Efficacy of imepitoin versus phenobarbitone as first line treatment for canine idiopathic epilepsy
Max is a 3 year old male neutered 45kg long-haired German Shepherd Dog who has been brought in during evening surgery as an emergency. He was in status epilepticus when he arrived but has now stopped seizuring and is stable. You are talking to his owners, Mr and Mrs Wood, who are really worried about him. They saw him have one fit last week and think he maybe had another about 3 weeks ago. They had left him at home alone and when they got back he was agitated, had urinated in the house and was covered in saliva which is unusual. On the basis of your history, clinical and neurological examination, you suspect Max has idiopathic epilepsy. You discuss the treatment options available in your practice (Pexion or Epiphen) and Mr & Mrs Wood ask which is best at controlling seizures. You have been to some commercially sponsored talks in the last couple of years about these drugs but you have never checked the evidence base yourself. You wonder if one really is more efficacious than the other.....
3-Part Question (PICO)
In [dogs with idiopathic epilepsy] is [imepitoin compared to phenobarbitone] [better at reducing seizure frequency]?
Search Strategy and Summary of Evidence
MEDLINE(R) In-Process & Other Non-Indexed Citations and MEDLINE(R) 1946 to Present using the OVID interface
(dog.mp. OR dogs.mp. OR canine.mp. OR canines.mp. OR canis.mp. OR exp Dogs/)
(epilepsy.mp OR idiopathic epilepsy.mp OR idiopathic generalised epilepsy.mp OR IGE.mp OR epileptic.mp OR seizure.mp OR seizures.mp OR fit.mp OR fits.mp OR exp epilepsy/ OR exp seizures/)
(imepitoin.mp OR imepition.mp OR pexion.mp OR phenobarb*.mp OR epiphen.mp OR luminal.mp OR barbita.mp OR epirepress.mp OR ELB 138.mp OR ELB138.mp OR exp phenobarbital/)
CAB Abstracts 1910 to Present using the OVID interface
(dog.mp. OR dogs.mp. OR canine.mp. OR canines.mp. OR canis.mp. OR exp Dogs/ OR exp canis/)
(epilepsy.mp OR idiopathic epilepsy.mp OR idiopathic generalised epilepsy.mp OR IGE.mp OR epileptic.mp OR seizure.mp OR seizures.mp OR fit.mp OR fits.mp OR exp epilepsy/ OR exp seizures/ OR exp convulsions/)
(imepitoin.mp OR imepition.mp OR pexion.mp OR phenobarb*.mp OR epiphen.mp OR luminal.mp OR barbita.mp OR epirepress.mp OR ELB138.mp OR ELB 138.mp OR exp phenobarbital/)
- 232 papers found in MEDLINE search
- 223 papers excluded as they don't meet the PICO question
- 4 papers excluded as they are in a foreign language
- 1 papers excluded as they are review articles/in vitro research/conference proceedings
- 4 total relevant papers from MEDLINE
- 315 papers found in CAB search
- 294 papers excluded as they don't meet the PICO question
- 0 papers excluded as they are in a foreign language
- 17 papers excluded as they are review articles/in vitro research/conference proceedings
- 4 total relevant papers from CAB
Total relevant papers4 relevant papers from both MEDLINE and CAB Abstracts
Medline and CAB Abstracts searches identified a systematic review of treatment in canine epilepsy by Charalambous et al. (2014). Whilst it does include the other publications in this BET, it does not specifically answer the PICO question, therefore it is not included. The CAB Abstract search identified two copies of this review.
Two publications, Loscher et al. (2004) and Rieck et al. (2006) investigated the efficacy of imepitoin whilst it was still under its chemical name of ELB 138. It is clear from the subsequent paper by Tipold et al. (2014) that ELB 138 became imepitoin, so these two publications are therefore eligible for inclusion in this BET. It appears that the data relevant to the PICO question reported in Rieck et al. (2006) are identical to those reported in Loescher et al. (2004) where the study is described in better detail.
Two papers are therefore included in this BestBET: Loscher et al. (2004), and Tipold et al. (2014).
Summary of Evidence
Tipold et al. 2014, Germany (cases also recruited in Switzerland and France)
Clinical efficacy and safety of imepitoin in comparison with phenobarbital for the control of idiopathic epilepsy in dogs
Dogs > 5kg in weight with a diagnosis of idiopathic epilepsy based on clinical examination, neurological examination +/- MRI that had at least two seizures in the previous 6 weeks and had not received previous treatment.
Randomised controlled trial
The authors conclude that imepitoin is non-inferior to phenobarbital in reducing seizure frequency as the percentage of dogs whose monthly seizure frequency decreased by ≥50% was not significantly different between the treatment groups (75% for imepitoin dogs and 83% for phenobarbital treated dogs). Geometric mean monthly seizure frequency for dogs in the imepitoin group that fulfilled all inclusion criteria dropped from 2.060 at baseline to 0.429, whilst the geometric mean monthly seizure frequency for phenobarbital treated dogs that met inclusion criteria fell from 2.208 to 0.332 (see Table 2).
All data are reported in terms of geometric mean monthly seizure frequency; range data for the actual number of seizures experienced by dogs during the trial period would greatly aid interpretation of these data.
- The study design used does not necessarily appear to have been constructed for demonstrating non-inferiority (as stated in the methods section), but is focused on the aims stated at the end of the introduction section
- In a non-inferiority trial, a hypothesis about the expected efficacy of imeptoin, and the maximum acceptable extent of clinical non-inferiority should be stated. Neither pieces of data are included in this paper
- It is not stated that this study was ethically approved
- Neither a sample size justification nor a power calculation are provided
- It is unclear how randomisation was performed
- It is unclear whether owners were blinded to treatment groups, and it is unclear how or when unblinding was undertaken
- Dogs receiving phenobarbital needed more regular blood sampling than those receiving imepitoin so it is difficult to see how blinding was maintained during the study
- Recall bias may have affected the validity of baseline data collected
- The adverse event described as "hyperactivity" is not clearly defined
- Validated outcome measures were not used
- It is unclear which timepoints or inclusion criteria were used to calculate the monthly seizure frequencies (MSF)
- The clinical relevance of the MSF outcome measure is difficult to ascertain in the absence of range data for the number of seizures included in these calculations
- No explanation is provided for why geometric means were used on log-transformed data when reporting data relating to the MSF
- Outcomes were measured for only 12 weeks, so the study cannot draw conclusions on long-term efficacy
- A large number of dogs were withdrawn from the study for a variety of reasons; only 64/116 (55%) dogs enrolled in the imepitoin arm and 88/110 (80%) dogs enrolled in the phenobarbitol arm were included in the "per protocol" data analysis
- Only selected results are presented, and basic data are summarised only in the text
- The authors acknowledge that some dogs included may not have had idiopathic epilepsy
- It would be useful to see data for both intention to treat and per-protocol analyses to better understand the efficacy of the two treatments
- It is not clear who funded this study but the final author is an employee of Boehringer Ingelheim who hold the marketing licence for imepitoin
Loscher et al. 2004, Germany
Anticonvulsant efficacy of the low-affinity partial benzodiazepine receptor agonist ELB 138 in a dog seizure model and in epileptic dogs with spontaneously recurrent seizures
111 dogs presented to the Department of Small Animal Medicine in Hannover with spontaneously occurring tonic clonic, recurrent seizures that were diagnosed with idiopathic epilepsy. Of these, 12 dogs were treated with imepitoin monotherapy and 44 dogs in a historical group used as a control received phenobarbital monotherapy.
Unknown (standard critical appraisal utilised for quality assessment)
- The authors conclude that median seizure frequency was statistically significantly reduced versus baseline for dogs that responded to ELB138, but only if the “non-responders” (dogs whose seizure frequency remained unchanged or increased) were removed from the analysis (9 dogs responded, 3 did not).
- Median seizure frequency was reduced in dogs that received phenobarbital when both responders and non-responders were included in analysis.
- Dogs receiving phenobarbital that had generalized tonic-clonic seizures showed a significant decrease in seizure frequency, but this was not found in dogs with generalized seizures receiving ELB138.
- The main aim of this paper was not the main aim of our BET, so information relevant to this BET question is not well reported
- This is an unusual study design comparing dogs treated prospectively with ELB 138 with a retrospective group treated with phenobarbal, about which little information is given
- There is no statement about ethical approval
- There is poor description of basic data in both methods and results sections - little information are available about the dogs included or how they were selected, how dose escalation was performed or the nature of data collected in the owner diary
- No sample size justification or power calculation are included; only 12 dogs received ELB 138 so risk of Type II statistical error is high
- No validated measures were used in data collection
- There is a high risk of recall bias if owners were asked to accurately remember details about the pre-treatment seizure frequency and nature
- The baseline seizure frequency data used in the comparative efficacy analysis intentionally focused on the time period where that frequency was at its highest
- Multiple subgroup analyses are performed on very small sample sizes
- Medians are used to report seizure frequency data; as range data are not available it is not possible to determine what the clinical significance of this statistic is
- It is unclear what the baseline seizure frequency was of dogs that were free of generalised seizures in the treatment phase of the trial
- The trial was not designed to test non-inferiority, yet conclusions about non-inferiority are drawn on the basis that no statistically significant difference was found between phenobarbital and imepitoin groups
- There is no information about who funded this study
The two trials included in this BET have numerous limitations. The absence of sample size and power calculations is particularly problematic for interpreting the significance of the data presented in light of a small sample size (Loescher) and large number of drop-outs (Tipold). The study by Tipold does not appear to be best designed to determine non-inferiority (although this could reflect issues in reporting); as such their main results could be indicative of a Type II statistical error from an inadequate sample size. In both studies, the method by which baseline data were collected is unclear. If reliant on owner recall, these data may have been inaccurate, and since in both studies conclusions are based on comparison with these baseline data, this could be very problematic.
The BETs on this website are a summary of the evidence found on a topic and are not clinical guidelines. It is the responsibility of the individual veterinary surgeon to ensure appropriate decisions are made based on the specific circumstances of patients under their care, taking into account other factors such as local licensing regulations. Read small print
Charalambous M, Brodbelt D, Volk HA (2014) Treatment in canine epilepsy - a systematic review. BMC Veterinary Research 10: 257.
Loscher W, Potschka H, Rieck S, Tipold A, Rundfeldt C (2004) Anticonvulsant efficacy of the low-affinity partial benzodiazepine receptor agonist ELB 138 in a dog seizure model and in epileptic dogs with spontaneously recurrent seizures. Epilepsia 45: 1228-1239.
Rieck S, Rundfeldt C, Tipold A (2006) Anticonvulsant activity and tolerance of ELB 138 in dogs with epilepsy: a clinical pilot study. The Veterinary Journal 72: 86-95.
Tipold A, Keefe TJ, Loscher W, Rundfeldt C, de Vries F. (2014) Clinical efficacy and safety of imepitoin in comparison with phenobarbital for the control of idiopathic epilepsy in dogs. Journal of Veterinary Pharmacology and Therapeutics 38: 160-168.
About this BET
A BET is a simple method of searching for and appraising evidence around a very specific clinical situation.
BETs don’t tell you what to do, they tell you about the evidence on a certain topic.
Not a Vet?
This website has been designed to help vets use the best, most relevant, up to date science when they make decisions about their patients.