Clinical efficacy and safety of imepitoin in comparison with phenobarbital for the control of idiopathic epilepsy in dogs



The anticonvulsant activity and safety of imepitoin, a novel antiepileptic drug licensed in the European Union, were evaluated in a multicentre field efficacy study as well as in a safety study under laboratory conditions. Efficacy of imepitoin was compared with phenobarbital in 226 client-owned dogs in a blinded parallel group design. The administration of imepitoin twice daily in incremental doses of 10, 20 or 30 mg/kg demonstrated comparable efficacy to phenobarbital in controlling seizures in dogs. The frequency of adverse events including somnolence/sedation, polydipsia and increased appetite was significantly higher in the phenobarbital group. In phenobarbital-treated dogs, significantly increased levels of alkaline phosphatase, gamma-glutamyl-transferase and other liver enzymes occurred, while no such effect was observed in the imepitoin group. In a safety study under laboratory conditions, healthy beagle dogs were administered 0, 30, 90 or 150 mg/kg imepitoin twice daily for 26 weeks. A complete safety evaluation including histopathology was included in the study. A no-observed-adverse-event level of 90 mg/kg twice daily was determined. These results indicate that imepitoin is a potent and safe antiepileptic drug for dogs.


Epilepsy is a common neurological disorder in dogs characterized by spontaneous recurring seizures with or without loss of consciousness. Seizures may occur as the result of an acquired brain lesion (symptomatic or secondary epilepsy) or for unknown or genetic reasons, in which case it is referred to as idiopathic epilepsy (Mariani, 2013). Idiopathic epilepsy is a frequent inherited condition and common in a wide range of breeds (Oberbauer et al., 2003; Thomas & Dewey, 2008). Estimates of the incidence of epilepsy in dogs range between 0.5% and 1% in referral hospital populations (Löscher et al., 1985; Steinmetz et al., 2013), and idiopathic epilepsy has been estimated to represent about 60–70% of all cases of epilepsy in dogs (Patterson, 2013).

Only a limited number of antiepileptic drugs are licensed for long-term treatment of epilepsy in dogs, among them formulations of primidone and phenobarbital in North America and some European countries, respectively. Although considered efficacious, the outcome of therapy with these drugs is not always satisfactory. Thus, more than 50% of epileptic dogs do not become seizure-free, and side effects are very common (Frey & Löscher, 1985; Frey & Schwartz-Porsche, 1985; Schwartz-Porsche et al., 1985; Löscher, 1993, 1994, 2003). The side effects reported include sedation, ataxia, polyphagia, polydipsia, polyuria, elevation of hepatic enzymes and induction of drug metabolizing liver enzymes (Graham et al., 2002; Fukunaga et al., 2009). Therefore, there is a need for other medications that not only are highly efficacious but also have a safer profile to those which are currently available.

Imepitoin [1-(4-chlorophenyl)-4-(4-morholinyl)-2,5-dihydro-1H-imidazol-2-one; previously named ELB138 and AWD131-138] is a new drug substance, which has been reported to have potent anticonvulsive and anxiolytic effects (Rostock et al., 1998a,b,c,d; Rundfeldt & Löscher, 2014). The compound acts as a low affinity partial agonist at the benzodiazepine site of the GABAA receptor (Sigel et al., 1998). Imepitoin exerted significant anticonvulsant efficacy in a canine seizure model, and a clinical pilot study in epileptic dogs with spontaneously recurrent seizures demonstrated antiepileptic efficacy (Löscher et al., 2004; Rieck et al., 2006).

The purpose of the present investigations was to evaluate both the efficacy and safety of a new oral preparation of imepitoin for the treatment of idiopathic epilepsy under field conditions in dogs and the safety of up to five times the maximum recommended therapeutic dose over a period of 6 months in healthy dogs under standard laboratory conditions. These studies formed the basis for the marketing authorization of imepitoin, marketed under the trade name Pexion®, in the European Union (European Medicines Agency, 2012).

Materials and Methods

Evaluation of clinical efficacy and safety under field conditions

The study was conducted as a multicentre, randomized, blind, controlled parallel group clinical field trial with client-owned animals in compliance with good clinical practice (GCP), aimed at demonstrating noninferiority of imepitoin (100 and 400 mg tablets), (Boehringer Ingelheim Vetmedica GmbH, Ingelheim, Germany) compared with commercially available formulations of phenobarbital.

Study design

Dogs ≥5 kg bodyweight with newly diagnosed idiopathic epilepsy and with at least two generalized convulsive seizures within a documented retrospective 6-week baseline period were eligible for enrolment in the study. The diagnosis of idiopathic epilepsy was based on a clinical and neurological examination according to principles of GCP. Magnet resonance imaging (MRI) was not made mandatory for the diagnosis. Dogs were specifically excluded for enrolment if their case history included any of the following: seizure due to symptomatic epilepsy or reactive seizures; history of only partial seizures; history of status epilepticus, defined as a state of continuous seizure activity lasting for 30 min or longer or repeated seizures with failure to return to normal; previous antiepileptic treatment (with the exception of a single treatment with diazepam provided it was administered at least 12 h prior to enrolment). Pregnant or lactating bitches were not included, as were dogs with a history or clinical signs of hepatic disease (because this is a contraindication for phenobarbital). Dogs with a history or clinical signs of renal, cardiac, gastrointestinal or other disorders were excluded, if the condition in the opinion of the investigator would have exposed the dog to an unacceptable risk or compromised the evaluation of the study. Investigators in 29 clinical centres in three European countries (Germany, France and Switzerland) enrolled a total of 226 dogs in the study.

Treatment with imepitoin involved an initial dose of 10 mg/kg twice daily. It was permitted to increase the dose from 10 to 20 mg/kg twice daily and, if required, from 20 to 30 mg/kg (maximum allowed dose) twice daily if the seizures were considered to be uncontrolled at the lower dose, following strict titration rules. The comparative group received an initial dose of 2 mg/kg phenobarbital twice daily, which was permitted to be increased from 2 to 4 mg/kg twice daily and, if required, from 4 to 6 mg/kg (maximum allowed dose) twice daily if the seizures were considered to be uncontrolled at the lower dose, following the same titration rules. The products were identically packaged and labelled using secondary outer packaging to blind the investigators when allocating the study animals to the treatment groups.

The treatment period was divided into two phases: a titration phase of 8 weeks followed by the evaluation phase of 12 weeks, to enable a statistical comparison of both treatments for noninferiority. At each visit during this phase, the occurrence of seizures as reported by the owner was recorded by the investigator and counted in the seizure frequency calculation for the evaluation of efficacy. It was permitted during both phases to increase the dose of the study treatments until the maximum allowed dose was reached.

Prior to enrolment, the investigators, in collaboration with the owners, documented retrospectively the number of generalized convulsive seizures that had occurred during the six previous weeks (baseline). In the case of a serial seizures event, that is, in case of occurrence of two or more generalized seizures within 12 h, each single generalized seizure of the event was counted. After enrolment, the owners were required to keep a record of the occurrence of seizures and to present the dog to the investigator for a clinical examination at 4-week intervals. Unscheduled visits were foreseen if seizures were considered uncontrolled. To identify dogs with uncontrolled seizures, at each visit, the number of seizure events that had occurred at the current dose after an equilibration phase of 2 weeks following start of dosing or dose escalation was compared with the total number of seizures that had occurred during the baseline (N-max). The equilibration phase was required to enable phenobarbital to reach steady-state. Seizures were considered to be uncontrolled if the number of cumulated seizure events at the current dose had exceeded N-max at any time after enrolment or if at least one serial seizures event (i.e. at least two generalized seizures within 12 h) had occurred. In the event of any of these two conditions being fulfilled, the investigator was required to increase the dose of the study treatment according to the dosing schedule described above. As the retrospective baseline phase was 6 weeks, while the evaluation phase of the study was 12 weeks, exceeding N-max during the evaluation period indicates that the monthly seizure frequency in the evaluation phase was not reduced by 50%.

Safety evaluation and premature termination

During each visit, the investigator conducted a clinical examination to determine the general health status of the dog. In addition, the patient owners were asked for observations, and the occurrence of adverse events since the last visit was recorded. In addition, blood samples for the measurement of biochemical indicators of hepatic and renal function were collected from all dogs on enrolment and at the end of the study. Serum samples were prepared and submitted to a single central laboratory for the determination of concentrations of alanine aminotransferase (ALT), alkaline phosphatase (AP), aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT), glutamate dehydrogenase (GLDH), total bilirubin, bile acid and albumin as indicators of hepatic function, urea (BUN) and creatinine as indicators of renal function and total protein, sodium, potassium, calcium and phosphorus using routine standard analytical procedures. Serum concentrations of phenobarbital were monitored on a monthly basis in the case of dogs that were treated with this drug, to identify patients with serum concentrations ≥45 μg/mL, the upper permitted concentration for phenobarbital. Due to the absence of dose-limiting toxicity of imepitoin, no serum concentration determination was conducted in the imepitoin group.

Premature termination of treatment was permitted after enrolment in accordance with predetermined criteria that fell under a number of categories such as withdrawal of owner consent, occurrence of a seizure frequency greater than N-max and/or of serial seizure events despite having received the maximum treatment dose allowed, status epilepticus, concomitant use of other antiepileptic drugs, further diagnostic workup during the study that revealed symptomatic epilepsy or reactive seizures, or severe adverse reaction which made further participation impossible. Termination of treatment of dogs in the phenobarbital treatment group was planned if serum concentration of phenobarbital reached ≥45 μg/mL.

Statistical evaluation of efficacy and safety in the field study

The monthly seizure frequency (MSF) during the evaluation phase of the study for each dog of the ‘per protocol’ (PPS) population was the primary clinical efficacy parameter for noninferiority evaluation. A mean difference in one seizure per month was considered as the clinically relevant difference.

The PPS population was defined as all cases included in the study that had no relevant deviations from the study protocol and stayed in the evaluation phase for at least 6 weeks. In addition, cases removed earlier than 6 weeks after the beginning of the evaluation phase because the seizures could not be controlled at the maximum dose of study treatment and cases that had been removed earlier than 6 weeks after the beginning of the evaluation phase because of a status epilepticus occurrence were also included in the PPS population.

Monthly seizure frequency during the baseline period was compared between treatment groups using analysis of variance (anova). MSF during the evaluation period was compared using analysis of covariance (ancova) with treatment as the single main effect and the baseline MSF as a covariate. Because of the skewness of the distribution of MSF, data on MSF were log-transformed to more closely meet the assumptions of normality and homoscedasticity inherent to anova and ancova, from which geometric means were calculated. To test the noninferiority hypothesis, the difference in geometric mean MSF, as well as the 95% confidence interval (CI) for this difference, was calculated.

In addition to the PPS population, these calculations were also repeated for all animals with idiopathic epilepsy that had entered the evaluation phase, representing an extended efficacy population. As secondary endpoints, the proportion of dogs with ≥50% MSF reduction and the proportion of dogs free of generalized seizures during the evaluation phase in the PPS population were compared between the two treatment groups via Fisher's exact test.

The safety of the study treatments was assessed based on the suspected adverse drug reactions (ADRs) reported for all cases enrolled in the study (i.e. during both the titration and evaluation of efficacy phases) following classification according to the VeDDRA list of preferred terms of the system organ classes (European Agency for the Evaluation of Medicinal Products (EMEA), 2004; European Medicines Agency, 2011). The standard Z-test for equality of two independent proportions was used to compare frequencies of ADRs for individual categories.

anova was used to compare differences in blood chemistry parameters between the imepitoin and phenobarbital treatment groups prior to treatment (visit 0) and among the six treatment dose groups at the last visit. A test for linear trend for dose dependence of effects for each treatment group was based on a general contrast of means within the latter anova. In addition, the frequency of blood chemistry measurements outside the reference values for the analytical laboratory was compared using chi-squared tests at both visit 0 and the last visit.

Evaluation of safety of imepitoin under laboratory conditions

A total of 16 male and 16 female healthy beagle dogs aged approximately 5.5 months at the time of initiation of the study, housed individually in stainless steel cages, were used, following strictly USDA Animal Welfare Act (9 CFR Parts 1, 2 and 3) and the Guide for the Care and Use of Laboratory Animals (National Academy Press, Washington DC, 1996). The environmental conditions of the animal room (temperature: 18–29 °C; humidity 30–70%; lighting 12-h light/dark cycle and ventilation: at least 10 air changes per hour with 100% fresh air) were continuously monitored throughout the study. Each dog had ad libitum access to water and was offered once daily an amount of feed (Harlan Teklad Global Diet 2027) appropriate for its size and age.

The males and females were allocated randomly to four groups (four male and four female per group) and received tablets twice daily (approximately 12 h apart) under fasted conditions for 26 consecutive weeks. Representing 0× , 1× , 3×  and 5×  the maximum recommended therapeutic dose, each dose contained 0, 30, 90 or 150 mg/kg imepitoin and was administered in placebo, 100 or 400 mg imepitoin tablets (Boehringer Ingelheim Vetmedica GmbH, Ingelheim, Germany). The animals were evaluated at designated intervals during the study for changes in clinical signs, food consumption, bodyweight, ophthalmology, ECG and clinical pathology indices (haematology, serum chemistry and urinalysis). A gross and histopathological examination of organs was performed at the end of the study. The study was conducted in compliance with the OECD Principles of Good Laboratory Practice and with the Guideline on Target Animal Safety for Veterinary Pharmaceutical Products.

The study data were evaluated primarily using descriptive statistics and, where appropriate, ancova with treatment and sex as main effects and their interaction and with the corresponding pretreatment value as the covariate, followed by pair-wise comparisons of each active treatment group with the control group in the event that the treatment effect was statistically significant in the ancova.


Clinical efficacy under field conditions

A total of 226 dogs with newly diagnosed idiopathic epilepsy fulfilling the inclusion and exclusion criteria were randomly assigned to the two treatment groups and entered the titration phase. This population was analysed for safety evaluation. A total of 12 cases were enrolled in a single centre in Switzerland, 47 cases in 14 centres in France and 177 cases in 14 centres in Germany. The number of dogs recruited in each treatment group was similar in each of the three countries. The sexes were also comparably represented in each treatment group although overall, the number of males enrolled was greater (150) than for females (76). Approximately 70 different breeds were represented in the cases enrolled in the study. The breeds most frequently represented were ‘mixed breed’ (24.3%), golden retriever (9.3%), Labrador retriever (8.0%), and border collie (4.4%). The distribution of the breeds was similar in the two treatment groups. The groups were also similar with respect to age at enrolment with mean values (±standard deviation) in the imepitoin and phenobarbital groups of 4.2 (±2.9) and 3.8 (±2.4) years, respectively.

Prior to unblinding, the data for each enrolled case were checked for compliance with the eligibility criteria specified for inclusion in the evaluation of efficacy in both the PPS and the extended efficacy population. Eight cases met the exclusion criteria for other causes of seizures and were not further considered for efficacy evaluation, and 23 cases were excluded during the titration phase due to various reasons (Table 1). The remaining 195 dogs were included in the evaluation phase and hence comprised the extended efficacy population. Further, 43 dogs were excluded from efficacy evaluation in the PPS population because they were considered to be ineligible for evaluation of efficacy due to either nonadherence to the predetermined exclusion criteria or protocol deviations. The remaining 152 cases comprised the PPS population for the evaluation of the primary efficacy endpoint (Table 1).

Table 1. Distribution of cases in the imepitoin and phenobarbital treatment groups. Distribution of cases in the treatment groups and categorization of the reasons for exclusion of cases in the course of the titration and evaluation phases of the study. The extended efficacy population comprises all patients, which reached the evaluation phase, regardless of protocol violations or other criteria for exclusion. The ‘per protocol population’ (PPS) represents the primary efficacy population
Treatment groupImepitoinPhenobarbitalTotal
Number of cases enrolled in titration phase (safety population)116110226
Ineligible for efficacy evaluation due to other seizure causes808
Owner consent withdrawn (reasons unknown)505
Owner consent withdrawn due to adverse event5510
Status epilepticus202
Death (unrelated to treatment)112
Noncompliance/use of prohibited co-medication123
Serial seizures101
Total number of cases entering evaluation phase (extended efficacy population)93102195
Number of cases determined to be ineligible for the pivotal evaluation of efficacy:291443
Loss to follow-up7411
Prohibited co-medication303
Treatment scheme errors191029
Number of cases for pivotal evaluation of efficacy (PPS population)6488152

The geometric mean MSF of imepitoin and phenobarbital during baseline was nearly identical indicating successful randomization. Following treatment, the difference in geometric means between the treatment groups for the PPS and the extended efficacy population support noninferiority of imepitoin vs. phenobarbital. In particular, the difference between the imepitoin and phenobarbital groups in the geometric means for the monthly seizure frequency, adjusted via ancova for the baseline MSF, was less than 0.1 with an upper limit of the 95% confidence interval of only 0.33 for the PPS population – considerably less than the noninferiority margin of 1. Although the difference of 0.47 between the imepitoin and phenobarbital groups for the extended efficacy population was larger than that for the PPS population, the upper limit of the 95% confidence interval (0.76) was still less than the noninferiority margin of 1, indicating noninferiority also for the extended efficacy population (Table 2).

Table 2. Evaluation of noninferiority of imepitoin compared with phenobarbital. Geometric mean monthly seizure frequency of imepitoin and phenobarbital during baseline and following treatment, as well as the difference in geometric means between the treatment groups along with its 95% confidence interval (CI), for the primary efficacy population treated per protocol (PPS) and the extended efficacy population, representing all animals with idiopathic epilepsy which reached the evaluation phase
Treatment groupGeometric meana Monthly Seizure Frequency (MSF)Differencea in Geometric Mean MSF95% CI for Difference in Geometric Mean MSF
  1. a

    The standard error of the geometric mean or the difference in geometric means is shown in parentheses.

PPS population: baseline2.060 (0.135)2.208 (0.120)  
PPS population: treatment0.429 (0.093)0.332 (0.069)0.097 (0.116)−0.133–0.329
Extended efficacy population: baseline2.250 (0.129)2.216 (0.122)  
Extended efficacy population: treatment0.853 (0.130)0.388 (0.076)0.465 (0.150)0.169–0.762

For dogs in the PPS population, the per cent of dogs with a reduction of ≥50% in monthly seizure frequency was not significantly different (P = 0.308) between the imepitoin and phenobarbital treatment groups (75% [48/64] and 83% [73/88], respectively). Additionally, the per cent of dogs with complete suppression of generalized seizures in the imepitoin treatment group (46.9% [30/64] and in the phenobarbital treatment group (58.0% [51/88]) was not significantly different (P = 0.191). These results further support the finding that the efficacy of imepitoin is not inferior to that of phenobarbital in the treatment of idiopathic epilepsy.

The majority of patients of the PPS population were treated at the lowest dose, that is, no dose escalation was required to control the seizures; 64.1% of patients were administered 10 mg/kg imepitoin, while 73.9% of patients were administered 2 mg/kg phenobarbital throughout the study. Only 15.6% of patients in the imepitoin group and 9.1% of patients in the phenobarbital group reached the highest doses of 30 mg/kg and 6 mg/kg, respectively. The mean dose for all patients in the treatment groups was 15.2 mg/kg of imepitoin and 2.7 mg/kg of phenobarbital.

The number of cases that had to be excluded from the primary efficacy evaluation was unexpectedly high (see Table 1 for a complete list). Consequently, an extended efficacy population was evaluated. The enlargement of the populations resulted in an increased variability, but the efficacy of imepitoin was comparable to phenobarbital; the noninferiority margin was not exceeded.

Safety under field conditions

The safety evaluation included all 226 cases that had obtained at least one dose of active drug (Table 1). At least one adverse event was observed for 46.6% of the dogs in the imepitoin group and for 57.3% of the dogs in the phenobarbital group. Of the ADRs, almost 50% were observed in four categories: increased appetite, somnolence/sedation, polydipsia and polyuria. The remaining ADRs were distributed in 18 other categories. The frequency of increased appetite (P = 0.063), somnolence/sedation (P = 0.009), neurological disorders (P = 0.015) polydipsia (P = 0.015), polyuria (P = 0.004), renal/urinary disorders (P = 0.012) and diarrhoea (P = 0.045) was on average 50% higher in the phenobarbital treatment group reaching 175 listings, compared with the imepitoin group with 114 listings. In contrast, the frequency of hyperactivity was higher in the imepitoin group with 19 listings compared with seven listings in the phenobarbital group (P = 0.001), however, rated in most cases to be mild.

At the time of enrolment of the dogs in the study, there were no significant differences between the treatment groups with respect to the blood chemistry indicators of hepatic and renal function, and in both treatment groups, similar frequencies of animals with individual values outside the reference values of the analytical laboratory were found. However, there were notable differences between the treatment groups at the end of treatment. The liver enzymes AP, GGT, ALT and GLDH were found to be significantly (P < 0.001) increased in the phenobarbital group along with a significant (P < 0.05) trend for dose dependence, whereas neither such increase nor dose dependence trend was seen in the imepitoin group (Fig. 1a, b for AP and GGT values).

Figure 1.

Serum concentrations of alkaline phosphatase (AP, Fig. 1a) and gamma-glutamyl transferase (GGT, Fig. 1b) at inclusion and at the end of treatment with imepitoin (I) and phenobarbital (P). Displayed are median ± 1st and 3rd quartile (box) as well as the range (whiskers) for animals treated with imepitoin (open boxes) or phenobarbital (grey boxes) prior to the start of treatment at the first visit and at the last available visit during the treatment period. For animals that did not reach the end of the evaluation phase at day 140, the serum levels at the last available visit were used. The treatment groups are subdivided by the dose reached at the last visit (10, 20 or 30 mg/kg imepitoin or 2, 4 or 6 mg/kg phenobarbital, respectively) to evaluate dose dependence. The upper limit of the reference range for the respective enzyme (81 IU for AP, 6 IU for GGT) is indicated by a dashed line. The per cent of animals with values above the reference value of the respective enzyme is given for each group.

The rise in liver enzyme measurements resulted also in increased frequencies of animals with values above the reference range on the last visit in the phenobarbital groups compared with the imepitoin groups, reaching level of significance (P < 0.01) for AP, GGT and GLDH. Prior to inclusion, 9.2% of dogs from the phenobarbital group had AP values above the reference range, whereas 35.6% of the animals treated with 2 mg/kg and 81.8% of animals treated with 6 mg/kg phenobarbital had AP values above the reference range at the end of treatment. A similar dose-dependent significant increase in the frequencies of values above the reference range was seen for GGT and GLDH in dogs treated with phenobarbital, but not with imepitoin (Fig. 1b). The only noteworthy result for the indicators of renal function was the significant increase in mean concentration of creatinine at the end of treatment for dogs in the imepitoin treatment group; however, only 6.2% of the dogs in the imepitoin group had creatinine concentrations higher than the reference value at the end of treatment. Serum concentration of phenobarbital did not reach ≥45 μg/mL in any of the dogs in the phenobarbital treatment group.

Study on safety under laboratory conditions

Under laboratory conditions, imepitoin was well tolerated. The no observable effect level (NOEL) for imepitoin in this study was found to be 90 mg/kg twice daily (i.e. 3× the maximum recommended therapeutic dose). Toxicokinetic evaluation of plasma concentrations revealed dose-dependent exposure. Clinical signs of toxicity observed at the highest dose (i.e. 150 mg/kg twice daily) were mild and infrequent, being mostly CNS or gastrointestinal system related. These included relaxed nictitating membranes, vomiting, salivation and white material in the faeces, and rarely eyelid closure, lacrimation, eye dryness, eye discharges, ataxia, loss of righting reflex, intermittent tremors, decreased activity and nystagmus. In this dose group, the final bodyweight and mean cumulative bodyweight gain during the course of the study were numerically but not statistically less for both sexes in comparison with the control group.

There were no treatment-related changes in food consumption, ophthalmology or ECGs for any of the treated groups compared with the control group. Clinical biochemistry revealed an increase in creatinine levels in the both mid- and high-dose group, but these biochemical changes were not associated with changes in other renal parameters or in kidney histopathology. There were no other treatment-related changes in clinical pathology parameters for any of the groups. No statistically significant changes in urinalysis parameters or microscopic urinalysis parameters for any treated group were observed when compared to the control group or pretest values. No changes occurred upon either gross or histopathological examination.

Discussion and Conclusions

The purpose of the current studies was to demonstrate noninferiority of imepitoin to phenobarbital and to compare the safety profile of both drugs in dogs with idiopathic epilepsy. The diagnosis of idiopathic epilepsy was based on a clinical and neurological examination according to current classification in veterinary medicine (Mariani, 2013). MRI was not made mandatory for the diagnosis. The prevalence of clinically significant MRI abnormalities in dogs below the age of 6 years with seizures without interictal neurological deficits was previously found to be one of 46 dogs (Smith et al., 2008), indicating that the majority of the included dogs, aged 4.2 (±2.9) and 3.8 (±2.4) years, had indeed idiopathic epilepsy. However, it cannot be excluded that individual dogs with other undetermined causes of epilepsy had been initially included. Indeed, eight dogs were excluded from efficacy evaluation prior to unblinding due to other causes of seizures.

Phenobarbital was chosen as a positive control because it has traditionally been the primary choice as an antiepileptic agent for the use in the control of epilepsy in dogs and has therefore been considered to be the standard care. There were both ethical and practical reasons for choosing this design rather than comparison with a placebo control. Because the disease itself has a naturally occurring progression with the frequency of seizures increasing in the absence of treatment intervention (Löscher, 1994, 2003), the use of a placebo would be expected to incur an additional risk for the patients. The standard approach to compare efficacy of two drugs is a noninferiority design. The noninferiority margin of less than one seizure per month was selected because a frequency of one seizure per month was considered clinically meaningful. According to Podell (1998), an antiepileptic therapy should be initiated if seizures occur at a rate of two seizures/month; the noninferiority margin represents 50% of this seizure rate.

Treatment of epilepsy requires individual adaptation of the administered dose, to balance efficacy and tolerability (Podell, 1998). In fact, previous studies have shown that less than 50% of treated patients achieve complete seizure control and up to 30% of canine patients do not experience significant seizure frequency reduction with the most commonly used antiepileptic drugs, phenobarbital and potassium bromide (Potschka et al., 2013). Many of these patients are euthanized because of the severity of seizures or because of severe side effects from antiepileptic drugs (Trepanier et al., 1998; Berendt et al., 2007). A reduction of at least 50% in the monthly seizure frequency is therefore accepted as a measure for seizure control in dogs (Schwartz-Porsche et al., 1985; Löscher et al., 2004; von Klopmann et al., 2007). To enable individual selection of the dose, a titration regime enabling dose adjustment based on the level of seizure control achieved was introduced, allowing a two-step titration if seizures were not controlled at the initial dose. In this PPS population, primary efficacy endpoint, that is, the geometric mean MSF under treatment, was nearly identical between the phenobarbital and the imepitoin group, indicating that both drugs were similarly effective. This efficacy result is in line with previous data. In an experimental model of pentylenetetrazole-induced seizures in dogs, comparable efficacy but superior tolerability of imepitoin over phenobarbital could be seen (Löscher et al., 2013). In a pilot efficacy study in dogs with newly diagnosed epilepsy, the efficacy obtained with imepitoin was also similar to that obtained from phenobarbital-treated dogs (Rieck et al., 2006).

During the evaluation phase, most of the dogs in both treatment groups received only the lowest drug dose indicating good efficacy for the lowest dose of both treatments. The mean dose of imepitoin administered in this study was slightly lower than the mean dose of 20 mg/kg administered in a pilot study to dogs with different types of epilepsy (Rieck et al., 2006).

One caveat of the current study is that despite the enrolment of 226 dogs, the number that proceeded into the evaluation of efficacy consisted of only 152 dogs. The relatively high number of exclusions was most probably due to strict patient enrolment requirements and the complicated strict, but individual titration regime – not necessary under normal practical use – applied to fulfil the study protocol for the efficacy evaluation phase. Thus, patients were highly standardized during the study to meet the requirements for a clear statistical analysis to evaluate the efficacy of imepitoin. Patient exclusions were not triggered by the product but rather by nonadherence to the study protocol, such as titration errors or inclusion of inadequate patients. In fact, the most frequent reasons for exclusion were treatment scheme errors, loss to follow-up, and identification of other seizure causes, together accounting for 48 dogs to be excluded (Table 1). The frequency of exclusion due to these causes was higher in the imepitoin group than in the phenobarbital group. In contrast, only five dogs each in both groups were excluded because owners withdrew consent due to adverse events.

To evaluate whether the unbalanced exclusion rate had an influence on efficacy, an enlarged efficacy population was analysed, that is, an extended efficacy population comprising all patients with idiopathic epilepsy that had entered the evaluation phase. Even in this enlarged efficacy population, imepitoin was comparable to phenobarbital in efficacy, supporting the results from the primary efficacy population.

However, in the enlarged population, there was a tendency that the reduction in MSF was slightly less compared with phenobarbital. One reason for this tendency relates to the pharmacological characteristics of phenobarbital. While phenobarbital is a strong inducer of metabolic liver enzymes resulting in auto-induction and metabolic tolerance (Frey & Löscher, 1985; Löscher et al., 1985), it is also well known that repeated dosing of phenobarbital often leads to functional tolerance towards the anticonvulsant activity of phenobarbital (Schmidt et al., 1978). This functional tolerance is recognized as cause for escape from seizure control and late treatment failure in dogs (Podell, 1996). Due to the development of tolerance, phenobarbital results initially in stronger anticonvulsant effects as compared to long-term treatment. In contrast, no such development of metabolic tolerance and no functional tolerance induction have been reported for imepitoin (Löscher et al., 2004; Rundfeldt et al., 2014). This initially stronger anticonvulsant activity of phenobarbital may have contributed to an overestimation of the clinically relevant long-term anticonvulsant activity. A different cause may be found in the study design in relation to the characteristics of imepitoin. In fact, the elimination kinetics of imepitoin is rapid, and upon repeated dosing, no accumulation occurs, thereby enabling rapid titration (Rundfeldt et al., 2014). The slow titration required according to the study protocol based on the pharmacokinetics of phenobarbital prohibited an adequate rapid adaptation of the imepitoin dose in cases with insufficient seizure control. Indeed, in a pilot study, the mean imepitoin dose administered in dogs was 20 mg/kg (Rieck et al., 2006), while in this study, the dose was only 15.2 mg/kg.

As spontaneous remission in epileptic dogs is rare, lifelong treatment is necessary in most dogs with seizures induced by idiopathic epilepsy. Therefore, the safety of an antiepileptic drug is of importance. There was clear indication that imepitoin has a better safety profile than phenobarbital in clinical use, as observed in frequency and type of adverse events or in clinical blood biochemistry. The most frequently reported adverse events were somnolence/sedation, polydipsia, polyuria and increased appetite. In most reported adverse events, the incidence in the imepitoin treatment group was significantly lower than that in the phenobarbital treatment group.

Only for hyperactivity, the frequency was higher for imepitoin than for phenobarbital, however, rated in most cases to be mild. While the duration of these adverse events was not systematically assessed in this study, it is known from a previous study that the behavioural adverse effects following imepitoin were only observed at the beginning of the study but disappeared during the treatment course (Rieck et al., 2006). An increased activity of the dogs under imepitoin treatment compares favourably to somnolence/sedation observed frequently in the phenobarbital group and may be even an indication of normalization. In human, epilepsy is often associated with neuropsychiatric clinical signs, interictal depression being the most frequent (Wiglusz et al., 2012). It is known that epilepsy in dogs is frequently associated with neuropsychiatric clinical signs, such as anxiety and aggression (Shihab et al., 2011). The anxiolytic activity of imepitoin (Rundfeldt & Löscher, 2014), in addition to its anticonvulsant effect, may result in alleviation of such clinical signs resulting in the observation of increased activity of the animals. However, benzodiazepine receptor ligands are also known to induce hyperactivity in dogs (Wismer, 2002), and this spectrum of activity may have also contributed to hyperactivity, although no such hyperactivity was observed in the safety study in beagle dogs.

A major difference between imepitoin and phenobarbital was seen in blood chemistry parameters. Imepitoin did not cause any effects on liver enzymes, but for phenobarbital, a dose-dependent and consistent effect was evident. Specifically, whereas phenobarbital was associated with significantly increased levels of the liver enzymes ALT, AP, GGT and GLDH, none of these enzymes was found to have increased in the imepitoin treatment group. The effect of phenobarbital was dose dependent, and at the highest dose of 6 mg/kg, 82% of dogs reached AP values above the respective reference range. The increased serum levels of these enzymes can be indicators of subclinical liver injury, which can result in potentially fatal clinically manifested liver disease (Dayrell-Hart et al., 1991; Gaskill et al., 2005).

The excellent safety profile of imepitoin was also evident in the safety study in healthy beagle dogs. Six months of oral administration of imepitoin tablets to male and female beagle dogs at doses of 0, 30, 90 and 150 mg/kg/twice daily produced mild and infrequent toxicological effects at the highest dose (150 mg/kg/bid). The NOEL was 90 mg/kg/bid, which represents threefold the maximum recommended therapeutic dose. Even at the highest dose of 150 mg/kg bid administered, no effects on liver enzymes were found. Only an increase in creatinine levels in the both 90 mg/kg and 150 mg/kg group, but not in the 30 mg/kg group, was seen, although this biochemical change was not associated with changes in urine parameters or in kidney histopathology. A minor increase in creatinine was also seen in the clinical field study. Because only a marked damage of functioning nephrons can result in increased creatinine levels in serum (Bostom et al., 2002), and because such damage was not seen even following dosing of 150 mg/kg imepitoin for 6 months, the increased serum creatinine level may be a result of increased muscular turnover and not an indicator of kidney dysfunction.

In summary, based on the results of the clinical field study and supported by the results of the safety study in beagle dogs, the administration of imepitoin twice daily in incremental doses over the range of 10, 20 or 30 mg/kg is seen to be similarly effective as phenobarbital in controlling seizures in dogs with newly diagnosed idiopathic epilepsy and demonstrates a clinically significant superior safety profile. In view of the mild and infrequent clinical signs of toxicity observed in response to administration of 150 mg/kg twice daily to young healthy dogs over a period of 6 months and the NOEL of 90 mg/kg, the safety profile of imepitoin can be expected to be excellent in long-term clinical use at the recommended therapeutic dose.


We would like to thank Dr. Wibke Stansen, Dr. Ingo Lang and Dr. Glen Wolfrom for constructive scientific discussions and contributions. Moreover, we wish to thank all participating veterinarians and their staff members. Please check the applicable regulatory status when using imepitoin.

Conflict of Interest

Prof. Dr. Andrea Tipold has been principle investigator of the described clinical study. Dr. Thomas J. Keefe was involved in biostatistical evaluation of the data obtained in this study and serves as scientific advisor to Boehringer Ingelheim. PD Dr. habil. Chris Rundfeldt and Prof. Dr. Wolfgang Löscher are co-inventors in a medical use patent for use of imepitoin in canine epilepsy, but they do not possess any rights in this patent. Both are scientific advisors to Boehringer Ingelheim. Dr. Frerich deVries is employed by Boehringer Ingelheim, Germany.