The Evidence Behind the Diagnostic Investigation of Canine Idiopathic Epilepsy

[...]

Appraisal, application and reflection

Idiopathic epilepsy is a diagnosis of exclusion. The studies included in this summary supportthe fact that a thorough investigation of history and dog’s signalment are vital “starting points” for excluding other potential underlying causes of seizures. In all the studies the vast majority of dogs with confirmed or, at least, presumptive idiopathic epilepsy had an age onset less than 6-7 years. Armaşu et al. (2014) found that 89% of dogs with idiopathic epilepsy had an age of seizure onset <6 years. Similarly, Smith et al. (2008) reported that only 2.2% of dogs <6 years old with unremarkable inter-ictal neurological examination had significant lesion (identifiable on MRI), compared to 26.7% of dogs >6 years old.Pákozdy et al. (2008) provided a more limited scale for the age of seizures onset (<5 years). Podell et al. (1995) reported that the diagnosis of idiopathic epilepsy was more probable when the dog experienced the first seizure(s) between 1 and 5 years of age and was a large breed (>15 kg). Viitmaa et al. (2013), Kloene et al. (2008), Casal et al. (2006) and Patterson et al. (2005) found that the median age of seizure onset in their study population was 3 years. De Risio et al. (2015) combined and analyzed the data from Pákozdy et al. (2008) and Armaşu et al. (2014) and found that there was a significant association between age of onset and cause of epilepsy for dogs under 6 years of age at epileptic seizure onset (Chi-squared = 5.136, n = 431, p = 0.023) when the cut-off was set at 6 months. Dogs aged between 6 months and 6 years were significantly more likely to be affected by idiopathic than structural epilepsy in comparison to the dogs aged beyond this range.

Various breeds have been considered to be prone to idiopathic epilepsy. Multiple genes and recessive modes of inheritance have been investigated. Seppälä et al. (2012), Ekenstedt, K. et al. (2011), Kloene, J. et al. (2008), Pákozdy et al. (2008), Licht et al. (2007), Casal, M. et al. (2006), Patterson et al. (2005), Patterson et al. (2003), Kathmann et al. (1999), Jaggy et al. (1998a) and Hall et al. (1996) reported various breeds. Also, the consensus statement by Hülsmeyer et al. (2015), reviewed all the current evidence available for breeds that have been identified as being predisposed to idiopathic epilepsy with a proven or suspected genetic background. Specifically, breeds include German shepherds, Australian Shepherds, Belgian Shepherds, Bernese mountain dogs, Beagles, Border Collies, Border Terriers, Cavalier King Charles Spaniels, Dachshunds, Dalmatians, English Springer Spaniels, Finnish Spitz, Golden Retrievers, Hungarian Vizslas, Lagotto Romagnolo, Labrador Retrievers, Irish Wolfhounds, Italian Spinone, Petit Basset Griffon Vendeen, Shetland Sheepdogs, Standard Poodles and Keeshonds. Jokinen et al. (2007) reported juvenile epilepsy in Lagotto Romagnolo with mainly focal seizures and seizure onset of5 to 9 weeks. Rusbridge et al. (2004) reported that idiopathic epilepsy in Cavalier King Charles spaniels is more frequent in lines originating from whole-colour dogs. The latter characteristic was also considered to influence the development of occipital hypoplasia.

Distribution of epilepsy has been considered to be affected by gender. Most reports suggest males have an increased likelihood to develop seizures compared to females. Viitmaa et al. (2013), Jaggy and Bernadini (2008), Pákozdy et al. (2008) and Casal et al. (2006) found that males were predisposed to idiopathic epilepsy. Fredsø, N. et al. (2014) reported that neutered male dogs with idiopathic epilepsy had a significant shorter survival (median: 38.5 months) compared to intact male dogs (median: 71 months). Van Meervenne et al. (2014) also reported and that there was an over-representation of male dogs with idiopathic epilepsy but no conclusions could be drawn as far as the effect of sterilisation status in seizures is concerned. In a retrospective case series study by Van Meervenne et al. (2014), it was suggested an association between oestrus and seizures onset in intact female dogs with presumptive idiopathic epilepsy. However, Pákozdy et al. (2008) found no correlation of seizures with oestrus as well as stress or excitement. In addition, the relation between lunar cycle and seizures has been investigated by Browand-Stainback et al. (2011) and Pákozdy et al. (2008) who showed no relationship between the two.

Apart from the signalment and history, the cornerstone for diagnosing idiopathic epilepsy is a normal inter-ictal neurological examination. Prior to the neurological examination, though, a general clinical examination should be performed to detect possible signs that could be related to or even be confused with seizures. In all the studies the dogs with confirmed or presumptive idiopathic epilepsy had normal inter-ictal neurological status (only a few dogs had neurological signs but these were considered as postictal). Indeed, Armasu et al. (2007) reported that there are further risk factors, besides signalment, that increase or decrease the risk of intracranial pathology or provide one with a diagnosis of idiopathic epilepsy. Precisely, the seizure severity (e.g. cluster seizures) and abnormal neurological examination findings (which was considered one of the most important) were highly associated withstructural epilepsy. The same authors reported that 84% of dogs with idiopathic epilepsy had a normal neurological examination. Smith et al. (2008) and Pákozdy et al. (2008) also supported that unremarkable inter-ictal neurological findings in combination with the age of seizure onset are important factors for diagnosing idiopathic epilepsy. Specifically, Pákozdy et al. (2008) reported that status epilepticus, cluster seizures, partial seizures, vocalisation during seizure and impaired neurological status were more readily seen with structural epilepsy. Ghormie et al. (2015) found that in 99 dogs ≥ 5 years of age at seizure onset, an abnormal neurologic examination had 74 % sensitivity and 62 % specificity to predict structural epilepsy. Armaşu et al. (2014) found that dogs with neurological abnormalities interictally were 16.5 and 12.5 times more likely to have an asymmetrical structural cerebral lesion and a symmetrical structural cerebral lesion, respectively, rather than idiopathic epilepsy.

Magnetic resonance imaging (MRI) of the brain, clinicopathological tests, i.e. haematological, biochemistry profile and urinalysis as well as cerebrospinal fluid (CSF) analysis can be considered an important part in the diagnostic investigation of idiopathic epilepsy. De Risio et al. (2015) suggested that clinicians should perform brain MRI and CSF analysis, after exclusion of reactive seizures, in dogs with age at epileptic seizure onset <6 months or >6 years, inter-ictal neurological abnormalities as a result of intracranial lesion, status epilepticus or cluster seizure at epileptic seizure onset, or a previous presumptive diagnosis of IE and in refractory cases. The findings from these results are expected to be unremarkable and non-indicative of any known underlying cause of seizures.

In the plasma and CSF, in particular, various studies have been performed to reveal potential biomarkers that would help to identify epilepsy in dogs, either in earlier or later stages of the disease. Bartels et al. (2014) showed that chemokines (e.g. CCL19) were increased in dogs with idiopathic epileptic compared to healthy individuals; but compared to dogs with other neuro-inflammatory diseases, chemokines were markedly decreased. Hasegawa et al. (2014) showed that metabolites including glutamic acid and ascorbic acid in CSF might be useful for the diagnosis of canine epilepsy. Merbl et al. (2014) found higher CSF concetrations of tumour necrosis factor-α (TNF-α) and interleukin-6 (IL-6) in dogs with naturally occurring seizures compared to a control group of healthy dogs. Wessmann et al. (2010) found epithelial cells in 6.5% of dogs of the study population affected by idiopathic epilepsy, although it was considered as a non-specific incidental finding. Goncalves et al. (2010) reported that seizures could initially result in a mild increase oftotal nucleated cell count; thus, this fact should be considered when taking CSF straight after a seizure (false positive elevation). Podell, M. et al. (1997) reported that altered gamma-aminobutyric acid and glutamate values in CSF might be indicative of a state of chronic overexcitation in the brain of dogs with idiopathic epilepsy. Similarly, Ellenberger et al. (2004), reported that CSF concentrations of gamma-aminobutyric acid and glutamate were significantly lower in Labrador Retrievers with genetic epilepsy compared to control group dogs or in non-Labrador Retrievers with idiopathic epilepsy; the same study showed that CSF concentration of aspartate was significantly lower in all the epileptic dogs. Creevy et al. (2013) and Gesell et al. (2013) found that glutamate and endocannabinoids anandamide (AEA) concentrations, respectively, were higher in CSF of dogs with idiopathic epilepsy compared to a control group of healthy dogs. Calvo (2012) measured the C-reactive protein in the blood of dogs with idiopathic epilepsy and, contrary to dogs suffering from other causes of seizures as well as healthy dogs, detected increased concentrations within 24 hours but a decline after that period. Further CSF and/or plasma indicators that were investigated failed to contribute towards the diagnosis of idiopathic epilepsy. Specifically, Weber et al. (2012), Fuente et al. (2012), Fujiwara et al. (2008) and Lobert, V. et al. (2003) showed that CSF glucose level/glucose ratio, D-dimers, glial fibrillary acidic protein autoantibodies and pyruvate/lactate levels respectively were not useful for supporting the diagnoses of idiopathic epilepsy. In all, based on these results, researchers succeeded or failed to establish certain plasma and/or CSF biomarkers associated with seizures in epileptic dogs. However, there is still research that could be performed in the future, either for the above or new biomarkers for epilepsy.

Electroencephalogram (EEG) is regularly used as one of the diagnostic procedures in humans and its utility in dogs has been assessed in a few studies. Jaggy et al. (1998b) and Srenk et al. (1996) reported that, despite anaesthesia, interictal EEG features were consistent and unique in dogs with idiopathic epilepsy. Holliday and Williams (1998) reported that interictal EEG might be useful diagnostic technique in dogs with idiopathic epilepsy. Viitmaa et al. (2014)supported the use of fluoro-d-glucose positron emission tomography (FDG-PET) and to less extend, EEG in epileptic dogs as diagnostic tool.However, Akos et al. (2012) revealed that interictal EEG rarely showed epileptic discharges and therefore the diagnostic value of the EEG in the diagnosis of epilepsy appeared to be low. Brauer et al. (2012) found that interictal EEG was not a useful diagnostic method because it could detect epileptic activity in less than one third of all seizuring dogs (including symptomatic epilepsy) of the study population. All in all, there are quite a few challenges of using EEG routinely in animals and further work need to be performed.

In conclusion, diagnosis should be based on history, signalment (age of onset (>6months and <6years), breed, sex), normal interictal neurological examination, seizure type, unremarkable complete blood count, biochemistry profile and urinalysis in the first instance. This can be supported by excluding structural lesions with advanced brain imaging techniques (i.e. MRI) and an unremarkable CSF analysis and cytology. EEG for identification of the characteristic patterns of epileptic seizures is highly recommended as a confirmation of the diagnosis.Based on the recent consensus statement by De Risio et al. (2015), all these diagnostic features and tests were categorized based on their value in criteria for the diagnosis of idiopathic epilepsy are described in a three-tier system.Precisely, Tier I is based on signalment, history, general and neurological examination as well as minimum data base blood tests and urinalysis. Tier II is based on tier I, plus unremarkable fasting plus post-prandial bile acids as well as brain MRI and CSF analysis. Tier III is based on tier I and II, plus identification of electroencephalographic abnormalities characteristic for seizure disorders.

Implications for the future: Advance diagnostic procedures, such as MRI and EEG willbecome more widely available in order to improve the quality of diagnosis of canine epilepsy. Recently, the consensus statements by Rusbridge et al. (2015) and Matiasek et al. (2015) recommended specific MRI and diagnostic pathology protocol, respectively, for investigating idiopathic epilepsy. Lastly, further studies with a high quality design (i.e. blinded randomised controlled studies), low overall risk of bias and greater number of dogs investigating established or new diagnostic methods (e.g. CSF or serum biomarkers) for idiopathic epilepsy are needed because the current evidence in veterinary medicine is relatively weak.

Limitation of the summary: The main limitation of this summary is that we could not obtain full access to a few papers included in the summary of evidence. These included: Hasegawa, T. et al. (2014), Browand-Stainback, L. et al. (2011), Ekenstedt, K. et al. (2011), Oberbauer, A. et al. (2003), Kathmann, I. et al. (1999), Holliday, T. and Williams, D. (1998) and Hall, S. et al. (1996)

[…]