TYPE Original Research

PUBLISHED 15 January 2025

DOI 10.3389/fvets.2024.1507861

High-field MRI findings in

OPEN ACCESS epileptic dogs with a normal
inter-ictal neurological
EDITED BY
Paul Mandigers,
Utrecht University, Netherlands

REVIEWED BY
Yoshihiko Yu,
examination
Nippon Veterinary and Life Science University,
Japan
Fabio Stabile,
Stephanie Phillipps * and Rita Goncalves
Wear Referrals Veterinary Specialist and Small Animal Teaching Hospital, Institute of Infection, Veterinary and Ecological Sciences, University
Emergency Hospital, United Kingdom of Liverpool, Neston, United Kingdom
*CORRESPONDENCE
Stephanie Phillipps
[email protected] Introduction: Epilepsy is one of the most common chronic neurological
RECEIVED 08 October 2024 conditions affecting dogs. Previous research exploring the likelihood of a
ACCEPTED 30 December 2024 structural cause of epilepsy specifically in dogs with a normal inter-ictal
PUBLISHED 15 January 2025
examination is limited to a small population of dogs using low-field MRI. The
CITATION aims of this study were to establish high-field (1.0T and 1.5T) MRI findings in
Phillipps S and Goncalves R (2025) High-field
MRI findings in epileptic dogs with a normal
dogs presenting with epileptic seizures and a normal inter-ictal examination.
inter-ictal neurological examination. Methods: Medical records were retrospectively searched for dogs presenting
Front. Vet. Sci. 11:1507861.
doi: 10.3389/fvets.2024.1507861
with at least two epileptic seizure events more than 24 h apart. To be included
in the study, patients had to have a normal neurological examination, high-field
COPYRIGHT
© 2025 Phillipps and Goncalves. This is an MRI of the brain and have had metabolic and toxic causes excluded.
open-access article distributed under the Results: Four hundred and twelve dogs were eligible for inclusion. Crossbreeds
terms of the Creative Commons Attribution
License (CC BY). The use, distribution or were most commonly affected (n = 63, 15.3%) followed by Border collies
reproduction in other forums is permitted, (n = 39, 9.5%) and Labrador retrievers (n = 26, 6.3%). Seventy-six dogs (18.5%)
provided the original author(s) and the had abnormalities detected on MRI, 60 (78.9%) of which were considered to
copyright owner(s) are credited and that the
original publication in this journal is cited, in be incidental. Overall, 16 dogs (3.9%) had a structural cause of their epileptic
accordance with accepted academic seizures including neoplasia (n = 13, 81.3%), anomalous (n = 2, 12.5%) and
practice. No use, distribution or reproduction meningoencephalitis of unknown origin (MUO) (n = 1, 6.3%). When split into age
is permitted which does not comply with
these terms. group at first epileptic seizure structural lesions were documented in 0/66 dogs
aged <1 year, 4/256 (1.6%) dogs aged ≥1 year ≤6 years (three neoplastic and
one anomalous), 3/51 (5.9%) aged >6 years ≤8 years (two neoplastic and one
MUO), and 9/39 (23.1%) dogs aged >8 years (eight neoplastic, one anomalous).
Multivariate analysis identified two risk factors for structural disease: increasing
age at first epileptic seizure (p < 0.001, OR = 4.390, CI 2.338–8.072) and a
history of status epilepticus (p = 0.049, OR = 4.389, CI 1.010–19.078).
Discussion: Structural lesions are an uncommon cause of epilepsy at any age in
dogs with a normal inter-ictal examination.

KEYWORDS

dog, epilepsy, structural, normal, MRI, inter-ictal

Introduction
The worldwide lifetime prevalence of active epilepsy (defined as ongoing seizures or
seizures requiring ongoing treatment) in humans is estimated to be 0.76%, with around 50
million people worldwide affected (1). In dogs, the incidence of at least one epileptic seizure
event is estimated to be 0.82% (2), with epilepsy of unknown origin diagnosed in approximately
0.62% of dogs attending UK primary care practices (3). With an estimated 11 million dogs
owned in the UK alone (4), this equates to over 90,000 dogs experiencing a single epileptic

Frontiers in Veterinary Science 01 frontiersin.org

Phillipps and Goncalves 10.3389/fvets.2024.1507861

seizure in a one-year period and 68,000 diagnosed with epilepsy of 407 dogs diagnosed with idiopathic epilepsy presenting to a referral
unknown origin. Unsurprisingly then, epileptic seizure disorders are institution is 41% and status epilepticus in 10% (17). In a recent study
the most common presenting neurological complaint in dogs (5), and investigating risk factors relating to outcome of patients with cluster
will be encountered regularly by most small animal clinicians. seizures or status epilepticus, 41/93 (44.1%) were diagnosed with
Epileptic seizures may be secondary to a number of different disease idiopathic epilepsy (18) and 50/124 (40.3%) in another focussing on
processes, including metabolic or toxic aetiologies (termed ‘reactive status epilepticus alone (19). These findings are fairly consistent with
seizures’), structural, and idiopathic (genetic, suspected genetic and previous literature documenting idiopathic epilepsy as the cause of
unknown) causes (6). Current recommendations made by the 40.3% of cluster seizures and between 13.2–37.5% of status epilepticus
International Veterinary Epilepsy Task Force (IVETF) are that dogs (7, 20). None of these studies assessed risk in only patients with a
aged below 6 months or above 6 years of age at the onset of their first normal neurological examination that were yet to be diagnosed with
epileptic seizure should have magnetic resonance imaging (MRI) of the the cause of their epileptic seizures, and so it remains unknown
brain performed (6). The presence of an abnormal inter-ictal whether the presence of cluster seizures or status epilepticus increases
neurological examination has been demonstrated to be strongly the risk of structural disease in this patient population.
associated with an increased risk of structural disease as the cause of Performing a brain MRI results in significant financial cost for pet
epileptic seizures in dogs (7), however IVETF imaging recommendations owners and also requires general anaesthesia or profound sedation,
relating to age are regardless of the outcome of the inter-ictal which are not risk-free procedures (21). Additionally, there may
neurological examination. These guidelines were based on previous be limited access to this imaging modality depending on geographic
studies which suggested statistically significant increase in risk of a location as reported in human medicine (22). With the cost of living
structural lesion outside of these age limits (8). As the growing body of crisis and the increasing demand from pet owners that veterinarians
research into canine epilepsy continues to expand however, the scientific offer rational diagnostic tests which are specific to that patient, rather
rigor on which these recommendations were made has been called into than recommending so called ‘gold standard’ workup in every case,
question. Some of the studies cited in reaching this age cut off did not there is a need for up-to-date, accurate information on the diagnostic
differentiate between dogs with normal or abnormal inter-ictal utility of brain MRI in epileptic patients with a normal
neurological examinations (8, 9) or did not require reliable neurological neurological examination.
examinations to have been performed in all patients at all (10). Other The aims of this study were therefore to (1) report the high-field
studies also included neurologically abnormal dogs in the ‘neurologically MRI findings in a large group of epileptic dogs with a normal inter-
normal’ population, such as mentation changes and cervical ictal neurological examination, (2) establish the prevalence of
hyperaesthesia (11) or post-ictal changes and abnormalities likely structural disease within this patient population and any association
attributed to anti-seizure drugs (ASDs) (12) included as neurologically with signalment or seizure type.
normal. Although the overall reported incidence of structural disease
in some of the papers used in the recommendations is high (10, 13), if
dogs with abnormal neurological examinations are excluded then Materials and methods
structural disease varies from 5.3% (14) to 31% (13) rather than the 26
and 59% reported in the IVETF guidelines (6). Additionally, many of Study design was retrospective. Digital medical records from
these studies used small patient populations, which are even smaller the University of Liverpool Small Animal Teaching Hospital were
when only neurologically normal animals are included. searched for dogs presenting for investigation of epileptic seizures
More recently, a large study of 900 dogs undergoing MRI at multiple between December 2008 and May 2024. Inclusion criteria required
institutions to investigate the cause of recurrent seizure disorders found that medical records were complete for review (clinical history,
statistically significant association between the diagnosis of idiopathic treatment while hospitalized, signalment), that dogs had
epilepsy and the patient age group 6 months to 6 years (15). Again, experienced more than one seizure event (>24 h apart), had
patients did not have to be neurologically normal in between epileptic undergone a complete neurological examination by a diplomate of
seizures in order to be included in the study and 70.9% of dogs aged the European College of Veterinary Neurology (ECVN) and/or
over 6 years had structural lesions found on MRI, but only 3.4% of dogs ECVN resident, as well as MRI of the brain and having had
less than 6 months of age. With the increasing availability of high-field metabolic and toxic causes of seizures excluded through clinical
MRI, it has been postulated that more subtle lesions may be found due history taking and routine bloodwork (haematology, biochemistry,
to the improved spatial and image resolution compared to low-field bile acid stimulation test as had been deemed necessary by the
scans (16). Additionally, earlier lesions may be discovered, particularly clinician responsible for the case). Dogs were excluded if a full
given how the IVETF recommendations may affect which animals are neurological examination could not be performed, if there were any
recommended to have an MRI scan performed. abnormalities found on neurological examination (including
Recommendations by the IVETF are also to perform MRI scans bilaterally symmetrical signs that may have been attributable to
in all patients that experience cluster seizures or status epilepticus (6). post-ictal changes or side effects of commonly prescribed ASDs), if
Previously reported prevalence of cluster seizures in a population of it was unclear from the history whether epileptic seizures or other
paroxysmal events were responsible for the clinical signs, if the
clinical history provided by the owner was consistent with
Abbreviations: ASD, Antiseizure drug; CSF, Cerebrospinal fluid; ECVN, European neurological abnormalities other than epileptic seizure activity and
College of Veterinary Neurologists; IVETF, International veterinary epilepsy task if medical data or MRI were not available for review. The study was
force; MRI, Magnetic resonance imaging; RBC, Red blood cell; TNCC, Total approved by the University of Liverpool Veterinary Research Ethics
nucleated cell count. Committee (reference number VREC752).

Frontiers in Veterinary Science 02 frontiersin.org

Phillipps and Goncalves 10.3389/fvets.2024.1507861

Information on signalment, clinical history, neurological on MRI. Before multivariable analysis, all variables were assessed for
examination findings, clinicopathological findings, MRI findings, correlation using Spearman’s rank correlation coefficients. If
treatment administered, outcome and follow-up information were Spearman’s rank correlation coefficient was >0.8, the most statistically
collected from the records. Follow-up information was recorded at significant or biologically plausible variable was selected. Independent
re-examinations. Breeds were recorded both individually and variables demonstrating liberal association (p < 0.2) were taken
purebreds were also grouped by skull conformation type. For statistical forward for multivariate modelling in which statistical significance
analysis, breeds with more than 10 animals represented were analysed was set at p < 0.05.
separately and all other dogs (including crossbreeds) were grouped as
‘other’. Dogs were grouped by age at the time of their MRI as follows:
<6 months, ≥6 months <1 year, ≥1 year ≤6 years, >6 years ≤8 years Results
and > 8 years. The same age grouping was also performed for age at
onset of seizures. Based on the owner’s, referring veterinarian’s Four hundred and sixteen dogs met the inclusion criteria. Four
description or hospital records, seizure types were grouped into dogs were excluded from analysis because the results of their MRI
generalized epileptic seizures only, focal epileptic seizures only, or could not definitively be categorised as either causative of epileptic
both generalised and focal epileptic seizures. Any cases which seizures or an incidental finding (two dogs with chronic lacunar
experienced focal epileptic seizures with secondary genersalisation infarcts, one dog with a small meningeal nodule which was stable at
were included in the latter group. Video records were utilised where follow-up MRI 7 months later and one dog with leukoaraiosis). Four
available by the attending veterinarian to characterise epileptic seizure hundred and twelve dogs were therefore included in the study.
type. Presence of cluster seizures (>1 seizure in a 24 h period) and Crossbreeds were most commonly affected (63/412, 15.3%), and the
status epilepticus (a single event lasting >5 min or two events without Border collie (39/412, 9.5%), Labrador retriever (26/4132, 6.3%) and
regaining full consciousness in between) were also recorded. Cavalier King Charles spaniels (19/412, 4.6%) were the most
Cerebrospinal fluid (CSF) analysis was evaluated where available but commonly reported purebreds (Table 1). Ninety-eight dogs (23.8.0%)
was not a prerequisite for inclusion in the study. were brachycephalic, 214 (51.9%) were mesocephalic and 39 (9.5%)
Magnetic resonance images were acquired from December 2008 were dolichocephalic. One hundred and thirty-one (31.8%) dogs were
to July 2015 with a 1.0T MRI scanner (Magnetom 1.0T; Siemens, female (85/131, 64.9% neutered) and 281 (68.2%) were male (175/281,
Forchheim, Germany) and from August 2015 to May 2024 with a 1.5T 62.3% neutered).
scanner (Ingenia 1.5TCX; Philips Medical Systems, Eindhoven, the Median age at diagnosis was 51 months (IQR 25–76 months,
Netherlands). As the 1.0T scanner is a super conducting magnet all range 2–187). Age group split at diagnosis was <6 months 15 dogs
images were considered high field (23). Sequences performed included (3.6%), ≥6 months <1y 26 dogs (6.3%), ≥1 year ≤6 years 254 dogs
a minimum of T2-weighted images (T2W) (in transverse and sagittal (61.7%), >6 years ≤8 years 66 dogs (16.0%), > 8 years 51 dogs (12.4%).
planes at least), fluid attenuation inversion recovery (FLAIR) and pre- The median age at first epileptic seizure was 40 months (IQR
and post-contrast (intravenous injection of 0.1 mmol/kg of 19–70, range 1–187). Age group split at first epileptic seizure was
gadopentetate dimeglubine) T1-weighted images (T1W) in the <6 months 25 dogs (6.1%), ≥6 months <1y 41 dogs (10.0%), ≥1 year
transverse plane. Images were reviewed by a board-certified diagnostic ≤6 years 256 dogs (62.1%), >6 years ≤8 years 51 dogs (12.4%),
imager and a board certified and/or a residency trained neurologist. >8 years 39 dogs (9.5%).
A diagnosis of peri-ictal change was made based on the presence of Two hundred and thirty-six (57.3%) of dogs had MRI performed
bilaterally symmetrical T2W/FLAIR hyperintensity of the with the 1.5T magnet and 176 (42.7%) with the 1T magnet. Seventy-six
hippocampus, piriform lobes +/− cingulate gyri (24). Cerebrospinal
fluid collection was performed under the same general anaesthesia
and analysis comprised a red blood cell (RBC) count and total TABLE 1 Most frequent breeds of dog with epilepsy and a normal inter-
ictal neurological examination.
nucleated cell concentration (TNCC), protein concentration
measurement and a cytological examination with a differential cell Breed Number (%) No (%) with
count. Normal CSF TNCC was defined as less than 5 cells/μl and structural disease
normal CSF total protein as less than 0.45 g/L (25). Crossbreed 63 (15.3) 2 (3.2)
Descriptive statistics were performed using Excel (Microsoft Border collie 39 (9.5) 0 (0)
Corportation, Redmond, Washington, USA 2020).
Labrador retriever 26 (6.3) 0 (0)
Statistical analysis was performed using the software SPSS 27.0
(SPSS Inc., Chicago, Illinois, USA). Continuous data were assessed for Cavalier King Charles spaniel 19 (4.6) 0 (0)

normality using the Shapiro–Wilk test. For normally distributed data, Staffordshire bull terrier 18 (4.4) 4 (22.2)
the range and mean values are presented and in abnormally distributed Cocker spaniel 14 (3.4) 0 (0)
sets, the range, median value and interquartile range (IQR) are
French bulldog 14 (3.4) 1 (7.1)
provided. Excel data was then imported into SPSS for
Pug 13 (3.2) 0 (0)
statistical analysis.
Binary logistic regression modelling was utilised to evaluate the Border terrier 12 (2.9) 4 (33.3)
univariable associations between risk factors (age group at first English springer spaniel 12 (2.9) 0 (0)
epileptic seizure, length of time between epileptic seizure onset and Chihuahua 11 (2.7) 0 (0)
MRI, breed, sex, neuter status, generalised vs. focal seizures, presence
Other purebreed 171 (41.5) 6 (3.5)
of cluster seizures and status epilepticus) and structural abnormalities

Frontiers in Veterinary Science 03 frontiersin.org

Phillipps and Goncalves 10.3389/fvets.2024.1507861

(18.4%) of the 412 dogs had abnormalities detected on MRI, 60 for a longer duration—one for 6 months and one for 20 months. The
(78.9%) of which were considered to be incidental. Incidental findings patient with the history of epileptic seizures of over 1 year experienced a
included Chiari-like malformation (n = 30, 50%), otitis (n = 20, single epileptic seizure 20 months prior to diagnosis, but no further
33.3%) (15 otitis media, five otitis externa), syringohydromyelia epileptic seizures until the 3 months prior to diagnosis.
(n = 13, 21.7%) dilation of the ventricular system (n = 6, 10%), peri- Two hundred and eighty-three (68.7%) dogs had generalised
ictal change (n = 5, 8.3%), mild cortical atrophy (n = 4, 6.7%), epileptic seizures only, 68 (16.5%) had focal epileptic seizures only and
atlantoaxial band (n = 3, 5%), absent septum pellucidum (n = 3, 5%) 61 (14.8%) dogs had both generalised and focal epileptic seizures
mild supracollicular fluid accumulation (n = 2, 3.3%) and one each recorded. One hundred and seventy-four (42.1%) dogs had a history
(1.7%) of: caudal calvarial malformation, ear mass (not affecting the of cluster seizures, including 164/396 (41.4%) dogs with no structural
para aural tissues), vascular malformation in the fourth ventricle cause of epileptic seizures and 10/16 (62.5%) dogs with structural
(extra-parenchymal and not resulting in hydrocephalus), dental disease. Thirty-three dogs had experienced status epilepticus including
disease, atlantoaxial band, pterygoid muscle lesion and temporal 30/396 (7.6%) dogs with no structural cause of their epileptic seizures
muscle lesion. and 4/16 (25%) dogs with structural disease.
Sixteen (3.8%) dogs had clinically significant structural changes on Cerebrospinal fluid was sampled in 274 (66.5%) dogs and was
MRI: 13 (81.3%) were suspected neoplastic (nine suspected gliomas, abnormal in 20 (7.3%) of sampled dogs, though blood contamination
two meningiomas, one suspected histiocytic sarcoma and one calvarial significant to preclude meaningful analysis was present in seven (35%) of
mass), two (12.5%) anomalous conditions (one cortical dysplasia and these cases, leaving 13 animals with abnormal CSF results. Blood
the other porencephaly) and one (6.3%) inflammatory disease contamination was attributed as the cause of abnormal CSF findings in
(meningoencephalitis of unknown origin). Breeds affected by structural 2/13 (15.4%) of these dogs (RBC 2920 with a TNCC of 8 cells/μl in one
lesions included four Staffordshire bull terriers (all gliomas), four dog with a glioma and RBC 9280 with a TNCC of 7 in a dog diagnosed
Border terriers (all gliomas), two crossbreeds (one suspected histiocytic with idiopathic epilepsy). In the remaining 11 dogs, TNCC was increased
sarcoma, one porencephaly), one Old English sheepdog (cortical in nine (81.8%) cases (median 8 cells/μl, range 6–18 cells/μl). The highest
dysplasia), one French bulldog (MUO), one Soft coated Wheaten terrier count of 18 cells/μl was found in a dog with peri-ictal changes on MRI. All
(suspected meningioma), one Boxer (glioma), one Miniature Poodle 11 of these dogs with an elevated TNCC were ultimately diagnosed with
(meningioma) and one English bulldog (glioma) (Table 1). idiopathic epilepsy, received no immunosuppressive medications and had
When split into age at diagnosis groups, the distribution of no neurological abnormalities at follow up. Abnormal protein
structural causes of epileptic seizures was as follows: 0/15 (0%) dogs concentration was reported in four (30.8%) of the dogs with abnormal
<6 months, 0/26 (0%) dogs ≥6 months <1y, 4/254 (1.6%) dogs ≥1 year CSF results (median 0.5 g/L, range 0.46–0.58 g/L). The dog with the
≤6 years (three neoplastic and one anomalous), 3/66 dogs (4.5%) highest total protein result of 0.58 g/L was diagnosed with idiopathic
>6 years ≤8 years (two neoplastic and one inflammatory), and 19/51 epilepsy and concurrent Chiari-like malformation and
(17.6%) dogs >8 years (eight neoplastic and one anomalous). syringohydromyelia. One dogs with elevated total protein was diagnosed
When split into age at first epileptic seizure groups, the with a structural cause of their epilepsy (porencephaly).
distribution of structural causes of epileptic seizures was as follows: The Spearman’s rank correlation coefficient between age at
0/25 (0%) dogs <6 months, 0/41 (0%) dogs ≥6 months <1y, 4/256 diagnosis and age at first epileptic seizure was 0.847 so only age at first
(1.6%) dogs ≥1 year ≤6 years (three neoplastic and one anomalous), epileptic seizure was included in the logistic regression model. On
3/51 (5.9%) dogs >6 years ≤8 years (one inflammatory and two univariable analysis, breed, age at first epileptic seizure, the presence
neoplastic), 9/39 (23.1%) dogs >8 years (eight neoplastic and one of cluster seizures and the presence of status epilepticus were
anomalous) (Table 2). associated with a diagnosis of structural disease. Only two variables
Overall, the median time between the onset of epileptic seizures and remained significant at the multivariate level: age at first epileptic
the time of diagnosis was 2.5 months (range 0.1–86 months, IQR seizure was associated with a diagnosis of structural disease, with risk
1–7 months), but in the structural group it was 1 month (range increasing as age group increased (p < 0.001, OR = 4.390, CI 2.388–
0.1–20 months, IQR 0.2–1.6 months). The majority of patients with a 8.072); history of status epilepticus was also associated with an
structural cause of epileptic seizures had been experiencing epileptic increased risk of a diagnosis of structural disease (p = 0.049,
seizures for 2 months or less (13/16, 87.5%), however two patients OR = 4.389, CI 1.010–19.078).
diagnosed with intracranial neoplasia had experienced epileptic seizures

TABLE 2 Diagnosis of structural disease in dogs with epilepsy and a Discussion

normal inter-ictal neurological examination.
From a total of 412 dogs that underwent an MRI of the brain to
Age at First No. (%) with Disease type
investigate causes of epilepsy with a normal inter-ictal neurological
Seizure structural disease
examination, 16 (3.9%) had a structural cause of their epileptic
<6 months 0/25 (0)
seizures. This is considerably less than that previously reported in the
≥6 months <1 year 0/41 (0) literature of 11.8–31% (11–13). Patients with structural disease tended
≥1 year ≤6 years 4/256 (1.6%) 3 neoplasia, 1 anomalous to be older, with no animals under the age of one having a structural
>6 years ≤8 years 3/51 (5.9%) 2 neoplasia, 1
cause of their epileptic seizures, consistent with those findings of Bush
inflammatory
et al. (11). There were only 66 patients less than 1 year old at the time
of the onset of their epileptic seizures in the current study, which may
>8 years 9/39 (23.1%) 8 neoplasia, 1 anomalous
explain the low numbers, however in a previous study looking

Frontiers in Veterinary Science 04 frontiersin.org

Phillipps and Goncalves 10.3389/fvets.2024.1507861

specifically at the cause of epileptic seizures in 136 dogs under 1 year this study aids understanding of the importance of a full neurological
of age, of those with a normal neurological examination, only 6/114 examination and thorough history taking when considering most
(5.3%) had a structural cause of their epileptic seizures (14). Juvenile likely differential diagnoses in these patients, and that this will result
epilepsy has previously been thought to be more likely secondary to in logical and rational decision making regarding further workup
structural disease (6) including congenital brain malformations (14). and treatment.
Our findings in conjunction with previously published works however Of the 12 Border terriers included in our population, four (33.3%)
indicate that young patients with a normal inter-ictal neurological had structural lesions and of the 18 Staffordshire bull terriers, four
examination are most commonly diagnosed with idiopathic epilepsy, (22.2%) has structural lesions, with all lesions suspected to
with other causes very uncommon. MRI is recommended early in the be neoplastic (gliomas). These breeds were also in the top three most
course of epilepsy in young children, as imaging changes secondary common breeds with intra-axial neoplasia as a cause of their seizures
to ongoing epileptic seizure activity may complicate the diagnosis of in the study examining causes of epileptic seizures in 900 dogs,
lesions that may respond well to surgical management (22). In alongside boxers (15). A previous study recommended that all Boxer
veterinary medicine, although surgical management of epilepsy has dogs that experience epileptic seizures should undergo an MRI of the
been described (26), this is currently rarely offered, and therefore brain due to a high prevalence of structural disease as the cause of
performing MRI in these cases, unlike in human medicine, is mainly epileptic seizures in this hospital population (32). There is incomplete
to exclude structural causes rather than for surgical assessment and information on the number of patients with a normal neurological
planning and the human recommendations are therefore examination in this study, but a high proportion (at least 44/74 dogs)
not applicable. had an abnormal examination. There were only six Boxers in our
The proportion of patients with structural disease increased from population, one (16.7%) with a structural lesion. Although care must
1 year of age, with patients over 8 years of age at the onset of their be taken not to give undue significance to these results given the lack
epileptic seizures having the highest proportion of structural disease of significant association between breed and structural disease, the
(n = 16, 23.1%). This finding is consistent with previously published relatively higher percentage of dogs with structural abnormalities
work indicating that older animals with a normal inter-ictal within these breeds may be worthy of further exploration with a
neurological examination are at increased risk of structural disease larger study.
compared with younger animals (12, 13). Within our patient Occurrence of status epilepticus was associated with an increased
population however, the risk of structural disease in dogs over 6 years risk of a structural lesions in our patient population. Although status
of age was 13.3% (12/90) which is lower than the previously reported epilepticus can be secondary to any cause of epileptic seizures (18–20,
26.7% in dogs aged over six (12) and 31% in dogs aged seven or older 33), a region of abnormal parenchyma may logically prevent intrinsic
(13). This likely reflects the more rigorous exclusion of dogs with inhibitory neural networks from preventing further seizure
inter-ictal abnormalities in the current study which is unique in the propagation. Additionally, neoplasia may lead to upregulation of
methodology compared to these previous works. P-glycoprotein with resultant decreased concentrations of ASDs
Full neurological examination cannot exclude significant reaching their therapeutic targets in the brain, exacerbating status
structural disease, particularly in so-called ‘silent’ areas of the brain epilepticus (33). It is possible that diagnostic yield of performing brain
such as the olfactory region and piriform lobes (27). The most MRI for the investigation of epileptic seizures in patients with a
common structural lesion in this patient population was neoplasia, normal inter-ictal examination may therefore be greater in patients
affecting 13/16 (81.3%) of patients with structural disease. Slowly that are older at the onset of their epileptic seizures and have
progressive disease processes such as slow-growing neoplasms may experienced status epilepticus.
be particularly challenging to detect clinically given the brain’s Incidental findings on MRI were relatively common, with 60/412
remarkable adaptation to slow change (28). These disease processes patients (14.6%) having abnormalities not expected to cause epileptic
are more likely to affect the middle-aged to older canine population seizures reported. MRI is a highly sensitive diagnostic imaging
(29), which reflects the recommendations that animals over a certain technique, capable of lesion detection with high sensitivity (34).
age may benefit from advanced imaging to exclude structural lesions. Despite the use of high-field MRI in the current study, the proportion
It is important to appreciate however that 30/39 of dogs (76.9%) aged of dogs diagnosed with a structural lesion was less than that in the
over 8 years old at the time of their first epileptic seizure in this study comparative low-field study. This likely reflects the rigorous exclusion
were diagnosed with epilepsy of unknown cause. In human medicine, of neurologically abnormal animals in the current study design,
the risk of developing epilepsy is bimodally distributed, with children however the improved resolution of 1 and 1.5T images compared to
aged <1 year of age having the highest risk, normalising to adult values the previous 0.3T study might also result in increased confidence in
by the age of 10 and then increasing again at around 60 years of age, determining incidental from causative findings (35). It is possible
with the highest risk >85 years (30). Our results suggest that geriatric however that smaller anatomical variations such as cortical dysplasia
onset epilepsy is also common in dogs, and that new onset epileptic and early changes such as focal hippocampal sclerosis may be better
seizures in an older pet should not be met with fear of a guarded visualised with 3T images or super high-field MRI (34), particularly
prognosis in animals that are neurologically normal. Although it is given the small volume of the canine brain compared to humans.
sensible to discuss the merits of MRI given the relative risk of clinically Limitations of the current study include the retrospective nature
silent structural lesions in older dogs, most of these patients will have which prevented individual case follow-up. Disease that was
a normal scan. Neurological disorders are among the top three causes microscopic at the time of diagnosis could therefore not be completely
of death in dogs in the UK (31), and this may be unwittingly biased by excluded, for example paraneoplastic causes. Although MRI protocol
clinicians unduly considering neoplasia as the only likely cause of was consistent throughout the study period at our institution, many
clinical signs, with its associated guarded prognosis. Our hope is that patients did not undergo CSF collection and analysis. Cerebrospinal

Frontiers in Veterinary Science 05 frontiersin.org

Phillipps and Goncalves 10.3389/fvets.2024.1507861

fluid sampling has been reported to be very safe in dogs (36), however in accordance with the local legislation and institutional requirements.
the diagnostic utility in dogs with a normal inter-ictal examination is Written informed consent was obtained from the owners for the
low (37, 38). Additionally, CSF abnormalities may be secondary to participation of their animals in this study.
epileptic seizures themselves, rather than a marker of underlying
aetiology (39). Although inflammatory disease cannot be completely
excluded in the patients without CSF analysis, the authors propose that Author contributions
the diagnosis would be unlikely to be changed in these patients given
the low value of this test in dogs with normal neurological examinations SP: Data curation, Formal analysis, Funding acquisition,
cited in other works. Finally, the diagnosis of epileptic seizure activity Investigation, Methodology, Project administration, Visualization,
largely relied on owners descriptions of events. Electroencephalography Writing – original draft, Writing – review & editing. RG:
(EEG) was not performed in any cases and although video footage was Conceptualization, Formal analysis, Methodology, Supervision,
also reviewed where possible, cases of other types of paroxysmal event Validation, Writing – review & editing.
such as paroxysmal dyskinesia cannot be excluded (40). The study
period spanned over nearly two decades and paroxysmal dyskinesia in
particular is becoming more frequently recognised in veterinary Funding
neurology across many breeds of dog (41, 42). As paroxysmal dyskinesia
episodes are rarely associated with structural disease, this may falsely The author(s) declare that financial support was received for the
lower the percentage of animals with no structural lesions on MRI if research, authorship, and/or publication of this article. This
these cases are included. As our understanding of these conditions manuscript was financially supported by the University of Liverpool
deepens, better testing to increase confidence in episode type (epileptic for Open Access Publishing.
seizures vs. other) might help this diagnostic conundrum. Currently
although frequently reported to be feasible in practice (43), few centres
routinely offer EEG as part of the epileptic seizure workup. Acknowledgments
The authors acknowledge the role of all clinicians and support
Conclusion staff responsible for the examination, investigation and treatment of
these patients and their documentation of these processes.
This study found that the most common cause of epileptic seizures
in dogs with a normal inter-ictal neurological examination is epilepsy
of unknown cause/idiopathic epilepsy, regardless of the age of onset Conflict of interest
of epileptic seizures. Although a normal neurological examination
cannot completely exclude significant structural disease, most animals The authors declare that the research was conducted in the
presenting with epileptic seizures and a normal inter-ictal examination absence of any commercial or financial relationships that could
will have a normal MRI scan. The age at diagnosis was associated with be construed as a potential conflict of interest.
a significant increase in risk of a structural lesion in the over 8 years
group, as did the presence of status epilepticus. It may be pertinent
therefore to consider risk on a patient–patient basis when Generative AI statement
recommending an MRI scan in these cases.
The author(s) declare that no Gen AI was used in the creation of
this manuscript.
Data availability statement
The raw data supporting the conclusions of this article will Publisher’s note
be made available by the authors, without undue reservation.
All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
Ethics statement organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
The animal studies were approved by the University of Liverpool claim that may be made by its manufacturer, is not guaranteed or
Veterinary Research Ethics Committee. The studies were conducted endorsed by the publisher.

References
1. World Health Organisation (2019). Epilepsy: a public health imperative. Available 3. Kearsley-Fleet L, O'Neill DG, Volk HA, Church DB, Brodbelt DC. Prevalence and
at: https://www.who.int/publications/i/item/epilepsy-a-public-health-imperative risk factors for canine epilepsy of unknown origin in the UK. Vet Rec. (2013) 172:338.
(Accessed December 8, 2021). doi: 10.1136/vr.101133
2. Erlen A, Potschka H, Volk HA, Sauter-Louis C, O'Neill DG. Seizure occurrence in 4. PDSA (2023). PDSA Animal Wellbeing (PAW) Report 2023. Available at: https://
dogs under primary veterinary care in the UK: prevalence and risk factors. J Vet Intern www.pdsa.org.uk/what-we-do/pdsa-animal-wellbeing-report/paw-report-2023
Med. (2018) 32:1665–76. doi: 10.1111/jvim.15290 (Accessed September 1, 2023).

Frontiers in Veterinary Science 06 frontiersin.org

Phillipps and Goncalves 10.3389/fvets.2024.1507861

5. Licht BG, Lichgt MH, Harper KM, Lin S, Curtin JJ, Hyson LL, et al. Clinical 24. Nagendran A, McConnell JF, De Risio L, José-López R, Quintana RG, Robinson
presentations of naturally occurring canine seizures: similarities to human seizures. K, et al. Peri-ictal magnetic resonance imaging characteristics in dogs with suspected
Epilepsy & behavior. (2002). 3, 460–70. idiopathic epilepsy. J Vet Intern Med. (2021) 35:1008–17. doi: 10.1111/jvim.16058
6. De Risio L, Bhatti S, Muñana K, Penderis J, Stein V, Tipold A, et al. 25. Suñol A, Garcia-Pertierra S, Faller KME. Cerebrospinal fluid analysis in dogs:
International veterinary epilepsy task force consensus proposal: diagnostic Main patterns and prevalence of albuminocytological dissociation. Vet Rec. (2021)
approach to epilepsy in dogs. BMC Vet Res. (2015) 11:148. doi: 10.1186/ 188:e27. doi: 10.1002/vetr.27
s12917-015-0462-1
26. Hasegawa D, Saito M, Kitagawa M. Neurosurgery in canine epilepsy. Vet J. (2022)
7. Armaşu M, Packer RMA, Cook S, Solcan G, Volk HA. An exploratory study using 285:105852. doi: 10.1016/j.tvjl.2022.105852
a statistical approach as a platform for clinical reasoning in canine epilepsy. Vet J. (2014)
27. Schwartz M, Lamb CR, Brodbelt DC, Volk HA. Canine intracranial neoplasia:
202:292–6. doi: 10.1016/j.tvjl.2014.08.008
clinical risk factors for development of epileptic seizures. J Small Anim Pract. (2011)
8. Podell M, Fenner WR, Powers JD. Seizure classification in dogs from a nonreferral- 52:632–7. doi: 10.1111/j.1748-5827.2011.01131.x
based population. J Am Vet Med Assoc. (1995) 206:1721–8. doi: 10.2460/
28. Rossmeisl J, Pancotto T. Intracranial neoplasia and secondary pathological effects
javma.1995.206.11.1721
In: S Platt and L Garosi, editors. Small Animal Neurological Emergencies. Barcelona:
9. Pákozdy Á, Leschnik M, Tichy A, Thalhammer J. Retrospective clinical comparison Manson Publishing (2012). 461–78.
of idiopathic versus symptomatic epilepsy in 240 dogs with seizures. Acta Vet Hung.
29. Miller AD, Miller CR, Rossmeisl JH. Canine primary intracranial Cancer: a
(2008) 56:471–83. doi: 10.1556/avet.56.2008.4.5
Clinicopathologic and comparative review of glioma, meningioma, and choroid plexus
10. Ghormley TM, Feldman DG, Cook JR. Epilepsy in dogs five years of age and older: 99 tumors. Front Oncol. (2019) 9:1151. doi: 10.3389/fonc.2019.01151
cases (2006–2011). J Am Vet Med Assoc. (2015) 246:447–50. doi: 10.2460/javma.246.4.447 30. Beghi E. The epidemiology of epilepsy. Neuroepidemiology. (2019) 54:185–91. doi:
11. Bush WW, Barr CS, Darrin EW, Shofer FS, Vite CH, Steinberg SA. Results of 10.1159/000503831
cerebrospinal fluid analysis, neurologic examination findings, and age at the onset of 31. O’Neill DG, Church DB, McGreevy PD, Thomson PC, Brodbelt DC. Longevity
seizures as predictors for results of magnetic resonance imaging of the brain in dogs and mortality of owned dogs in England. Vet J. (2013) 198:638–43. doi: 10.1016/j.
examined because of seizures: 115 cases (1992–2000). J Am Vet Med Assoc. (2002) tvjl.2013.09.020
220:781–4. doi: 10.2460/javma.2002.220.781
32. Loncarica T, Balducci F, Bernardini M. Prevalence of idiopathic epilepsy and
12. Smith PM, Talbot CE, Jeffery ND. Findings on low-field cranial MR images in structural epilepsy in 74 boxer dogs in a referral hospital. Front Vet Sci. (2022) 9:956648.
epileptic dogs that lack interictal neurological deficits. Vet J. (2008) 176:320–5. doi: doi: 10.3389/fvets.2022.956648
10.1016/j.tvjl.2007.03.003
33. Charalambous M, Bhatti SFM, Volk HA, Platt S. Defining and overcoming the
13. Schwartz M, Muñana KR, Nettifee-Osborne J. Assessment of the prevalence and therapeutic obstacles in canine refractory status epilepticus. Vet J. (2022)
clinical features of cryptogenic epilepsy in dogs: 45 cases (2003–2011). J Am Vet Med 283-284:105828. doi: 10.1016/j.tvjl.2022.105828
Assoc. (2013) 242:651–7. doi: 10.2460/javma.242.5.651
34. van Lanen R, Colon AJ, Wiggins CJ, Hoeberigs MC, Hoogland G, Roebroeck A,
14. Arrol L, Penderis J, Garosi L, Cripps P, Gutierrez-Quintana R, Gonçalves R. et al. Ultra-high field magnetic resonance imaging in human epilepsy: a systematic
Aetiology and long-term outcome of juvenile epilepsy in 136 dogs. Vet Rec. (2012) review. Neuroimage Clin. (2021) 30:102602. doi: 10.1016/j.nicl.2021.102602
170:335. doi: 10.1136/vr.100316
35. Arnold TC, Freeman CW, Litt B, Stein JM. Low-field MRI: clinical promise and
15. Hall R, Labruyere J, Volk H, Cardy TJ. Estimation of the prevalence of idiopathic challenges. J Magn Reson Imaging. (2023) 57:25–44. doi: 10.1002/jmri.28408
epilepsy and structural epilepsy in a general population of 900 dogs undergoing MRI
for epileptic seizures. Vet Rec. (2020) 187:e89. doi: 10.1136/vr.105647 36. Fentem R, Nagendran A, Marioni-Henry K, Madden M, Phillipps S, Cooper C,
et al. Complications associated with cerebrospinal fluid collection in dogs. Vet Rec.
16. Hori M, Hagiwara A, Goto M, Wada A, Aoki S. Low-field magnetic resonance (2023) 193:e2787. doi: 10.1002/vetr.2787
imaging: its history and renaissance. Investig Radiol. (2021) 56:669–79. doi: 10.1097/
RLI.0000000000000810 37. Gilbert SE, Cardy TJ, Bertram S, Taylor-Brown F. Diagnostic utility of cerebrospinal
fluid analysis in dogs with suspected idiopathic epilepsy. Aust Vet J. (2021) 99:1–5. doi:
17. Monteiro R, Adams V, Keys D, Platt SR. Canine idiopathic epilepsy: prevalence, 10.1111/avj.13018
risk factors and outcome associated with cluster seizures and status epilepticus. J Small
Anim Pract. (2012) 53:526–30. doi: 10.1111/j.1748-5827.2012.01251.x 38. Coelho AM, Maddox TW, Sanchez-Masian D, Gonçalves R. Diagnostic value of
cerebrospinal fluid analysis in a population of dogs with suspected idiopathic epilepsy.
18. Cagnotti G, Ferrini S, Ala U, Bellino C, Corona C, Dappiano E, et al. Analysis of Vet Rec. (2019) 185:539. doi: 10.1136/vr.105438
early assessable risk factors for poor outcome in dogs with cluster seizures and status
epilepticus. Front Vet Sci. (2020) 7:7. doi: 10.3389/fvets.2020.575551 39. Gonçalves R, Anderson TJ, Innocent G, Penderis J. Effect of seizures on
cerebrospinal fluid analysis in dogs with idiopathic epilepsy. Vet Rec. (2010) 166:497–8.
19. Fentem R, de Stefani A, Quintana RG, Alcoverro E, Jones GMC, Amengual-Batle doi: 10.1136/vr.b4812
P, et al. Risk factors associated with short-term mortality and recurrence of status
epilepticus in dogs. J Vet Intern Med. (2022) 36:656–62. doi: 10.1111/jvim.16353 40. Packer RMA, Berendt M, Bhatti S, Charalambous M, Cizinauskas S, De Risio L,
et al. Inter-observer agreement of canine and feline paroxysmal event semiology and
20. Zimmermann R, Hülsmeyer V-I, Sauter-Louis C, Fischer A. Status epilepticus and classification by veterinary neurology specialists and non-specialists. BMC Vet Res.
epileptic seizures in dogs. J Vet Intern Med. (2009) 23:970–6. doi: (2015) 11:39. doi: 10.1186/s12917-015-0356-2
10.1111/j.1939-1676.2009.0368.x
41. Cerda-Gonzalez S, Packer RA, Garosi L, Lowrie M, Mandigers PJJ, O'Brien DP,
21. Shoop-Worrall SJ, O'Neill DG, Viscasillas J, Brodbelt DC. Mortality related to et al. International veterinary canine dyskinesia task force ECVN consensus statement:
general anaesthesia and sedation in dogs under UK primary veterinary care. Vet Anaesth terminology and classification. J Vet Intern Med. (2021) 35:1218–30. doi: 10.1111/
Analg. (2022) 49:433–42. doi: 10.1016/j.vaa.2022.03.006 jvim.16108
22. Bernasconi A, Cendes F, Theodore WH, Gill RS, Koepp MJ, Hogan RE, et al. 42. Mandigers PJJ, Santifort KM, Lowrie M, Garosi L. Canine paroxysmal
Recommendations for the use of structural magnetic resonance imaging in the care of dyskinesia—a review. Front Vet Sci. (2024) 11:11. doi: 10.3389/fvets.2024.1441332
patients with epilepsy: a consensus report from the international league against epilepsy
43. Lyon E, Pochat H, Blot S, Troupel T, Van Caenegem N, Besnard S, et al. Use of
neuroimaging task force. Epilepsia. (2019) 60:1054–68. doi: 10.1111/epi.15612
video-electroencephalography as a first-line examination in veterinary neurology:
23. Westbrook C, Roth CK, Talbot J. Instrumentation and Equipment' in MRI in development and standardization of electroencephalography in unsedated dogs and cats.
practice. 4th ed. West Sussex, UK: Wiley-Blackwell (2012). Front Vet Sci. (2024) 11:1326165. doi: 10.3389/fvets.2024.1326165

Frontiers in Veterinary Science 07 frontiersin.org