Epidemiology of recurrent seizure disorders and epilepsy in cats under primary veterinary care in the United Kingdom

Abstract Background Little epidemiological evaluation of recurrent seizure disorders in cats currently exists in veterinary literature. Objectives To report the prevalence and risk factors for recurrent seizure disorders (RSD) and epilepsy in cats presented to primary care veterinary practices in the United Kingdom (UK). Animals A total of 285 547 cats under veterinary care during 2013 presenting to 282 primary care clinics in the UK. Methods Cohort study using multivariable logistic regression modeling for risk factor analysis. Results There were 458 confirmed RSD cases, giving a 1‐year period prevalence of 0.16% (95% confidence interval [CI], 0.15‐0.18). A subset of 114 (24.89%) cases was recorded as having epilepsy, giving a 1‐year period prevalence of 0.04% (95% CI, 0.03‐0.5). Increasing age was significantly associated with increasing odds of RSD. Breed, sex, neuter status, and body weight were not associated with RSD. Epilepsy was most frequently diagnosed in adult to middle‐aged cats. Cats aged 3.0 to <6.0 years had 3.32 times higher odds of epilepsy diagnosis compared to cats <3.0 years of age. Insured cats were more likely to be diagnosed with epilepsy compared to noninsured cats. Conclusions and Clinical Importance Although less common than in dogs, RSD and epilepsy still comprise an important disorder group in the UK cat population. Aging is a significant risk factor for these disorders in cats.

to a veterinary neurologist, with the prevalence of seizures in referral populations of cats reported as 2.1% in Germany 3 and 3.5% in Austria. 4 Cats with seizures represent an important and challenging presentation for the general veterinarian, and these conditions have the potential to cause great concern and emotional distress for owners. 5 Recurrent seizure disorders (RSD) including epilepsy have been the topic of extensive veterinary research in companion animal species over the last 75 years, but almost exclusively based on caseloads from referral populations and teaching hospitals. However, the reliability of generalization from these secondary and tertiary care subsets to the wider cat population has been questioned. 6 Research on RSD in dogs utilizing datasets from primary practice electronic patient records (EPR) have reported a 1-year period prevalence for seizure occurrence in dogs of 0.82% and for suspected idiopathic epilepsy of 0.62% to 0.75%. 7,8 However, to date, no similar studies have described RSD in cats.
Recurrent seizure disorders in cats represent a diagnostic challenge for several reasons. Seizure type in cats is often atypical. [9][10][11] Cats frequently present with focal epileptic seizures, 12,13 which can be misleading for both owners and practitioners and may be mistaken for other paroxysmal events. 14,15 Second, debate about the true prevalence and clinical relevance of spontaneous genetic epilepsy in cats is ongoing.
Traditionally, genetic epilepsy was believed to be rare in cats, 12,16 but a genetic basis since has been identified in a laboratory cat colony and now is considered an important differential diagnosis in cats. 11,17,18 Confusion about diagnosis of epilepsy in both cats and dogs is exacerbated by inconsistent use of the diagnostic terms "epilepsy" and "idiopathic epilepsy" in the veterinary literature. 16,19 The International Veterinary Epilepsy Task Force (IVETF) addressed these issues for dogs, but not for cats. 5 Fluidity of these definitions has led some authors to adopt the term "epilepsy of unknown cause" to describe RSD with a clinically unremarkable diagnostic evaluation, with the aim of maintaining a neutral stance on whether these disorders are truly "idiopathic" or if the diagnostic techniques available in veterinary medicine are currently unable to elucidate the etiology. 20 Although concise definitions that can be universally agreed on undoubtedly will aid future research efforts by use of precise terms to differentiate specific etiologies, overzealous definitions should not detract from insight that is currently available through more generalized use. 5 Defining disorders based on their phenotypic signature (eg, RSD) rather than reliance on uncertain biomedical terms (eg, epilepsy) applied by clinicians may offer increased reliability for research into disorder frequency. 21,22 Evidence relevant to the general population of animals under primary veterinary care should be derived from the general population of animals under primary veterinary care. 23 Consequently, several large projects are now underway that aim to merge anonymized clinical data from primary care veterinary clinics into single databases for research. [24][25][26][27] Research using primary care veterinary clinical records benefits from contemporaneous recording at the time of the clinical events by veterinary professionals across the spectrum of species and disorders recorded during their care. 28 We aimed to estimate the prevalence of RSD and epilepsy in the wider cat population under primary veterinary care in the United Kingdom (UK) and to evaluate demographic risk factors for their occurrence. A secondary aim was to explore risk factors associated with diagnosis of epilepsy among the subset of cats with RSD. This information could promote understanding of the clinical rationale applied by clinicians when assigning biomedical diagnostic terms to neurological clinical cases in cats.

| METHODS
The VetCompass Programme collates deidentified EPR data from primary care veterinary practices in the UK for epidemiological research. 26  Given an a priori awareness of the likely inconsistency of clinical diagnoses applied to seizure disorders in cats, 16,19 we applied a phenotypic signature approach to searching and assigning the RSD cases that were included in the analysis. 22 Case inclusion criteria for RSD required that at least 1 of the following criteria applied during 2013: at least 2 episodes of seizure events with a minimum of 24 hours between the first and final events, a final diagnosis of epilepsy or synonym (eg, epileptic) was recorded in the EPR, or was prescribed an antiseizure drug (ASD) to manage a seizure-related disorder (based on an assumption that ASD treatment was unlikely to be prescribed for a solitary seizure event). Cats with only extracranial reactive seizures (ie, seizures secondary to a primary condition external to the brain, such as hepatic encephalopathy, electrolyte imbalances, toxicity, or agonal seizuring) were excluded. The case definition for epilepsy required that a final diagnosis of epilepsy or synonym (eg, epileptic) was recorded in the EPR by the primary care veterinary team. Case finding involved initial screening of all EPR for candidate RSD cases using a bank of search terms including epil*, seiz*, seizure2, had 1 fit, had 2 fits, phenob*, epiphen2, anti-epil*, anti-convuls*, potassium bromide, KBr, levetira*, keppra, zonisam*, 2 fits, short fit, lyrica, pregabalin, and gabapentin. Candidate cases were randomized and the clinical notes of all candidates were reviewed to evaluate for case inclusion. Additional data were extracted on confirmed cases to describe the date of the first recorded seizure for the RSD overall, and whether the case was diagnosed with epilepsy by the first opinion practitioner. Confirmed RSD cases were grouped as "RSD cases" and all remaining study cats were grouped as "noncases." A "purebred" variable categorized all cats of recognizable breeds as "purebred" and the remaining cats as "crossbred." A "breed" variable included individual breeds that had at least 1 RSD case or that were represented by at least 1000 study animals, a grouped category of all remaining purebreds, and a general grouping of crossbred cats. This approach was taken to allow focus on commonly affected breeds and on common breeds, and to facilitate statistical power for the individual breed analyses. 32 A "neuter" variable described the status of the cat (neutered or intact) recorded at the final EPR. An "insurance" variable described whether a cat was insured at any point during the study period. An "age" variable categorized age (years): <3.0, 3.0-<6.0, 6.0-<9.0, 9.0-<12.0, 12.0-<15.0, ≥15.0, not recorded. Age (years) was calculated for RSD cases at the first recorded seizure event and for noncase cats at December 31st, 2013 (the latest date the cat was known to be seizure-free). This approach was taken so that the age results would reflect the odds of "becoming" a case rather than the odds of "being" a case. An "adult body weight" variable categorized adult body weight: Adult body weight described the maximum body weight recorded during the study period for cats >6 months of age. "Dominant color" defined a categorical variable that included all colors that were recorded as comprising some or all of the coloration for at least 5000 study animals along with a grouped category of all remaining less-common colors. The dominant color ascribed for individual cats was determined by the first color term used to describe the cat. "Self-color" defined a binary variable describing whether the cat was recorded as self-colored (ie, with only a single solid color) or not. 33 After data checking and cleaning in Excel (Microsoft Office Excel 2013, Microsoft Corp.), analyses were conducted using Stata Version 13 (Stata Corporation). The 1-year period prevalence describes the proportion of all study animals recorded with the disorder during a specified 12-month period, which was 2013 in the current study. 34 The 95% confidence interval (CI) estimates were derived from standard errors, based on approximation to the binomial distribution. 35 Descriptive statistics characterized the purebred status, breed, sex, color, neuter status, insurance, age and adult body weight for the RSD cases and noncases. The chi squared test was used for statistical comparison between categorical variables. 35

| Risk factors for epilepsy in the overall population of cats
These methods were repeated to evaluate risk factors associated with epilepsy in the overall population after dropping RSD cases that were not classified as epilepsy.

| Risk factors for epilepsy among cats with RSD
These methods also were applied to explore risk factors associated with diagnosis of epilepsy among the subset of cats that were recorded with RSD.

| RESULTS
The   compared with cats aged <3.0 years. Insurance was retained as a confounder in the final model (Table 6).

| DISCUSSION
Ours is the first study to explore RSD and epilepsy in cats by analyzing data from a multicenter primary care research database in the UK.
Our study of 285 547 cats attending UK primary care practices in 2013 reports a 1-year period prevalence of 0.16% for RSD and 0.04% for epilepsy as classified by primary care veterinary practitioners.
These results confirm that although RSD are less frequently diagnosed in cats than in dogs, 7,8,37 RSD are a relatively common clinical presentation for cats evaluated by veterinary practitioners, which is consistent with previous reports. 38 Despite substantial progress in recent decades on the classification of neurological diseases in companion animals 5 along with efforts to formally define internationally-accepted diagnostic lexicons and definitions, 19 reluctance remains for primary care practitioners to evaluate and identify their neurological clinical caseloads. 44 This reluctance for standardized diagnosis recording means that many primary care cases are recorded using phenotypic signature (ie, seizure disorder) rather than a formal diagnosis term (eg, epilepsy). This phenomenon was highlighted in a recent study of seizure disorders in dogs presenting to primary care clinics where only 10.7% of idiopathic epilepsy cases that met IVETF criteria were recorded as epilepsy or idiopathic epilepsy in the clinical records. 45 One impact for scientific research arising from this tendency toward informal diagnosis recording is that studies that rely on formally-recorded diagnosis terms are likely to substantially underreport the true frequency of these disorders. Consequently, and in an effort to include all true cases in our study, we chose not to rely on formal diagnosis terms such as epilepsy or idiopathic epilepsy as recorded in the clinical notes but instead to focus more on a phenotypic signature that would identify cats with RSD.
A formal diagnosis of epilepsy was recorded by the first opinion veterinary practitioner for 114 (24.89%) of the 458 RSD cases, although it is likely that many more of these cases met the internationally-agreed definition for epilepsy despite not being recorded as such by the attending veterinarians. 19 The precise clinical criteria applied for the epilepsy diagnoses recorded in the clinical records was not explored in our study, but these could be the topic of future research exploring how closely primary care clinicians apply IVETF diagnosis guidelines. 44 It is likely that a final pathophysiological rationale for the disorder was not reached in many of the cases that were not recorded as epilepsy, and therefore the attending veterinarians recorded the condition based on its phenotypic signature rather than spuriously recording a biomedical diagnosis term that they could not be confident was correct. Our study applied the diagnosis term "epilepsy" to include subclassifications including idiopathic and structural epilepsy as well as epilepsy of unknown cause. 19 Diagnosis of "epilepsy" in veterinary medicine classically requires ≥2 seizures that are 24 hours apart and cases without a known underlying cause generally are recorded as "idiopathic epilepsy." 46 As with all idiopathic conditions, idiopathic epilepsy is a diagnosis of exclusion, and therefore relies on the clinical acumen and resources available to the relevant veterinary teams. 47 In dogs aged 6 months to 6 years with normal interictal neurological examination, lack of clinically relevant abnormalities on routine blood serum biochemical and urinalysis tests with typical seizure presentation is consistent with Tier I level of diagnostic confidence. 44 A similar consensus view on diagnostic confidence for cats currently is not available. In cats, idiopathic epilepsy historically has been considered relatively rare, but in recent years more investigators have used the term routinely for cats and reported that up to 57% of cats with epilepsy might be classified as having idiopathic epilepsy. 13  Although the odds of diagnosis specifically with epilepsy were highest in cats aged 3 to 6 years, the odds of diagnosis with RSD in general increased progressively with age. Structural causes of RSD are reported in 40% to 70% of affected cats. 13 The RSD caseload in our study likely includes several underlying conditions, including neoplastic and vascular disorders that predominantly affect older animals. 9,43 Additionally, although recognized cases of reactive seizures were excluded from our study, some may have remained given that not all cases were comprehensively evaluated by complete serum biochemistry, urinalysis and blood pressure measurement. Because reactive seizures are also predominantly caused by conditions that preferentially affect older animals, 39  hyperadrenocorticism in dogs (×4.0), 54 and corneal ulceration in dogs (×1.6). 55 This insurance bias must be considered carefully when generalizing the results of studies that are based entirely on data from insured animals. 37 We did not report the proportion of animals referred for specialist management or that underwent advanced imaging, but it would have been useful to know these because a normal brain MRI and cerebrospinal fluid (CSF) analysis would increase confidence in a diagnosis of idiopathic epilepsy and insured animals are more likely to undergo this level of diagnostic evaluation.
Our findings do not support an association between breed and RSD in cats.  57 Further investigation of genetic causes of epilepsy in cats would benefit from similar development of rigorous diagnostic criteria of epilepsy of unknown cause in cats, as has been achieved in dogs. 5,19 No evidence was found of an association between sex and either RSD or epilepsy in cats in our study. This finding is consistent with previous studies in cats. 3,4 In dogs, males may have increased odds of developing seizure disorders, 7,8,58 but this observation is complicated by effects of neutering as well as different sex predisposition to seizures among specific dog breeds. 58,59 Despite the large numbers of cats affected by RSD in our study, no antiseizure medications currently are licensed for use in cats in the UK.
Several studies have shown that cats can respond well to medical treatment and furthermore that stopping treatment leads to recurrence of seizure activity in up to 75% of affected cats. 13,60 Our findings should prompt manufacturers to consider steps to undertake the licensing of antiseizure medications in cats and encourage practitioners to consider medical treatment of affected animals.
Our study had some limitations. Despite there being an estimated 1 million more owned cats than dogs in the UK, 40  perceived difficulties in medicating cats. Diagnosis of epilepsy in our study relied on the opinion of the primary care practitioner rather than the formal classification guidelines as laid out by IVETF, which were published after data collection for our study. 5,19 Only fair to moderate interobserver agreement has been found between neurologists and nonspecialists when diagnosing seizure activity in dogs and cats 15 and, as discussed above, poor concordance has been shown between primary care and retrospective classification of epilepsy cases in dogs. 45 Improvement in future studies using similar datasets would include extraction of information pertaining to ancillary testing so as to categorize animals into an IVEFT Tier 1 level of confidence of diagnosis of epilepsy. In addition to the clinical acumen of the individual veterinary practitioner and the available diagnostic tests, an accurate description of events by owners is also essential to generate the initial index of suspicion of seizure activity, 62 but such ability is likely to vary substantially among owners.

| CONCLUSION
Our large retrospective study using a multicenter primary practice dataset found a 0.16% prevalence of RSD in cats, highlighting that these disorders are not uncommon in this species. Recurrent seizure disorders increase in prevalence with increasing age. No evidence was found for sex and breed associations with RSD in cats, which may reflect multiple etiologies in this species. Epilepsy was recorded in 0.04% of cats by primary care practitioners, with cats aged 3 to 6 years being the most likely age group of cats with RSD to be diagnosed with epilepsy. Future studies using expanded datasets are needed to fully explore questions related to semiology and clinical outcomes. The veterinary neurology community should take efforts to better define the term "epilepsy" in cats and develop diagnostic criteria.