cause of remitting juvenile epilepsy in Lagotto Romagnolo dogs (Seppala et al., 2011). Most recently, homozygosity for a two single nucleotide polymorphism (SNP) haplotypes within the ADAM23 gene was associated with a high risk for epilepsy in Belgian Shepherds, which is a promising finding in light of the fact that ADAM23 interacts with the known epilepsy proteins LGI1 and LGI2 (Seppala et al., 2012). Yet, slow progress in the identification of canine idiopathic epilepsy genes suggests that, like in humans, epilepsy may represent a complex genetic disease with interaction of multiple genes and environmental factors (Ekenstedt et al., 2012).

Progressive myoclonic epilepsies

The canine epileptic syndromes resulting from Lafora disease and the various types of neuronal ceroid lipofuscinoses (NCLs) were recently reclassified as canine progressive myoclonic epilepsies due to their invariable progressive nature and the underlying neurodegenerative disease (Ramachandran et al., 2009; Ekenstedt et al., 2012). Clinical signs of Lafora disease described for three dog breeds consist of intermittent fine, whole-body tremors, and intermittent myoclonic contractions of the head and neck muscles (jerks) in response to visual and acoustic stimuli. During the course of the disease, progression may occur, with generalized seizures accompanying the myoclonic seizures (Gredal et al., 2003; Lohi et al., 2005). Clinical signs and the triggering of seizures by external stimuli in affected dogs reflect similar clinical manifestations of human Lafora disease.

NCLs are described in many dog breeds (Ekenstedt et al., 2012). Generalized epileptic seizures may occur during disease progression but are an inconsistent feature not reported in all affected breeds, and often occur only in the terminal phase. In contrast to idiopathic epilepsy in dogs, significant progress has been made in the identification of the genetic base of progressive myoclonic epilepsies of dogs. A biallelic expansion of a dodecamer repeat in the EPM2B gene on CA 35 proved to be the cause of myoclonic epilepsy in miniature wirehaired Dachshunds with Lafora disease. Mutations in the orthologous human genes, that is, chromosome 6q24 EPM2A or 6p22 EPM2B (NHLRC1), are responsible for one of the severest forms of teenage-onset epilepsy (Minassian, 2001; Chan et al., 2003a,b).

At present, eight causative mutations (PPT1/CLN1, TPP1/CLN2, ATP13A2, CLN5, CLN6, CLN8, CTSD/CLN10, ARSG), all in orthologs of human NCL genes, are identified in six dog breeds affected by NCL. To this date, eight causal genes with more than 265 mutations and 38 polymorphisms have been associated with juvenile NCLs, whereas late-onset NCLs in humans still lack gene identification (Kousi et al., 2012). The recent discovery of an ARSG mutation in American Staffordshire Terriers with adult-onset NCL has resulted in the nomination of human ARSG as a potential locus of Kufs disease, which is an adult onset NCL in humans (Abitbol et al., 2010). Furthermore, a single base-pair deletion (c.1620delG) and a truncating mutation within exon 16 of the ATP13A2 gene were recently discovered in a Tibetan Terrier with late onset NCL (Farias et al., 2011; Wohlke et al., 2011).

Pathology of canine epilepsies

In human epilepsy, a multitude of cellular and molecular alterations including neurodegeneration, neurogenesis, activation of microglia, and alterations of the blood–brain barrier have been described as a consequence of seizure activity and of initial brain insults triggering epileptogenesis (Pitkanen & Lukasiuk, 2011; Loscher, 2012). Several of these disease-associated alterations are discussed as factors contributing to cognitive deficits, psychiatric comorbidities, disease development, and progression (Loscher, 2012). Unfortunately, the number of studies characterizing respective neuropathologic alterations in canine epilepsy is rather limited.

The relevance of temporal lobe pathology remains a matter of debate in canine epilepsy (Palmer, 1972; Koestner, 1989; Buckmaster et al., 2002; Kuwabara et al., 2010). This aspect also needs to be taken into account when considering clinical studies in dogs with drug-refractory epilepsy, because temporal lobe epilepsy, although the most common type of refractory epilepsy in human adults, might not play a major role in dogs.

In our own analyses, we were able to detect obvious pathomorphologic alterations in canine hippocampal tissue from dogs with both idiopathic as well as symptomatic epilepsy (unpublished data). Based on these findings we are planning a thorough evaluation of hippocampal pathology in canine epilepsies with different etiologies.

Evidence exists that further molecular and cellular alterations in the canine epileptic brain also show face validity when compared to the pathology in human epilepsies. For instance the rate of neurogenesis proved to be altered in dogs with chronic epilepsy (von Ruden et al., 2012). However, considering that data are limited and often anecdotal, further studies are obviously necessary to compare epilepsy-associated molecular and cellular alterations in more detail, taking the etiology and different types of epilepsies into consideration.

Response to Antiepileptic Drugs and Related Limitations for Drug Trials

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Chronic pharmacologic treatment is the mainstay in the clinical management of canine epilepsy. Despite major disadvantages regarding drug interactions and adverse effects, phenobarbital remains to be the first-line choice in the treatment of focal and generalized seizures in dogs (Boothe et al., 2012). This fact can be attributed to pharmacokinetic and economic considerations as well as the pharmaceutical or medicines law in some countries, which regulates that treatment needs to be based on drugs marketed for the species and indication, and that other options including drugs only marketed for use in humans can only be used upon failure of respective drugs.

Earlier reports have suggested that seizure control can be achieved in 20–40% of dogs treated with phenobarbital (Schwartz-Porsche et al., 1985; Loscher et al., 2004). However, a recent study reported a percentage of 85% of dogs with newly diagnosed epilepsy in which seizures were controlled by phenobarbital monotherapy (Boothe et al., 2012). Differences in the outcome are likely to be related to discrepancies in the study population (e.g., seizure history, breed distribution, and etiology), inclusion and exclusion criteria, dose escalation, duration of the follow-up, and so forth. In particular, it needs to be considered that Boothe et al. (2012) have excluded dogs with symptomatic epilepsy from their study.

In dogs with phenobarbital-refractory epilepsy, the oldest antiepileptic drug, potassium bromide, is often tried as an add-on medication (Kluger et al., 2009). However, the combination of both drugs also fails to result in seizure freedom in a considerable subgroup of dogs (Schwartz-Porsche & Jurgens, 1991; Hulsmeyer et al., 2010; Boothe et al., 2012; Weissl et al., 2012).

During the last decade efforts have been made to integrate newer antiepileptic drugs into the clinical management of canine epilepsy. However, selection of drugs for polytherapy needs to consider species differences regarding tolerability and pharmacokinetics. It is well known that different antiepileptic drugs can cause species-specific adverse effects in dogs. For instance, lamotrigine metabolization in dogs results in the formation of a cardiotoxic metabolite (Wong & Lhatoo, 2000), and vigabatrin can induce hemolytic anemia and neurotoxic effects in dogs (Löscher, 2003). With respect to testing of novel drug candidates, it is a major advantage that safety pharmacology and toxicology data are generally available for dogs, since respective testing needs to be done in one nonrodent species during drug development and the dog is used most frequently.

Antiepileptic drugs used so far in polytherapy in dogs include levetiracetam, zonisamide, felbamate, topiramate, gabapentin, and pregabalin as well as the new drug imepitoin (ELB131-138) (Loscher et al., 2004; Platt et al., 2006; von Klopmann et al., 2007; Volk et al., 2008; Dewey et al., 2009; Munana et al., 2012). Clinical studies evaluating the efficacy of different antiepileptic drugs in dogs with phenobarbital-resistant epilepsy indicate that resistance can for instance extend to levetiracetam (one study 36% no response; only 21% seizure-free; another study 44% no response, only 17% seizure-free), zonisamide (one study 42% no response, only 17% seizure-free; another study 20% no response, only 20% seizure-free), gabapentin (one study 59% no response, only 18% seizure-free; another study 45% no response, only 18% seizure-free), and pregabalin (22% no response, 0% seizure-free) (Dewey et al., 2004; Govendir et al., 2005; Platt et al., 2006; von Klopmann et al., 2007; Volk et al., 2008; Dewey et al., 2009; Munana et al., 2012). The data suggest that, comparable to the situation in human patients, multidrug resistance is a major issue in canine epilepsy. However, it must be considered that the studies were performed in rather small groups of epileptic dogs and that several limitations in study design exist. In particular, it is emphasized that responder and nonresponder percentages are not comparable among the studies due to major differences in the study population, inclusion and exclusion criteria, duration of the study, and so on. Moreover, the lack of a placebo group in several studies further limits conclusions as placebo responses are seen to reach high rates with a high level of variance between studies (Munana et al., 2010, 2012). The placebo rates in canine patients might be attributed to various aspects. In addition to increased attention, which might affect the patient situation and welfare during a clinical study, expectations of the owner might also contribute to the high placebo responder rates, which have reached 29% according to a recent meta-analysis of three prospective, placebo-controlled canine epilepsy trials (Munana et al., 2010). Therefore, drug testing in epileptic dogs has to face a comparable problem to human trials in which response to placebo with a range of 0–39% has shown unexplained variability and recently a strong trend toward unusually high rates (Beyenburg et al., 2010; Loscher & Schmidt, 2011). The high placebo responder rates can pose a particular challenge for novel antiepileptic drug candidates (Fig. 2).

image

Figure 2. Summary of the promises and advantages as well as the major limitations of drug testing in dogs with drug-refractory epilepsy.

Download figure to PowerPoint

In this context, one needs to address the question of whether robust and adequately powered studies including multicenter collaborative projects are generally feasible. So far only few randomized, controlled-treatment studies have been published, with the largest one including 46 patients (Boothe et al., 2012; Munana et al., 2012). Another investigation was able to recruit 111 epileptic dogs for consecutive investigations from a single neurologic referral unit (Loscher et al., 2004). Evidence provided from investigations is often limited due to lack of adequate power calculation. Considering the known placebo effect (30% of responders), sample sizes with more than 70 patients in each treatment arm are required to assess positive and negative treatment effects (20% difference in percentage responders) with sufficient statistical power (α 0.05, β 0.8), and additional patients are needed to account for high drop-out rates. Nevertheless, the feasibility of conducting such large multicenter clinical trials in epileptic dogs, provided that sufficient financial and administrative support is available, was recently demonstrated during further investigation of the partial benzodiazepine receptor agonist imepitoin from Boehringer Ingelheim Vetmedica GmbH, which led to the adoption of a positive opinion from the Committee for Medicinal Products for Veterinary Use (CVMP) for the control of epilepsy in dogs (http://www.ema.europa.eu/ema/index.jsp?curl=pages/news_and_events/news/2012/12/news_detail_001676.jsp&mid=WC0b01ac058004d5c1). The general feasibility of robust and adequately powered clinical trials is also supported by the fact that a network of well-trained veterinary neurologists (www.ecvn.org; www.acvim.org) with routine access to MRI scanning is available in the United States and Europe.

Another concern regarding studies in canine patients is related to the fact that the assessment of seizure frequency has to rely on owner-based seizure monitoring and reporting. Therefore, it is critical to consider owner compliance and the percentage of time spend with the dog during day and night as an inclusion/exclusion criterion during patient recruitment. Moreover, it needs to be considered that owners might not recognize or report all seizures even during times of close observation. A comparable problem exists in human patients. Based on an analysis of computer-assisted ambulatory EEG monitoring, Tatum et al. (2001) reported that 38% of partial seizures remained unrecognized by patients. In addition, due to the problem of seizure reporting, the duration of the study needs to be considered as a critical aspect while taking the natural course of epilepsy into account.

Moreover, it needs to be considered that dogs tend to metabolize and excrete compounds more rapidly than humans, which affects drug choice for clinical use in dogs as well as the testing of novel drug candidates in dogs (Frey & Loscher, 1985; Potschka et al., 2009). As a consequence of this fact, it is often difficult to reach steady-state and maintain therapeutic plasma concentrations with reasonable oral administration intervals. Moreover, slow-release formulations developed for humans are not appropriate for dogs considering species differences in gastric physiology. In view of the relatively short plasma elimination half-lives of many antiepileptic drugs, conclusions from several published studies are also limited due to the lack of determination of plasma trough levels. Therefore, it remains difficult to make conclusions about the rate of true drug resistance versus pseudoresistance due to failure in reaching a steady-state with continuous therapeutic concentrations. This also requires careful attention in the diagnosis of drug resistance when recruiting dogs for add-on trials, and might restrict the feasibility of canine epilepsy studies for many developmental compounds.

Another problem is related to the fact that in dogs with drug-resistant epilepsy, add-on of a test compound in general occurs in combination with phenobarbital (Boothe et al., 2012), which has a high potential for drug interactions based on the induction of cytochrome P450 (CYP) enzymes. Therefore, pilot studies are necessary to evaluate respective interactions.

As mentioned above, recruitment of patients for studies needs to be based on a thorough evaluation of patient history and diagnostics to allow an accurate diagnosis and check the match with inclusion/exclusion criteria. Altogether, diagnostics, study management, communication, repeated follow-up clinical investigations, and drug concentration analytics require a high level of commitment and compliance of the responsible veterinary neurologists and pet owners. Therefore, respective studies of add-on therapy in canine patients with drug-resistant epilepsy are highly elaborate, time-consuming, and cost-intensive and therefore can be considered only as final proof-of-principle and translational studies for novel drug candidates, which have already showed promise in chronic rodent models of difficult-to-treat or drug-resistant epilepsy. Moreover, considering drug-testing in dogs would be worthwhile only if the therapeutic responsiveness and mechanisms of drug resistance are comparable in canine and human patients (see Fig. 1 for a summary on etiologic, face, and predictive validity). Considering the multifactorial nature of drug resistance, future studies are necessary to further elucidate and compare specific factors and mechanisms that might contribute to therapeutic failure and intrinsic disease severity in epileptic dogs.

Epileptic dogs are also considered for an evaluation of compounds for treatment of human status epilepticus. Leppik et al. (2011) have recently pointed out that many compounds tested in rodent status epilepticus models had potentially significant advantages over current drugs. However, as many of these candidates were not suitable for chronic treatment they were not further developed (Leppik et al., 2011). The authors suggested that a test species exhibiting naturally occurring status epilepticus is more similar to humans as compared to chemically or electrically induced status epilepticus models in rodents, and might therefore render a translational testing approach. In addition, canine posttraumatic epilepsy has recently been suggested as a translational platform for human posttraumatic epilepsy, thereby perhaps facilitating the development of antiepileptogenic drugs (Steinmetz et al., 2013). This suggestion is based on the high incidence of posttraumatic epilepsy indicated by the retrospective study based on hospital records (Steinmetz et al., 2013). The authors state that they plan to evaluate the most effective antiepileptogenic strategy identified in rat models in dogs after traumatic brain injury.

Future Perspectives

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Canine epilepsy might be considered as a tool for the identification of novel candidate disease genes as well as drug testing. However, regarding drug testing several questions need to be addressed in future studies to provide more reliable conclusions regarding etiologic, face, and predictive validity of canine epilepsy in comparison with human epilepsy.

In particular, it is necessary to further evaluate the pharmacokinetics and drug interactions of newer antiepileptic drugs in dogs to avoid the diagnosis of pseudo-multidrug resistance and to match the International League Against Epilepsy (ILAE) definition of drug resistance (Kwan et al., 2010) as a requirement for studies with appropriate application of inclusion criteria. In this context, it will also be of significant relevance to further develop the quality of diagnostic procedures and to optimize and harmonize epilepsy classification in canine patients.

More specifically, efforts should be made to thoroughly compare the mechanisms of intrinsic disease severity, the mechanisms of drug resistance, as well as the histopathologic alterations in canine versus human epilepsy. As a basis for the design of clinical studies in dogs and the definition of inclusion/exclusion criteria, it will also be helpful to further study the effect of the mechanism of action in drug combinations in chronic rodent models during preclinical drug development.

Considering the current state-of-knowledge it remains questionable whether the funding and time investment related to canine clinical studies would be worthwhile regarding the validation of novel drug candidates.

Acknowledgments

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

The authors' own work was supported by a research grant from the Deutsche Forschungsgemeinschaft (DFG PO681/5-1), from the European Community's Seventh Framework Programme under grant agreement no. 201380 (EURIPIDES), and from the Gesellschaft für Kynologische Forschung. We thank Renee Marie Bogdanovic for language editing of the manuscript.

Disclosure

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Heidrun Potschka and her group received fees for consulting and giving presentations from Glaxo Smith Kline, UCB, and Hoffmann La Roche and funding for collaborative projects from UCB, Bial, and BioMas. Andrea Fischer's group has received funding for collaborative projects by Intervet and BioMas. The remaining authors have no conflicts of interest. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Need for a Translational Model?
  4. Etiology, Clinical Features, and Pathology
  5. Response to Antiepileptic Drugs and Related Limitations for Drug Trials
  6. Future Perspectives
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
epi12138-sup-0001-Supporting InformationS1.docxWord document25KData S1. The supporting information comprises a summary of findings from pharmacogenetic studies in dogs aiming to identify genes linked with drug resistance. Moreover, a table is presented comparing plasma elimination half-lives of selected antiepileptic drugs in dogs and humans. A list of references is given for both pharmacogenetic and pharmacokinetic data.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.