Parts of the study were presented at the “17. Jahrestagung der Fachgruppe Innere Medizin und Klinische Labordiagnostik der DVG, January 31 to February 1, 2009, Berlin, Germany.
Corresponding author: Andrea Fischer, Priv. Doz Dr med vet, Diplomate ACVIM (Neurology), Section of Neurology, Clinic of Small Animal Medicine, Ludwig-Maximilians-University Munich, 80539 Munich, Germany; e-mail: email@example.com
Background: A special form of epileptic seizures (ES) is the life-threatening condition of status epilepticus (SE), which requires immediate and specific treatment based on a correct diagnosis of the underlying disease condition.
Hypothesis/Objectives: The objectives of this retrospective study were to determine prevalence of ES and SE in dogs presenting at a veterinary teaching hospital, to identify the etiology and relative risk (RR) for SE in general and at the onset of seizures. Furthermore the outcome for dogs suffering from SE was to be evaluated.
Animals: Three hundred and ninety-four dogs that were admitted to a veterinary teaching hospital (January 1, 2002 to March 31, 2008) with ES.
Methods: All medical records of dogs with ES were identified by screening the clinical documentation system and evaluated for inclusion in this retrospective study.
Results: Dogs with reactive seizures caused by poisoning had a significantly higher risk of developing SE (P < .001; RR = 2.74), particularly as 1st manifestation of a seizure disorder (P= .001; RR = 1.97). After SE, dogs with symptomatic epilepsy had a significantly lower probability of survival than dogs with idiopathic epilepsy (P < .001) and reactive ESs (P= .005).
Conclusion and Clinical Importance: In dogs showing SE as the 1st manifestation of a seizure disorder, intoxication should always be considered and appropriate investigations undertaken. Dogs with SE owing to toxicosis have more favorable outcomes than dogs with symptomatic epilepsy (P < .001).
status epilepticus in course of a seizure disorder
Epilepsy is the most common neurological disorder in human medicine, affecting 50 million people worldwide, especially in developing countries.1 Seizure disorders are also the most common neurological disease in dogs.2
Epileptic seizures (ES) are commonly divided into 3 groups based on the underlying disease condition: idiopathic and symptomatic epilepsy and reactive ES. The term idiopathic or primary epilepsy is used when no underlying cause can be identified and a familial predisposition is presumed.3,4 If ES are the result of structural disease of the brain, they are by definition characterized as symptomatic or secondary epilepsy.5 Reactive ES are the healthy brain's reaction to a temporary systemic disorder.2 The underlying causes of reactive ES are endogenous metabolic diseases or exogenous poisonings.6
A special form of ES is the condition of status epilepticus (SE) that has been studied comparatively little in dogs to date.7 SE is characterized by prolonged seizure activity. Dogs of any sex, breed, and age can be affected and this condition can occur as a result of a number of underlying causes.7–9 SE occurs in dogs with known chronic seizure disorders, but there is a lack of exact data about the incidence.8 Apart from chronic seizure disorders, SE can be associated with inflammation of the central nervous system (CNS), brain tumors, metabolic-toxic disorders, trauma, and vascular events.10 Finally underlying disease conditions of SE are the same as in ES.
There is a failure of mechanisms that usually terminate seizure activity in SE.10 It is estimated that neuronal death in prolonged seizure activity of SE is mostly caused by excitotoxicity related to glutamate10,11 and extracellular calcium (Ca2+) that enters through the N-methyl-d-aspartate (NMDA) glutamate receptor.12,13 Furthermore, untreated seizure activity may lead to complications such as hyperthermia, hyperglycemia, hypoglycemia, hypoxia, acidosis, renal failure, disseminated intravascular coagulation, or cardiopulmonary collapse. Therefore, SE is classified as a life-threatening condition and is considered to be a medical emergency requiring immediate treatment.10,14 Therapy for SE should be determined on an individual basis based on the dog's clinical status.10
In the past, SE was often defined as a seizure activity lasting longer than 20–30 minutes. This timeframe is the result of estimation of the duration of seizure activity necessary to cause neuronal damage.15,16 It is necessary in practice to start therapy when seizures have lasted for <20 minutes.14 Therefore, today SE is defined as continuous seizure activity that lasts for more than 5 minutes, or as 2 or more discrete seizures between which there is incomplete recovery of consciousness.5,7,8,10,14,17–19 This definition distinguishes SE from isolated ES and cluster seizures (between which there is complete recovery of consciousness), which can be summarized as non-SE seizures (nSE).
The purpose of this retrospective study was to determine prevalence of ES and SE in dogs presenting at a veterinary teaching hospital and to identify the etiology of canine SE and therefore to assess the relative risk (RR) for SE in general and at the onset of seizures. Possible impact factors like sex, breed, and age at onset should be assessed. Furthermore the outcome for dogs suffering from SE was to be evaluated.
Materials and Methods
Dogs with ES admitted to the Clinic of Small Animal Medicine of Munich (LudwigMaximilians-University Munich [LMU]) between January 1, 2002 and Machr 31, 2008 were screened for entry in this study. The general inclusion criterion was a sufficient medical documentation concerning the sex, breed, and age at seizure onset, seizure type, and seizure history. For evaluation of etiology of SE and nSE seizures, only dogs with a sufficient and well-documented diagnostic work-up that led to a diagnosis according to defined criteria were eligible.
Review of Medical Records
The following clinical variables were extracted from medical records: breed, sex, age at seizure onset, seizure type, and final clinical diagnosis.
Seizures were classified by etiology as idiopathic epilepsy, symptomatic epilepsy (categorized as inflammatory brain diseases, cerebral tumors and “other” [including vascular disease, malformations, seizures immediately after craniocerebral trauma or posttraumatic epilepsy]) or reactive ES (caused by metabolic disorder or poisoning) depending on the diagnosis.
Idiopathic epilepsy was diagnosed in dogs younger than 6 years old at seizure onset that have recurrent seizures, were normal on interictal neurological and laboratory examination, and did not have any evidence of neurological disease other than seizures at any time during their lives. In some dogs this was further supported by normal results of cerebrospinal fluid (CSF) analyses and advanced imaging (magnetic resonance imaging [MRI] or computed tomography[CT] scan). Diagnosis of CNS inflammation required positive findings in CSF analysis or necropsy. Brain tumor (primary or secondary) was diagnosed with advanced imaging (MRI or CT scans), postmortem examination or strongly suspected if neoplasia was evident on thoracic radiographs. Metabolic disease was diagnosed if severe laboratory changes capable of causing ES (eg, hypoglycemia, hepatic encephalopathy) were evident. Poisoning was diagnosed by an unequivocal history of intake of poisonous material (eg, snail baits) or if qualitative toxicological screening (mass spectrometry, thin layer gas chromatography; Institute of Veterinary Pharmacology and Toxicology, LMU Munich) of stomach contents or body fluids identified a poison known to cause seizures.
The study population (Fig 1) was divided into dogs with and without SE (SE and nSE).
SE was defined as a continuous seizure activity lasting at least 5 minutes or 2 or more discrete seizures between which there is incomplete recovery of consciousness.5,7,8,14,17–19
Dogs with SE and a defined clinical diagnosis were further divided into 2 subcategories: patients with SE as a 1st manifestation of a seizure disorder (SE1) and patients that developed SE during the course of a seizure disorder (SEc).
The RRs for ES and subgroups SE and nSE were calculated for sex and the 10 most frequently represented dog breeds with ES in the study period.
The final outcome for dogs discharged from the hospital was assessed by phone calls with owners.
A χ2 analysis was used to evaluate associations on the one hand between etiology or sex and the occurrence of SE (versus nSE) and on the other hand between etiology and SE at seizure onset (SE1 versus SEc). Furthermore, patient groups were compared with the general hospital population during the study period. A Kruskal-Wallis test and Mann-Whitney U-test were applied to test for differences in the age at seizure onset between patient groups that were normally distributed (Kolmogorow-Smirnow test). Survival times were displayed graphically as Kaplan-Meier curves and evaluated by log-rank tests. The median and range of ages at seizure onset were displayed in box-and-whisker plots. The RR or the hazards ratio (HR) was specified and Cornfield's 95% confidence limits for RR were used. In contingency tables larger than 2 × 2, where any expected cell value was <5, a Yates correction was used. In 2 × 2 contingency tables with any expected cell values <5, the Fisher exact 2-tailed results were used. For all analyses, values of P < .05 were considered significant. All statistical analyses were performed using MedCalc (Version 188.8.131.52).
Of 15,449 dogs admitted to the hospital during the study period, 394 fulfilled the inclusion criteria. All dogs presented with generalized seizures (primary or secondary). Two dogs had partial seizures. The dogs were divided into 2 groups: SE (n = 114) and nSE (n = 280). In the SE group of dogs with a clinical diagnosis (n = 88), 51 dogs were presented with SE as their 1st seizure and in 37 dogs SE occurred during the course of the seizure disorder (Fig 1).
For the study period, the prevalence of ES and SE in the hospital collective was 2.6% (394/15,449) and 0.7% (114/15,449), respectively.
Eighty-eight of 114 dogs (77.2%) and 147/280 dogs (52.5%) in the SE and nSE group, respectively, received a diagnostic work-up that led to a clinical diagnosis according to defined criteria. Thus, dogs suffering from SE were more likely to receive a sufficient diagnostic work-up that finally led to a diagnosis (P < .001, RR = 1.47 [95% CI 1.27; 1.71]).
Sex, Breed, and Age
Sex. Of the hospital population (n = 15,449) in the investigated time 52.4% were male (40.0% sexually intact, 12.4% neutered) and 47.6% were female (30.0% sexually intact, 17.6% spayed). Of the dogs with ES (n = 394), 55.3% were male (40.6% sexually intact, 14.7% neutered), and 44.7% were female (22.1% sexually intact, 22.6% spayed). Looking at the SE group (n = 114), 54.4% were male (38.6% sexually intact, 15.8% neutered) and 45.6% were female (15.8% sexually intact, 29.8% spayed). Of the dogs in the nSE group (n = 280), 55.7% were male (41.4% sexually intact, 14.3% neutered) and 44.3% were female (24.6% sexually intact, 19.6% spayed females).
No significant differences concerning the risk of developing ES or SE could be found between male and female dogs. Yet, spayed females had a significantly higher risk of developing ES (P= .009, RR = 1.37 [95% CI 1.09; 1.73]) and SE (P= .001, RR = 2.00 [95% CI 1.34; 2.98]) compared with the hospital population. Furthermore, for spayed females, the risk of developing ES (P < .001, RR = 1.75 [95% CI 1.31; 2.35]) or SE (P < .001, RR = 3.24 [95% CI 1.83; 5.72]) was significantly higher compared with sexually intact females.
Breed. Dog breeds most frequently affected by ES were (listed in descending order): Golden Retrievers (6.9%), Dachshounds (5.8%), Labrador Retrievers (4.3%), Beagles, Boxers, German Shepherd Dogs (every 2.3%), Rottweilers, Westhighland White Terriers, and Yorkshire Terriers (every 2.0%) and Jack Russell Terrier (1.8%). The distribution of these breeds among the hospital population, the SE and nSE group was Golden Retrievers (3.5, 8.8, 6.1%), Dachshounds (4.5, 7.0, 5.4%), Labrador Retrievers (3.1, 4.4, 4.3%), Beagles (1.0, 2.6, 2.1%), Boxers (1.8, 1.8, 2.5%), German Shepherd Dogs (4.6, 7.0, 2.1%), Rottweilers (1.3, 0, 2.9%), Westhighland White Terriers (2.5, 0.9, 2.9%), Yorkshire Terriers (3.3, 0.9, 2.9%), and Jack Russell Terrier (2.9, 1.8, 1.8%).
Golden Retrievers (P < .001, RR = 2.04 [95% CI 1.39; 2.99]) and Beagles (P= .016, RR = 2.35 [95%CI 1.24; 4.47]) had a significantly higher risk of developing ES compared with the hospital population, whereas German Shepherd Dogs had an increased risk of developing SE (P= .016, RR = 2.17 [95%CI 1.39; 3.38]).
Age at 1st Onset of Seizure Activity. The age at seizure onset in the SE and nSE groups varied from 0.3 to 15.9 years (median 5.7 years) and from 0.2 to 16.3 years (median 4.5 years), respectively (P > .05).
Within the SE group, the age at onset of seizures differed between etiological groups (P < .001). Dogs with idiopathic epilepsy were significantly younger at the onset of seizures than dogs with symptomatic epilepsy (P < .001) or dogs with reactive ES (P= .009). In addition, dogs with reactive ES were significantly younger than dogs with symptomatic epilepsy (P= .009).
Age at 1st onset of seizure activity in dogs with SE went for dogs with primary epilepsy from 0.6 to 5.9 years (median 2.4 years), for dogs with symptomatic epilepsy from 0.6 to 15.9 years (median 9.8 years) and for dogs with reactive ES it ranged from 0.3 to 13.3 years (median 5.6 years). Age at 1st onset of seizure activity in the nSE group went for dogs with primary epilepsy from 0.2 to 5.9 years (median 2.4 years), for dogs with symptomatic epilepsy from 0.2 to 16.3 years (median 7.7 years) and for dogs with reactive ES it ranged from 0.2 to 16.2 years (median 9.4 years). Dogs with symptomatic epilepsy that experienced SE were significantly older at seizure onset than dogs that never experienced SE (P= .013) (Fig 2).
Dogs were furthermore categorized into 3 groups based on their age at seizure onset (<1 year, 1–5 and >5 years). There were significant differences between these age groups concerning their etiology (P < .001). In the <1- and 1–5-year groups, the most frequent cause for seizures was idiopathic epilepsy (57.5, 74.4% respectively), whereas in the >5-year group, symptomatic epilepsy (66.3%) was the most common cause for ES. Looking only at dogs with SE, the distribution was the same: idiopathic epilepsy was most common in the groups <1- and 1–5-year (60.0, 65.4% respectively) and symptomatic epilepsy was most common in the group >5 years (66.7%).
Idiopathic epilepsy was diagnosed in 113 dogs with ES (48.1%), symptomatic epilepsy in 90 dogs (38.3%; 33 inflammatory brain diseases, 43 cerebral tumors, 14 other causes), and reactive ES were observed in 32 dogs (13.6%; 18 metabolic disorders, 14 poisonings). Identified poisons were carbofurane (n = 3), metaldehyde (n = 3), paraoxone (n = 2), crimidine (n = 2), zinc phosphide (n = 1), strychnine (n = 1), diazinon (n = 1), and amphetamine (n = 1).
Table 1 shows the detailed etiological classification with regard to the occurrence of SE.
Table 1. Etiology of seizures in dogs with status epilepticus (SE; n = 88) and without status epilepticus (nSE; n = 147).
The etiologies of seizures for dogs with SE as 1st manifestation (SE1; n = 51) and dogs with SE in course of a seizure disorder (SEc; n = 37) are also shown.
Reactive epileptic seizures
Differences between SE and nSE Dogs. Dogs with reactive ES had a significantly higher risk of developing SE (P= .003, RR = 1.87 [95% CI 1.34; 2.60]). In particular, dogs that were poisoned had a 2.7 times higher risk of developing SE (P < .001, RR = 2.74 [95% CI 2.16; 3.64]) compared with all other dogs, and a risk that was more than 3 times higher than dogs with idiopathic epilepsy (P < .001, RR = 3.18 [95% CI 2.30; 4.39]).
Differences between SE1 and SEc Dogs. Dogs with reactive ES had a significant higher risk of developing SE as a 1st manifestation of a seizure disorder (P= .044, RR = 1.55 [95% CI 1.13; 2.14]), in particular, dogs that had been poisoned presented SE as their 1st seizure (P= .003, RR = 1.97 [95% CI 1.58; 2.47]). Compared with dogs with idiopathic epilepsy, dogs that had been poisoned had a more than 2 times higher risk to develop SE as 1st manifestation (P= .001, RR = 2.75 [85% CI 1.75; 4.32]).
Outcome of Dogs with SE
After admission to the hospital, all dogs were treated individually with rectal and/or IV diazepam (n = 19 and 50, respectively) and IV phenobarbital up to a loading dose (n = 64), depending on possible pretreatment by the referring veterinarian (n = 17 rectal diazepam, n = 11 diazepam IV, n = 23 phenobarbital up to a loading dose) and the clinical presentation. In cases of refractory SE anesthesia with pentobarbital (n = 43) or propofol (n = 1), continuous rate infusion was initiated. Ten dogs were discharged after initial seizure control was achieved on admission day, by the owners' requests; the remaining 78 patients were retained for inpatient treatment. Of the retained patients, 4/78 dogs died (5.1%), 26/78 dogs were euthanized (33.3%), and 48/78 dogs were discharged (61.5%). Of all dogs discharged from the hospital, 31/58 dogs were still alive at follow-up (53.5%), 3/58 dogs died (5.2%), 14/58 dogs were euthanized (24.1%), and 10/58 dogs were unavailable for follow-up (17.2%). Dogs of all etiological groups had further seizures after discharge from hospital, except for dogs that ingested poison, which never experienced any more seizures.
The follow-up time for patients still alive at the time of interview ranged from 0.4 to 7.7 years (median 2.5 years). Dogs that died or were euthanized after discharge had a survival time of 0.1–5.9 years (median 0.8 years). The general median age of dogs at death was 10.2 years.
The survival time of dogs with SE differed significantly between the etiological groups (Fig 3). Dogs with symptomatic epilepsy had a significant lower survival probability after SE than dogs with idiopathic epilepsy (P <.001, HR = 0.11 [95% CI 0.03; 0.37]) and dogs with reactive ES (P= .005, HR = 0.16 [95% CI 0.05; 0.58]). Especially dogs with poisonings had a significant higher survival probability than dogs with symptomatic epilepsy (P < .001, HR = 9.15 [95% CI 2.51; 33.34]). The point of time of SE manifestation (as 1st seizure or in course of a seizure disorder) had no influence on the survival probability after SE (P= .470).
ES can result in the life-threatening condition of SE. Yet, the prevalence of ES and SE in the hospital population is rarely described in veterinary medicine, and currently only limited data on the prevalence of these conditions are available.20 In human medicine, between 0.8 and 1% of the human population, respectively, are estimated to suffer from a form of epilepsy during their lives.21,22 In veterinary medicine, the prevalence of ES in dogs has been estimated to be between 0.5 and 5.0%, or between 1 and 2%,23,24 which is in agreement with our data (2.6%). One restriction is that all dogs admitted to our hospital presented with generalized seizures, so prevalence may even be higher because of partial seizures. Limited data concerning breed-specific prevalence of epilepsy are available in dog breeds affected with idiopathic epilepsy, and may reach up to 9.5%.20
Twelve to 13% of humans with diagnosed epilepsy first present SE.25 There is a lack of data concerning the prevalence of SE in veterinary medicine, but it was assumed that 0.44% of all dogs admitted into veterinary clinics suffer from SE.9 The prevalence of 0.7% in our study supports these data. One restriction is that the prevalence in the previous and our study was calculated using a hospital population. Studies of the prevalence of ES or SE in the general dog population can be difficult to estimate because of a lack of satisfying epidemiological data. Even in human medicine, epidemiological data of SE are still not resolved.26
The fact that dogs suffering from SE were more likely to receive a more comprehensive diagnostic work-up that finally led to a diagnosis may be not surprising. The life-threatening condition of SE impresses and scares the owners more than a short ES may do, so they will more likely investigate the cause of seizure activity.
There is a slight predilection of male dogs for ES in the literature.2,27 A predilection for males is reported in childhood epilepsy, and for idiopathic and symptomatic epilepsy in adult humans.28–30 We were unable to confirm this finding in dogs in the present study. No significant differences between male (sexually intact and neutered) and female dogs (sexually intact and spayed) concerning the risk of developing ES were found. As male dogs also dominated the hospital population in our study (52.4%), the slight overrepresentation of male dogs in the ES (55.3%), SE (54.4%), and nSE (55.7%) groups may only reflect the general sex distribution in the hospital population.
Results of this study did demonstrate that spayed females had a significantly higher risk of developing ES and SE compared with sexually intact females and compared with the hospital population. This finding corresponds with results from another study that also showed an underrepresentation of sexually intact females with ES compared with general hospital admissions.9 In contrast, previous studies report that sexually intact female dogs have an increased seizure frequency during estrus.31 The proconvulsive effect of estrogen and the anticonvulsive effect of progesterone on the seizure threshold of humans have been known and described for a long time.32 Recent evidence also suggests that different levels of these hormones (high estradiol to progesterone rates), caused by the menstrual cycle, lead to an increase in seizures. This circumstance is also known as a catamenial seizure pattern or catamenial epilepsy.8,33,34 According to these findings, one would expect that sexually intact female dogs have a higher risk of developing ES and SE, respectively. However, our data and the results of another study show the opposite.9 As female dogs are usually spayed via the complete removal of the dog's reproductive organs, the production of sex hormones ceases completely. Without the proconvulsive effect of estrogen, one would expect the risk of ES or SE to be decreased; yet our results show the opposite. No conclusive explanation can be made for this observation and further studies are required to elucidate this finding.
In many of the breeds that presented ES, SE, and nSE, predispositions for idiopathic epilepsy have previously been described.22,35–37 Thus, it is not surprising that Golden Retrievers (P < .001, RR = 2.04) and Beagles (P= .016, RR = 2.35) had a significantly higher risk of developing ES compared with the hospital population. German Shepherd Dogs showed an increased risk of developing SE (P= .016, RR = 2.17) compared with nSE seizures. As German Shepherds show a high disposition for a multitude of diseases other than ES, it may be difficult to distinguish between an actual increased risk caused by a lower threshold for seizures in German Shepherds and a higher risk that is provoked by predispositions of this breed for disorders that cause more severe seizures. Surprisingly, for other breeds with a known predisposition for idiopathic epilepsy, no increased risk was detected. This finding may be caused by the use of a hospital population.
Current opinion suggests that the age at onset of seizures provides a clue for the underlying etiology.2,38 Data from this study support this thesis, as 74.4% of dogs with ES and 65.4% of dogs with SE between 1 and 5 years were classified as idiopathic. Yet, in dogs younger than 1 year, idiopathic epilepsy (52.6% ES and 60.0% SE, respectively) was still the most common diagnosis. This finding may relate to different inclusion criteria for idiopathic epilepsy. On the other hand, symptomatic epilepsy (ES 66.3%, SE 66.7%, respectively) predominated in dogs older than 5 years. Interestingly, dogs with symptomatic epilepsy that experienced SE were significantly older than dogs that never experienced SE (P= .013). This may be because of more severe diseases such as cerebral tumors with a manifestly more fulminant progression as well as the generally poorer constitution of elderly dogs.
The 2 most recent epidemiologic studies of SE in dogs showed that 27.0% (28.0%) had idiopathic epilepsy, 35.0% (32.0%) had symptomatic epilepsy and 7.0% (12.0%) had reactive SE. The remaining dogs were classified into groups as undetermined and low antiepileptic drug concentrations.8,9 Data in our study, which showed that of the study subjects, 37.5% had idiopathic epilepsy, 39.8% had symptomatic epilepsy, and 22.7% had reactive ES, reflect these findings if one considers that the idiopathic (39%), symptomatic (44%), and reactive seizure group (17%) add to 100% in one of the mentioned studies.8 In our study, no additional groups, such as undetermined or low antiepileptic drug concentration were considered, so a similar allocation was assumed for the group of dogs with SE but no clinical diagnosis.
In the present study, 51 dogs presented SE as the first manifestation of a seizure disorder (58%, 51/88). This value is greater than previously reported in most other studies. In human medicine, 12–13% of recently diagnosed epilepsies are first presented as SE,14 whereas in dogs, SE was their first seizure in 44%.8 Differences may relate to the large proportion of poisonings in our study population (n = 13), particularly in dogs with SE as the 1st seizure manifestation. Others studies observed only 1, respectively, 4 dogs that suffered from poisonings and presented in SE as their 1st seizure.7,8
As a consequence, dogs suffering from intoxication had a risk of presenting SE that was 2.7 times higher (P < .001), particularly when SE was the 1st presentation of a seizure disorder, compared with dogs with SE due to other etiologies. In contrast, dogs with idiopathic epilepsy had a reduced risk of developing SE (P= .017), and particularly SE as a 1st seizure (P= .003), compared with other etiologies. The data are in support of another study showing that SE was more than 1.5 times as likely because of symptomatic and reactive causes than because of idiopathic epilepsy.8 The high risk of developing SE in dogs that experienced poisonings in our study emphasizes the need to consider poisonings as a cause for SE when it is the 1st manifestation of a seizure disorder. The spectrum of the proven poisonings can also be surprising; besides common poisonings from snail bait (metaldehyde), unusual poisons from substances such as crimidine, carbofurane, and strychnine were also detected.
The fact that dogs that were poisoned never experienced any seizures after discharge from hospital suggests that these dogs may have a very good prognosis if SE can be adequately managed and antidotes administered.
The current study had some limitations, caused by its retrospective nature and a bias in the study population. Different treatment, systemic complications, and prior health problems may also result in an outcome selection bias.
In conclusion, the data presented here suggest that a toxicological screen may be a useful tool in the diagnostic work-up of dogs presenting SE and especially SE as their 1st seizure, depending on geographic location, time period of evaluation, and dog breed (outdoor exposure). In dogs with suspected poisoning, neuroimaging with MRI and CSF analyses may even be postponed for a few days until toxicological results are available and SE is adequately controlled. This also avoids misinterpretation of MR images due to the appearance of reversible T2 hyperintense lesions on the MRI if imaging is performed immediately after severe seizures.39–41 Suspicion of poisoning may be especially true if the dogs had been neurologically normal up to the day of presentation and showed no evidence of a previous neurological disease culminating in SE and if toxin ingestion appears likely from the history.
Romina Zimmermann is supported by a dissertation elite grant (Graduiertenstipendium nach dem Bayerischen Eliteförderungsgesetz).