A multicenter randomized controlled trial of effect of medium-chain triglyceride dietary supplementation on epilepsy in dogs.

Abstract Background Medium‐chain triglyceride (MCT) enriched diet has a positive effect on seizure control and behavior in some dogs with idiopathic epilepsy (IE). Objective To evaluate the short‐term efficacy of MCTs administered as an add‐on dietary supplement (DS) to a variable base diet to assess seizure control and antiseizure drug's (ASD) adverse effect profiles. Animals Twenty‐eight dogs with International Veterinary Epilepsy Task Force Tier II (IVETF) level diagnosis of treated IE with 3 or more seizures in the last 3 months were used. Methods A 6‐month multicenter, prospective, randomized, double‐blinded, placebo‐controlled crossover trial was completed, comparing an MCT‐DS with a control‐DS. A 9% metabolic energy‐based amount of MCT or control oil was supplemented to the dogs' diet for 3 months, followed by a control oil or MCT for another 3 months, respectively. Dogs enrolled in this study satisfied most requirements of IE diagnosis stated by the IVETF II level. If they received an oil DS or drugs that could influence the metabolism of the investigated DS or chronic ASD, the chronic ASD medication was adjusted, or other causes of epilepsy were found, the dogs were excluded from the study. Results Seizure frequency (median 2.51/month [0‐6.67] versus 2.67/month [0‐10.45]; P = .02) and seizure‐day frequency were significantly (1.68/month [0‐5.60] versus 1.99/month [0‐7.42], P = .01) lower when dogs were fed MCT‐DS in comparison with the control‐DS. Two dogs were free of seizures, 3 had ≥50% and 12 had <50% reductions in seizure frequency, and 11 dogs showed no change or an increase in seizure frequency. Conclusions and Clinical Importance These data show antiseizure properties of an MCT‐DS compared to a control oil and support former evidence for the efficacy of MCTs as a nutritive, management option for a subpopulation of drug‐resistant dogs with epilepsy.


| INTRODUCTION
Seizures and epilepsy are the most common neurological signs in dogs affecting an estimated 0.6%-0.8% 1-3 of the population. The various cellular pathophysiological mechanisms leading to epilepsy, characterized by unprovoked recurrent seizures, remain poorly understood. As a result, the enduring predisposition of having epileptic seizures can only be managed by medication for seizure suppression 4 (antiseizure), instead of prevention of epilepsy 5,6 (antiepileptogenic). However, the chronic administration of antiseizure drugs (ASDs) presents difficulties in the balance between the beneficial seizure-suppressive effects and undesirable drug-related adverse effects. 7 Around one-third of dogs continue to seizure despite appropriately managed polypharmacotherapy. [8][9][10] In addition, behavioral and cognitive comorbidities occur in dogs with epilepsy, 11 such as anxiety 12,13 deficits in spatial memory, 14 or cognitive function. 15,16 Seizures, drug-related adverse effects along with changes to behavior and cognitive capabilities all contribute to the reduction in quality of life (QoL) for both dogs and their owners. [17][18][19][20][21] Hence, new treatment strategies should address all aspects of epilepsy to improve the overall welfare of dogs and their owners.
Manipulation of nutrition has attracted increasing considerations as an alternative approach to impacting seizure activity and behavior. [22][23][24] A cross-sectional study revealed that over 60% of owners changed their dog's diet, after being diagnosed with idiopathic epilepsy (IE) in an attempt to improve seizure control and protect their dog from ASD-related adverse effects. 25 The influence of dietary modification on seizure control has been extensively studied in human medicine [26][27][28][29] and rodent epilepsy models, [30][31][32][33] showing antiseizure effects to varying degrees. Similarly, there are effects of nutritional management and diets in dogs with epilepsy 23,24,[34][35][36][37][38][39] and paroxysmal dyskinesia. [40][41][42][43] A 6-month prospective, randomized, double-blinded, placebocontrolled crossover dietary trial comparing a medium-chain triacylglyceride (MCT) kibble diet (10% of the total formula calories from octanoic acid [C8], decanoic acid [C10], lauric acid [C12]) to a standardized placebo diet was completed in 21 chronically ASD-treated dogs with IE ( Figure 1). 24 Consumption of a MCT diet resulted in statistically significant elevation of serum beta hydroxybutyric acid (BHB) concentrations and improvement in both seizure control 24 and behavioral comorbidities. 23 Although many owners add oil supplements or change their dogs' diet, 25 it is currently unknown whether a MCT oil supplement might possess similar beneficial effects. 24 The study hypothesis for the current study was that dietary supplementation with a commercially available MCT oil would reduce seizure frequency in dogs with IE and be well tolerated. Therefore, the aims of this clinical study were to investigate antiseizure effects and tolerability of MCT dietary supplement (MCT-DS) in dogs with IE compared with a standardized control-DS.

| MATERIAL AND METHODS
This study was conducted in accordance with the guidelines of the

| Study design
The study design has been published separately. 44 In brief, the main study hypothesis was that MCTs administered as a DS would reduce seizure frequency. Thus, altered seizure frequency will serve as primary, measurable outcome variable to assess therapeutic effect. Changes in ASD adverse effects and quality of life of dogs with IE were assessed as a secondary outcome.
Based on the former research conducted 22 dogs completing the trial appeared sufficient to reach significance between groups. 24

| Dietary supplements
The experimental control and test products were commercially available DSs suitable for human consumption. The test oil was a MCT-DS purified from palm and rapeseed oil containing 50%-65% octanoic acid (C8) and 30%-50% decanoic acid (C10) with 8.37 kcal/mL from 93% saturated fatty acids (End GmbH, Germany, Batch No. L 16M12). The control oil was colorless, extra-virgin olive oil with 8 kcal per ml containing 11% saturated, 11% polyunsaturated, and 78% monounsaturated fat (Filippo Berio, Italy, Batch No LE194M04). Both oils were dispensed in brown bottles ensuring blinding of all involved. Prior to study setup, both oil supplements have been tested in a small cohort of healthy nonepileptic dogs (n = 19) on palatability, preference, side effects, and eating behavior confirming comparability.
The control and the MCT oil were stored at room temperature in a locked room at each test site. Owners were educated to keep the base diet consistent throughout the study. The quantity of DS-oil given daily was calculated according to the metabolic weight, age, and neuter F I G U R E 1 CONSORT flow diagram: flow diagram of the progress through the phases of this randomized clinical trial of 2 groups (ie, enrolment, intervention allocation, follow-up, and data analysis) status of each dog individually and represented 9% isoenergetic requirement per day. The base diet was recorded, but not individually checked to fulfill nutritional requirements. A deviation of ±10% of the base diet consumption (kg) was allowed taking into consideration differences in activity level, physical condition, and to prevent gain or loss of body weight. and bitches known or suspected to be pregnant or lactating were all excluded from the study. Dogs were enrolled as an individual experimental unit, and only 1 dog per household were enrolled onto this study. A unique study case number, consisting of a 3-digit number (study side number-study case number), ascending in a chronological order of enrolment, was assigned and used to identify each dog throughout the study. In total, 36 dogs were recruited onto this study.

| SEIZURE DATA
Information from a standardized seizure diary per dietary intervention period was used to assess the short-term effects of MCT-DS on seizure control. 44 At each study visit, the owners were repeatedly instructed, how to complete the seizure diary and each listed seizure event has been reevaluated by the owner's description. Only generalized seizures were counted toward the efficacy measurement and the primary outcome variable is the change of seizure frequency. Therapeutic success was defined as a significant difference (reduction) in 1 of seizure frequency (seizures/month), seizure days frequency per months, seizure cluster days frequency per months, seizure severity measured by number of episodes of cluster seizures or status epilepticus during a 3-month period between the MCT, and control periods. Finally, each dog will be classified as a "MCT responder" if it experiences at least 50% reduction in seizure frequency between both dietary periods.

| Visual analog score
Visual analog score for ataxia, sedation and overall QoL was completed by the owner at each study visit. The owner was asked to mark a measurement line ranging from 0 to 100 mm (0%-100%) perpendicularly with a secondary intersecting line that presented best the subjective severity (0% [better, than normal] to 100% [cannot be worse]). A perpendicular line at 0 mm, baseline (B), represented "asymptomatic/normal" and at 100 mm represented either "ataxia so severe dog is unable to walk" or "sedation to the extent dog only sleeps" or "QoL is so poor euthanasia is requested," respectively. The length from 0 to the perpendicular line in mm was compared between periods and expressed as a percentage value (0%-100%, negative = improvement, positive = deterioration).

| Owner-based epilepsy disease-specific quality of life aspects (EpiQoL)
In

| Adverse events and concomitant treatments
Any adverse, abnormal, or both events occurring after enrolment were reported to the investigator and documented (Table S1). Seizure occurrences throughout the study were not classified as adverse events.
Concomitant treatments administered throughout the study were recorded detailing indication, products used and strength, dose and length of treatments administered (Table S2). Concomitant changes of ASD medication or dosages throughout the study led to study exclusion.   (Table S3). The study population consisted of 16 males, of which were 12 neutered and 4 were intact, and 12 females, of which 10 were neutered and 2 intact. The population of dogs was on average 5.46 ± 2.61 years of age and weighed 25.6 ± 13.4 kg at start of trial.

| Antiseizure medication regimes
All 28 dogs were drug resistant to at least 1 of the ASD given and defined as partial ASD-nonresponders with less than 50% reduction in seizure frequency. Two dogs received imepitoin only, the rest were under chronic ASD combination treatment (N = 26; Table S4)
Potential biases such as treatment order (MCT-control, control-MCT) or the effect of the study center were considered in monovariate comparison and multivariate statistical mixed model.
On monovariate comparison, the average reduction rate in seizure frequency appeared not significantly different for treatment order or study center; and no association of the supplementation group with treatment order (r = .1793, X 2 = 0.03216, P = .36) or study center (r = 0.1629, X 2 = 0.02653, P = .41) has been identified.
Using the type of DS as fixed and treatment order and study location as random effects in a multivariate analysis as linear mixed model, seizure frequency (P = .02) and seizure day frequency (P = .01) appeared significantly reduced during MCT treatment period compared to control.

| Effects on VAS and adverse effect profiles of antiepileptic drug medication
There were significant differences in the VAS compared to baseline.

| Adverse events and disease-related death
Observation about any unfavorable, unintended health abnormality occurred after enrolment, regardless it was considered as MCT-DS-or control DS-related event, was reported to the investigator and documented (Table S1). Independent of the administered DS, no adverse event occurred more than once during 1 dietary period or was predominantly found associated to the type of DS. Seizure occurrences and disease-related death throughout the study were not classified as adverse events and considered as sudden unexpected death in epilepsy. 6,35,46

| Withdrawn and excluded dogs
Eight further dogs were recruited but did not complete the study, of which 2 were excluded at visit 1 prior to DS allocation. One dog had a significant increase in canine pancreatic lipase activity, while the other dog received chronic administration of cannabidiol. Furthermore, 1 dog was excluded after completion of the trial and 5 were withdrawn during the trial. For the dogs which withdrew, 2 dogs were on the MCT-DS period and 4 were on the control-DS period at time of withdrawal. Two dogs, withdrawn during the MCT-DS period, were euthanized before study completion due to uncontrollable cluster seizures and ownerdirected postictal aggression. On the other hand, 2 dogs were euthanized due to uncontrollable cluster seizures or status epilepticus during the control-DS period. One dog was withdrawn, during the control-DS period, as the owner requested to alter their ASD treatment due to uncontrollable cluster seizures. One dog was excluded from subsequent data analysis after completion of the trial as this dog had been diagnosed with pulmonary hypertension and dilatative cardiomyopathy, from which alterations in ASD clearance and elimination were expected due to chronic furosemide consumption.

| DISCUSSION
The objective of this clinical trial was to evaluate the antiseizure efficacy and tolerability of an MCT oil in dogs with IE chronically treated with standard ASDs. Twenty-eight dogs with epilepsy were included in the 6-month randomized, placebo-controlled, crossover prospective trial. The supplementation of commercially available MCT oil, containing C10 and C8, to a stable, individually variable base diet was well tolerated and decreased both seizure frequency and seizure day frequency compared to control oil (Figure 4).
Epidemiological studies suggest differences between the intrinsic severity of epilepsy among individuals linked to their response to medication and the long-term therapeutic success. In addition, individual seizure-associated alteration on pharmacodynamics and kinetics F I G U R E 3 Effects of the medium-chain triglyceride dietary supplement (MCT-DS) on ataxic gait, sedation and quality of life (QoL). Visual analog score was compared per dietary supplementation period to baseline records at enrolment visit to assess antiseizure drug-related adverse effect profiles in the course of the trial. The owner was asked to draw a secondary intersecting line perpendicular to the line of measurement that best represented the subjective severity (0 [better, than normal] to 100% [cannot be worse]). The figure shows the owner reported perception and in percentage normalized grading of each aspect illustrated in a column bar graph. The MCT-DS intake resulted in significant improvement of ataxia (P = .003), sedation (P = .002), and QoL (P = <.001) of the population (N = 28) of ASD has been discussed. 47 This might explain that in human patients only around 10% of the overall cohort are seizure free when a 2nd ASD is added to the treatment and less than 4% for any additional ASD given. 48 Some veterinary studies report an even lower seizure freedom when a 2nd or 3rd ASD is given. 49 When more than 50% reduction in seizure frequency is used as definition of ASD response, usually a third of the reported population in the literature will respond to the 2nd ASD. 8,10,50 In the current study, all 28 dogs in this trial were classified as resistant to at least 1 ASD and suffered from at least 1 seizure per month at trial start. Two dogs received 1 ASD (impetoin), while most dogs were on 2 (N = 9) or 3 drugs (N = 17) as antiseizure combination therapy. The overall reduction in seizure frequency between treatment groups was small and that only a small cohort of animals was seizure free or had more than 50% reduction in seizure frequency. Two dogs achieved seizure freedom, 3 additional dogs had ≥50% and 12 had <50% reductions in seizure frequency and 11 dogs showed no change or an increase in seizure frequency. The response to MCT treatment was better in the study reported by Law et al 24 with 14% becoming seizure free and 48% of the dogs having more than 50% reduction in seizure frequency. In both studies MCTs were very well tolerated and in the current study even an improvement is the ASD adverse-effect profile was noted.
Owner-reported improved level of sedation and ataxia and perceived is of interest to note that a fat-and protein-restricted diet resulted in increased serum alkaline phophatase activity. 57 The results presented in this study show that MCT oil and its respective metabolic constituents may also affect the absorption and excretion rates of PB. 58,59 Furthermore, it has been demonstrated that PB, when consumed with food, is not absorbed as well as in a fasting state and decreases the bioavailability. 54 MCT oil-induced perturbations on PB absorption could lead to decreased liver enzyme activity, such as alkaline phosphatase activity, and partly improved adverse effect profiles. 7,60 F I G U R E 4 Trial diet composition: the figure summarizes the distribution of the different main feeding regimes within the study population (N = 28). The in average calculated oil amount based on 9% of the ME requirement was 7.03 (SD 2.68) mL of MCT-DS or 7.01 (SD 2.69) mL of control-DS split into 2 portions a day added to the base diet. The oil amount was calculated from the patients metabolic energy requirement derived from the metabolic body weight (RER = 70 [10 kg] 3/4 , ME = RER × factor, life stages corresponding factors used to estimate daily energy needs for dogs). DF, dry food; HCF, home-cooked food; MBW, metabolic body weight; MCT-DS, medium-chain triglyceride dietary supplement; ME, metabolic energy; RER, resting energy requirement; RW, raw food; WF, wet food PB induces the expression of cytochrome P450 genes and increase liver enzyme activities, [61][62][63] and CYP152 family P450s to oxidatively decarboxylate fatty acids and induce expression levels. 64 MCT are metabolized by some CYP4 isoforms, which is a family of P450 enzymes. 65 The reduced PB concentration associated with reduced alkaline phosphatase enzyme activities might therefrom also result from MCT provoked higher cytochrome p450 activities in the liver. 66 Further in vitro studies are needed to confirm this observation and link it to potential metabolic pathways.
Although PB serum concentrations were significantly lower during MCT consumption, the relative reduction might or might not be clinically relevant for the improvements in both seizure control or owner-perceived adverse effects shown in this study. 67,68 The supplementation of MCT oil described in this trial resulted in a significant reduction of ASD adverse effects, which has not been previously  24 In a Drosophila invitro model, BHB reduces seizure-like neuronal activity, which is thought to be mediated via K ATP channel activity and GABA B signaling inhibition. 69 In a kainic acid rat seizure model intraperitoneal BHB administration prolonged the onset of induced seizures. In addition, neuronal cell death was alleviated and linked to protective effects. 70 More recent in vitro studies also suggest that BHB exerts beneficial neuroprotective functions by way of gene regulation and promotion of brain-derived neurotrophic factor expression. 71 Other than seizure suppressing effects, elevated blood BHB concentrations have anxiolytic effects in lab rats. 72 In humans receiving a ketogenic diet (KD), the increase in blood BHB concentrations and other ketone bodies is considered to play a crucial role for improved seizure control. 28,30,73 Here, the majority of dogs received over 50% of their daily calorie intake from carbohydrates (nitrogen-free extract: 56.59%; from cereals, potatoes, or other vegetables), fat (11.75%), and protein (21.8%). This macronutrient profile is in contradiction to the common KDs in human medicine which are usually of a high-fat, lowcarbohydrate composition. 74 Therefrom, the results of this study demonstrate also that MCT supplementation with a non-KD base diet can enhance BHB production. However, the dogs did not become ketotic, as observed in humans under KD diet. Accordingly, it might be suggestive that the rapid onset of antiseizure properties in dogs consuming the MCT-DS might be governed more by the direct effect of certain MCTs, rather than the downstream metabolites of BHB.
The MCT-enriched kibble diet reported by Law et al in 2015 contained medium-chain fatty acids C8, C10, and C12 which was the only compositional difference to the control diet. 24 C10 not only acts as a noncompetitive AMPA receptor antagonist resulting in direct inhibition of excitatory neurotransmission, and thus exerts an anticonvulsant effect, 75 but also has synergistic effects with a newly introduced ASD in human medicine. 76 Furthermore, C10 on its own might be able to modulate mitochondrial proliferation via PPARγ receptor 77 and improve brain mitochondrial energy homeostasis in a dose-dependent manner. 78 An enhanced anticonvulsant effect is observed in mice under seizure provocation by intravenous administrated pentylenetetrazol, when C10 and C8 was orally coadministered. 79 In multiple vitro seizure and status epilepticus models, the inactive unbranched compound C8 does not to play a direct role in provoking antiepileptic properties. 80,81 However, methyl-branched derivatives of C8 enhance seizure control and neuroprotection. 82 Thus far, the available scientific evidence indicates that C10 may be more relevant than C8 with regard to active interaction with cellular targets that can explain antiseizure effects from MCT consumption. 31,83,84 The dogs' baseline diets in this study have not been controlled to ensure a balanced nutrient supply as they varied between the trial participants and therefore must be considered a limitation of this diet trial study. In addition, the original MCT content cannot be evaluated in the baseline diet. However, the dietary access to MCTs is limited in the usual diet of dogs. Average sources used in production of commercial foods or the composition of a home-made or raw-feed normally contain relatively low concentrations or no MCTs. 85,86 In commercial pet food, MCTs are typically not added to diets due to reports on decreased food palatability. 87 Therefrom, a negligible impact on the results of the baseline on this trial could be expected.
Likewise, because MCTs are so rarely used in commercial pet food, its supplementation or integration may easily offer more options to manage chronic neurological diseases and should be explored further in the future.
The oral tolerability of MCTs in dogs has been widely explored with controversial findings and the occurrence of gastrointestinal signs or palatability issues were more likely. [87][88][89][90][91] The percentage of metabolic energy chosen as amount of MCT-DS was orientated on the previous study from Law et al in 2015, but also tolerance and palatability findings from other studies, [87][88][89][90][91] but also recommendation to meet nutritional guidelines established by the Association of American Feed Control Officials and the National Research Council. 92 In this trial, adverse events have not been found in higher frequency or predominantly associated to 1 type of DS in this trial. Thus, the seen adverse events were considered as not provoked by the dietary interventions of this study and as naturally occurring diseases.
One relevant limitation of this clinical study could be the study center effect. However, as the number in each of the centers was rather small and each dog experienced at each visit the same scenario, a center effect was expected to be small and was not identified on statistical analysis. We were also not able to identify a treatment order bias. In addition, the findings of the current trial confirm earlier findings in a single center study. 24 Another limitation of the analysis was that an intention to treat analysis (ITTA) was not conducted. Normally, ITTA is an analysis approach in which all recruited patients are evaluated as block randomized, regardless of the dietary intervention they actually received. 93 However, as in randomized clinical trial with epileptic patients, the dropout rate appears to be higher, especially in the beginning, all of the 8 dogs dropped out in the first leg of the study. Therefore, an ITTA would not be appropriate.
In summary, this trial provides additional evidence for the use of MCT oil as a DS to a stable base diet as a potential therapeutic management option for dogs with epilepsy. Although the overall significant statistical effect was small, some dogs had a good response.

CONFLICT OF INTEREST DECLARATION
Authors declare no conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
The study protocol and design were approved by the Clinical Research Ethical Review Board (CRERB) and ethical approval has been granted (URN 2016 1558). The data collected in this trial are collated and stored at the Royal Veterinary College in London (RVC).
Data were anonymized as appropriate, and only used for analysis. All dog's personal information was held and used in accordance with the GDPR 2018 and will not be disclosed to any unauthorized person or body.

HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.