Dogs as a Natural Animal Model of Epilepsy

Abstract

Epilepsy is a common neurological disease in both humans and domestic dogs, making dogs an ideal translational model of epilepsy. In both species, epilepsy is a complex brain disease characterized by an enduring predisposition to generate spontaneous recurrent epileptic seizures. Furthermore, as in humans, status epilepticus is one of the more common neurological emergencies in dogs with epilepsy. In both species, epilepsy is not a single disease but a group of disorders characterized by a broad array of clinical signs, age of onset, and underlying causes. Brain imaging suggests that the limbic system, including the hippocampus and cingulate gyrus, is often affected in canine epilepsy, which could explain the high incidence of comorbid behavioral problems such as anxiety and cognitive alterations. Resistance to antiseizure medications is a significant problem in both canine and human epilepsy, so dogs can be used to study mechanisms of drug resistance and develop novel therapeutic strategies to benefit both species. Importantly, dogs are large enough to accommodate intracranial EEG and responsive neurostimulation devices designed for humans. Studies in epileptic dogs with such devices have reported ictal and interictal events that are remarkably similar to those occurring in human epilepsy. Continuous (24/7) EEG recordings in a select group of epileptic dogs for >1 year have provided a rich dataset of unprecedented length for studying seizure periodicities and developing new methods for seizure forecasting. The data presented in this review substantiate that canine epilepsy is an excellent translational model for several facets of epilepsy research. Furthermore, several techniques of inducing seizures in laboratory dogs are discussed as related to therapeutic advances. Importantly, the development of vagus nerve stimulation as a novel therapy for drug-resistant epilepsy in people was based on a series of studies in dogs with induced seizures. Dogs with naturally occurring or induced seizures provide excellent large-animal models to bridge the translational gap between rodents and humans in the development of novel therapies. Furthermore, because the dog is not only a preclinical species for human medicine but also a potential patient and pet, research on this species serves both veterinary and human medicine.

Keywords: antiseizure medications; canine epilepsy; intracranial EEG; pharmacokinetics; responsive neurostimulation; seizures; status epilepticus.

Figure 1

Presumed causes of recurrent epileptic seizures in dogs with epilepsy. See Steinmetz et al. (38) and text for further details.

Figure 2

Epileptogenic processes and risk factors involved in the development of epilepsy after acute brain insults. Possibly depending on crucial modifiers or risk factors, the same brain injury can be epileptogenic or not. Immediately after brain injury, early (or provoked) seizures may occur; these acute symptomatic seizures are not indicating epilepsy but may increase the risk of developing epilepsy. In the majority of patients, brain insults do not cause epilepsy. The term epileptogenesis includes processes that render the brain susceptible to spontaneous recurrent seizures and processes that intensify seizures and make them more refractory to therapy (progression or “secondary epileptogenesis”). During epileptogenesis, multiple brain alterations occur, including altered excitability of neurons and/or neuronal circuits, activation of microglia, astrocyte dysfunction, alterations in expression and function of receptors and ion channels (in part recapitulating ontogenesis), loss of neurons, neurogenesis, axonal and dendritic sprouting, gliosis, inflammatory processes, and more. It is important to note that some of these alterations may be related to post-injury repair or recovery and not suited as targets to halt the epileptogenic process. The “latent period” is the time from the initiating epileptogenic brain injury to the first onset of spontaneous clinically obvious seizures. This latent period, during which the epileptogenic processes take place, may last days to months to years. The figure has been modified from previous versions (–131).

Figure 3

Relationship between CSF GABA concentrations and neuronal excitability. In the dog studies shown in (A,B,D), CSF was withdrawn from the subarachnoidal space by a suboccipital puncture during anesthesia. (A) GABA levels in the cerebral cortex, CSF, and plasma during the administration of the antiseizure drug valproate (VPA) in an anesthetized dog. Similar experiments were performed with vigabatrin and other GABA-T inhibitors to investigate the relationship between GABA levels in the brain parenchyma and those in CSF and plasma. Unexpectedly, these experiments showed that plasma GABA alterations reflect respective alterations in the brain and CSF. Also, note the correlation between brain and CSF GABA alterations. VPA is thought to increase GABA synthesis (191), which explains the GABA increases in dogs and other species, including humans (see text). Data are from Löscher (190). (B) Correlation between CSF GABA levels and pentylenetetrazole seizure threshold in 10 healthy dogs. Data are from Löscher (192). (C) CSF GABA levels in 20 adult unmedicated healthy volunteers and 21 adult epilepsy patients. All patients had more than three seizures a day despite chronic treatment with ASMs (phenytoin, phenobarbital, or primidone). Data are shown as individual lumbar CSF GABA levels and median; the significant inter-group difference is indicated by asterisks (P = 0.0003). Data are from Wood et al. (193). (D) CSF GABA levels in 34 adult healthy control dogs and 21 adult epileptic dogs. The CSF GABA levels in the epileptic dogs are also shown separately for untreated (n = 14) and treated (n = 7) dogs, respectively. Data are shown as individual CSF GABA levels and median; the significant inter-group difference is indicated by asterisks (P = 0.0075). CSF GABA levels in treated (phenobarbital or primidone) and untreated dogs did not differ significantly. Only one of the seven treated dogs was seizure-free at the time of CSF sampling. Data are from Löscher and Schwartz-Porsche (194).

Figure 4

Advances in understanding the causes of human epilepsy. (A) Left graph: Till the ~1990's, the majority of epilepsies were characterized as “idiopathic.” Right graph: Today, epilepsy of unknown cause comprises a much smaller proportion, owing to the discovery of autoimmune epilepsies, epilepsies with lesions that are only detectable by MRI, and, most importantly, the reclassification of many epilepsies previously considered idiopathic as having a genetic cause. The exact proportions of monogenic and complex or polygenic epilepsies remain uncertain. Based on Thomas and Berkovic (42) and modified recently by Jeff Noebels. (B) Mutations identified in human epilepsy genes by gene function. Modified from Simkin and Kiskinis (37).

Figure 5

Schematic illustration of the timed i.v. pentylenetetrazol (PTZ) seizure threshold test in an unrestrained dog. Before beginning the experiments, it is important to habituate the dogs to persons involved in the experiments as well as to the rooms and to handling. For seizure threshold determination, a 3% solution of PTZ (in 0.9% NaCl) is continuously infused at a rate of 3 ml/min by an infusion pump via a thin, flexible plastic catheter of about 1 m length, connected by a sharp cut-off end of an injection needle to the cephalic vein at a hind leg. The infusion is terminated immediately after the occurrence of the first generalized twitch (initial myoclonus), which usually takes on average 120 s after the onset of PTZ infusion. Before the myoclonic twitch, dogs typically exhibit tremors as a sign of increasing neuronal excitability. During PTZ infusion, the animal is only slightly restricted (or, in trained dogs, not restricted at all). The threshold dose of PTZ (in mg/kg body weight) is calculated from the infusion rate, the bodyweight of the animal, and the time necessary to produce the first myoclonic twitch (which occurs together with the first paroxysmal EEG activity). Typical PTZ seizure thresholds are in the range of ~15 mg/kg PTZ but may vary with the breed, sex, and age of the dogs. The potency of drugs to increase seizure threshold can be determined (and compared) by calculating the doses required to increase the threshold by 20% (TID₂₀) or 50% (TID₅₀), testing a range of doses in groups of dogs (see Figure 7). The same dogs can be repeatedly used at intervals of at least 1 week to avoid kindling (see text). Dogs rapidly adapt to the method and do not show any signs of discomfort or anxiety before or after the threshold determination. Before any drug experiments, the PTZ seizure threshold is determined once per week until reproducible and stable thresholds are obtained in all dogs. The figure was modified from Löscher (269). For details see Löscher et al. (274).

Figure 6

Simplified schemes of the inhibitory GABAergic synapse and the structure of the GABA_A receptor illustrating the site of action of benzodiazepines (BZDs) and other drugs acting via this site. The left part of the figure illustrates a GABAergic synapse showing synthesis, vesicular packaging, release, uptake, and degradation of GABA in GABAergic nerve terminals; uptake into astrocytes; and the pentameric subunit structure of a typical GABA_A receptor complex in the postsynaptic membrane, consisting of α-, β, and γ-subunits. Components of the GABAergic synapse shown include glutamic acid decarboxylase (GAD), the enzyme that catalyzes the decarboxylation of glutamate to GABA; GABA-containing synaptic vesicles (circles containing GABA molecules); the GAT-1 GABA transporter (cylinders); and conversion of GABA to succinic semialdehyde (SSA) by GABA transaminase (GABA-T). The right part of the figure illustrates schematically the pentameric structure of the GABA_A receptor within the plane of the neuronal membrane showing the relative positions of the transmembrane domains. Subunit interfaces are formed by M3 and M1. The interfacial locations of the two GABA and one BZD recognitions sites are shown. By binding to the BZD site, BZDs (e.g., diazepam, midazolam, and others), β-carbolines (e.g., abecarnil), and imidazolone derivatives (e.g., imepitoin) act as positive allosteric modulators of GABA leading to increased chloride channel opening frequency, increased chloride influx, and, consequently, to increased hyperpolarization of the membrane and thus inhibition of the postsynaptic neuron. Barbiturates such as phenobarbital also bind to the GABA_A receptor to potentiate GABA, but the exact binding site is less well-established. Bromide ions (as produced by administration of potassium bromide) enhance GABAergic inhibition but the mechanism is distinct from that of BDZs and barbiturates in that Br- ions compete with Cl⁻ ions for GABA-gated Cl⁻ channels and, at high concentrations, enter the neuron through these channels, thereby inducing a lasting hyperpolarization of the neuronal membrane. Once Br⁻ ions have entered a neuron, they can only very slowly be eliminated from the neuron, explaining the poor therapeutic ratio and risk of intoxication with potassium bromide. The figure was modified from Rundfeldt and Löscher (101). For comparison, also the chemical structures of GABA, phenobarbital, a BDZ (diazepam), imepitoin, and abecarnil are shown.

Figure 7

Effect of antiseizure drugs on the timed i.v. pentylenetetrazol (PTZ) seizure threshold test in dogs. (A) Effect of different doses of phenobarbital on the PTZ seizure threshold, shown as percent increase above control threshold. Each symbol presents the percent increase in seizure threshold in a group of 6–7 dogs. By nonlinear regression analysis, the doses increasing the seizure threshold by 20% (TID₂₀) and 50% (TID₅₀) were calculated. (B) Effect of different doses of imepitoin on the PTZ seizure threshold. Other details as in (A). (C) Effect of different doses of abecarnil on the PTZ seizure threshold. Other details as in (A). (D) Alterations in antiseizure efficacy during prolonged treatment in dogs. Four drugs were compared in groups of 4–7 dogs: diazepam, clonazepam, abecarnil, and imepitoin. These drugs were administered daily over 4 weeks. The first PTZ seizure threshold was determined after 1 week of treatment. Control thresholds were repeatedly determined in each dog before the onset of treatment. Data are shown as the drug-induced mean percent increase (± SEM) above control thresholds. Note the rapid decline of antiseizure efficacy of diazepam, indicating the development of tolerance. Tolerance, although less marked, was also observed with clonazepam, but not with imepitoin or abecarnil. Data are from Frey et al. (271), Scherkl et al. (272), Löscher et al. (273), and Löscher et al. (171, 274).

Figure 8

Schematic of a dog with an implanted ambulatory NeuroVista Seizure Advisory System (SAS). The implantable device for recording and storing continuous iEEG includes: An Implantable Lead Assembly (ILA) placed in the subdural space, an Implantable Telemetry Unit (ITU), and a Personal Advisory Device (PAD). The ILA, which acquires 16 channels of iEEG, detects and relays electrical activity in the brain to the ITU. The ITU receives data from the implantable leads, predicts seizure activity using an algorithm, and sends a wireless alert to the PAD. The PAD sends a wireless alert to the caregiver, which may lead to accelerated intervention and administration of seizure-stopping medication (see text). All iEEG data are stored on a flash drive and uploaded weekly via the internet to a central data storage site. Modified from Coles et al. (354).