Anticonvulsant therapy: the new and the old (Proceedings)


Successful control of seizures with anticonvulsant drugs reflects a balance in achieving seizure control while minimizing undesirable drug side effects. Variability in the disposition of anticonvulsants and interactions among them and other drugs are important confounders of successful therapy.

Successful control of seizures with anticonvulsant drugs reflects a balance in achieving seizure control while minimizing undesirable drug side effects. Variability in the disposition of anticonvulsants and interactions among them and other drugs are important confounders of successful therapy. This chapter reviews selected anticonvulsants, focusing on drugs most likely to control seizures in small animals. The proper use of anticonvulsants is discussed, with an emphasis on the differences in individual drug disposition, detection of these differences, and rational approaches to responding to these differences by dose modification. The primary topic of discussion is treatment of generalized, tonic-clonic seizures, the most common type afflicting small animals. Opinions regarding anticonvulsant therapy vary among clinicians. Most of the comments and recommendations offered in this discussion reflect personal observations in a therapeutic drug monitoring service and completed as well as ongoing clinical trials that focus on the use of anticonvulsants either alone or in combination with phenobarbital.

It is important to approach epilepsy as a clinical manifestation of an underlying disease. Thus, therapy is more likely to be effective if the underlying disease is treated. Such causes should be identified and appropriately treated, and if possible, before chronic anticonvulsant therapy is instituted. Undesirable side effects are often the limiting factor in the use of anticonvulsant drugs, and not all seizures necessarily need to be treated.

Certainly immediate, short-term anticonvulsant therapy is indicated for status epilepticus (see later definition) or cluster seizures. Chronic therapy is generally indicated for seizures that last more than 3 minutes, cluster seizures (for which there is no delineable interictal period), or seizures that occur more frequently than once a month. Seizures that are not sufficiently controlled can lead to additional seizuring (kindling) or to the development of a second "mirror" focus of seizure activity. This might be manifested as a decreasing interictal period or a worsening of seizure activity (including duration). Ideally, monotherapy is preferred to combination therapy; monotherapy should not be considered to have failed until either undesirable side effects emerge, or drug concentrations at the maximum acceptable range have been surpassed. Monitoring is indicated to confirm determine therapeutic failure, potential toxicity and to establish baseline concentrations such that proactive changes in anticonvulsants can be recognized. Monitoring at Auburn University can be found at

THE OLD: Bromide.

Bromide continues to be a drug of choice for first choice or combination anticonvulsant therapy in dogs. For rapid response, a loading dose can be given but it is important to match the loading dose with an appropriate maintenance dose, otherwise, drug concentrations will slowly decline over 2-3 months (with the majority of the decline in 15-21 days). A load of 450 mg/kg should yield concentrations of 1 mg/ml (the minimum end of the therapeutic range); for each 0.5 mg/ml increase in blood concentrations desired (maxiumum end of the range is 3.5 mg/ml), an additional 225-250 mg/kg loading dose should be given. If this loading dose is split over 5 days, the maintenance dose should also be given (30 mg/kg/day for 1 mg/ml; 15 mg/kg/day for each 0.25 to 0.5 mg/ml increase above that). Patients should be monitored 1 to 3 days after the loading dose and again at one month; if the two samples do not match, the maintenance dose should be changed accordingly. Our lab will increase bromide concentrations well above the recommended range if necessary to control seizures as long as the animal is not groggy or otherwise is intolerant to the drug. If groggy, our choice is to decrease phenobarbital concentrations first. Phenobarbital can be completely eradicated in some animals; in contrast, some animals will be controlled only at concentrations of both bromide and phenobarbital at the maximum end of the therapeutic range. Bromide can be made by dividing a 1kg bottle of the salt into 4 equal 250 gm parts (store in zip lock back, protect from humidity).. One package can be added to a 1 liter bottle of commercial spring water.:draw a line at the 1 liter mark, remove about 0.5 liter, add the bromide, and enough water to make the bromide dissolve, and fill the remaining volume to the line with either water or corn syrup for flavoring. The final solution is 250 mg/ml. Potassium bromide can be loaded following rectal administration over a 24 hour period (divide the loading dose into 4 administrations). IV administration is not recommended because of the risk of potassium overload.

Bromide has been studied in cats (as the potassium salt) when used at the canine maintenance dose. Although concentrations are similar to those achieved in dogs, in a retrospective study of 17 cats, 38% of seizuring cats developed signs consistent with feline bronchial asthma. The time to onset varied from 3 to 24 months and did not seem to be related to dose. Treatment with glucocorticoids may be helpful. Combination anticonvulsant therapy is a powerful tool for control of refractory seizures (defined as unacceptable seizure activity despite anticonvulsant concentrations at the maximum end of the therapeutic range). Several options exist. We have studied bromide as an add-on anticonvulsant and found it to be effective in eradicating seizures in 60% of dogs refractory to phenobarbital. Bromide is also effective as a sole anticonvulsant. Efficacy and safety of bromide (BR) were compared to phenobarbital (PB) in 46 dogs with spontaneous epilepsy using a parallel, randomized double blinded study design. Acceptance was based on seizure history, physical and neurologic examinations and clinical pathology. Dogs were loaded over a 7 day period to achieve the minimum end of the therapeutic range of the assigned drug. PB (3.5 mg/kg) or BR (15 mg/kg) was administered every 12 hours. Data (clinical pathology and drug concentrations) were measured at baseline and at 30 days intervals for 6 months. All but 3 patients completed the study. Seizures initially worsened in 3 dogs on BR but not in any PB patient. Mean seizure number, frequency and severity were reduced at 6 months compared to baseline for both drugs; seizure duration was shorter for PB but not BR. Seizure activity was eradicated in a greater percent of PB (85%) compared to BR (65%) patients, but successful control (at least 50% reduction in seizure number) did not differ between drugs at 6 months. Mean bid dose and drug concentrations were dose 4.1 ± 1.1 mg/kg and 27 ± 6 µg/ml, respectively for PB and 31 ± 11 mg/kg and 1.9± 0.6 mg/ml for BR. Both drugs caused abnormal behaviors. Weight increased by 10% in both groups. Changes in clinical pathology were limited to increased (but within normal) serum alkaline phosphatase and decreased (but within normal) serum albumin at 6 months for PB compared to baseline and compared to BR at 6 months. Side effects at one and six months, respectively for each drug were: ataxia (PB: 55 vs 5%;BR: 22 vs 9%), grogginess (PB: 50 vs 5%; BR: 35 vs 13%), polydypsia (PB: 40 vs 0%; BR: 39 vs 4%), polyuria (PB: 35 vs 0%; BR: 13 vs 0%), hyperactivity (PB: 35 vs 10%; BR 43 vs 4% [one failure]), polyphagia (PB: 30 vs 0%; BR: 43 vs 4%) and vomiting (PB: 20 vs 0%; BR: 57 vs 21% [one failure]). One PB dog failed due to neutropenia, a reported rare side effect (as is superficial necrolytic dermatitis in dogs.) The incidence of grogginess and vomiting were greater in BR compared to PB at 6 months. This study suggests that both PB and BR are reasonable first choices for control of epilepsy in dogs, although PB may provide better control. Side effects can be expected to be greater in BR following chronic dosing.


Clonazepam (Klonopin) is a benzodiazepine derivative that is more potent than diazepam and is used only in the emergency treatment of status epilepticus in the dog. Clonazepam is given intravenously in a dose of 0.05-0.2 mg/kg (the IV preparation is not available in the US). Accumulation occurs upon continued administration. However, tolerance develops due to hepatic enzyme induction within days to weeks after administration. Consequently, clonazepam, like diazepam, is unsatisfactory in long-term control of epilepsy in dogs, but may be useful in cats (0.016 mg/kd/day; target concentration 70 ng/ml). An alternative anticonvulsant in cats is clorazepate: 3.75-7.5 mg/cat q8-12h PO, which can also be used in dogs long term.

"Alternative" Drugs


Felbamate was approved in the United States in the late 1990's for treatment of human epilepsy as either the sole drug, or in combination with other anticonvulsants. Similar to meprobamate in chemistry, felbamate's mechanism of action appears to be inhibition of NMDA receptor-mediated calcium or sodium influx (inhibition of excitatory signals) as well as potentiation of GABA receptor-mediated chloride (negative) influx.141 Thus, the drug should have a broad mechanism of anticonvulsant activity with an action that might be considered complementary to phenobarbital.

The drug was proved very safe and efficacious in the treatment of partial and generalized seizures in experimental animals142,143 and humans, particularly children.144,145 Initially studied as monotherapy treatment of partial seizures, the drug has since proved useful as monotherapy for other seizures including generalized seizures.145-148 When added to Phenobarbital in refractory canine epileptics, it reduced seizure severity and numbers by.149

Felbamate is well absorbed after oral administration, although bioavailability in pediatric animals may be as little as 30% of that in adult dogs, necessitating a higher dose.150 The drug is eliminated by hepatic metabolism to metabolites that are largely inactive.150 Initial studies in the dog revealed felbamate to be completely absorbed after oral administration of 16 and 1000 mg/kg, At each dose, Cmax was 12.6 to 168.4 µg/ml, respectively, in dogs with Tmax occurring at 3-7 hr, respectively. Plasma elimination half-life was 4.1n to 4.5 h at both doses. After multiple oral doses of 50 mg/kg, plasma concentrations did not appear to change, and much (at least 50%) of the drug (based on [14C] felbamate) was eliminated in the urine (58-87.7%), with at least 7% eliminated in feces and the remainder in bile.151 Volume of distribution was 0.72 L/kg and binding to plasma proteins was at most 36%. Plasma clearance was 108; renal clearance of unchanged drug was between 20 and 35% and hepatic clearance due to metabolism was between 65-80% of overall clearance.151

In adult dogs, the half-life of felbamate is 4 to 8 hours (mean of 5.2 hours); the elimination half-life is shorter in pediatric (beagle) dogs (mean of 2.5 hours) probably.151,152 Felbamate also was studied in adult and pediatric dogs at 60 mg/kg orally once a day for 10 days. Oral bioavailability was less in pediatric dogs compared to adults, apparently due to more rapid clearance. Bioavailability also decreased by day 10 compared to day 1.Safety has been experimentally documented at doses ranging from 15 mg/kg divided twice daily (the starting therapeutic dose) to 300 mg/kg. Oral bioavailability is markedly variable; peak felbamate concentrations following oral administration of 60 mg/kg in dogs was 12.6 to 168 µg/mL. The oral disposition of felbamate has been described in adult and pediatric male and female dogs after single and multiple dosing. Felbamate was characterized by a lower Cmax (33 and 37 mcg/ml for male and females, respectively) and AUC and shorter half-life (2.87 ± 0.52 [M] and 2.93 + 0.33 [female] in pediatric animals compared to adults. Clearance increased to 0.98 0.11 L/h/kg), resulting in a decrease in half-life from 8.48 and 6.17 hr (male and female respectively) to 6.1 and 5.1 hr (M, F, respectively) after multiple dosing (10 days

Sedation, polyuria, polyphagia, and polydipsia, side effects typical of most anticonvulsants, do not appear to occur in dogs. Aplastic anemia due to bone marrow suppression, however, developed in 10 of 100,000 patients (human) receiving felbamate,154 leading to marked curtailment of the use of the drug in humans. Felbamate has not been sufficiently used in dogs to detect a similar side effect, although it is probably as likely to occur in dogs as in humans. Hepatotoxicy has been reported in dogs receiving both Phenobarbital and felbamate and has occurred in 3/15 patient treated with combination therapy by the author. However, we have administered felbamate to dogs also receiving Phenobarbital at doses as high as 300 mg/kg divided daily for over 6 months with no apparent initial adverse effects. However, in one of 15 dogs receiving this regimen, progressive liver disease that ultimately was fatal developed 1.5 years into the combination therapy. Because felbamate is a drug metabolized by the liver, prudence suggests that combination with phenobarbital, particularly in patients requiring high serum concentrations of phenobarbital to control seizures, be avoided. In humans, interactions have lead to increases in phenobarbital by felbamate,155 although this appears to be due to selective inhibition of a cytochrome P450 enzyme in only 25% of the population. Phenobarbital concentrations should be measured in order to detect any drug interaction that might lead to an increase in drug concentrations and a subsequent increased risk of liver disease. Greater risk may reflect decreased felbamate concentrations due to induction by phenobarbital.155 Currently, there is no easy, cost-effective means for assaying felbamate, although selected laboratories may offer the service. Human therapeutic concentrations range from 20 to 100 µg/ml, with trough concentrations ideally remaining in the range of 60 to 80 µg/ml for best efficacy.156


Gabapentin is an anticonvulsant approved in 1994 for treatment of partial seizures with or without generalization in humans with epilepsy.157-159 It has been used in dogs and anectodally in cats. It appears to act by a novel mechanism by promoting the release of GABA, although the actual mechanism of release is not known. Although gabapentin is absorbed well after oral administration, its absorption appears to be dose dependent, relying on a saturable transport process. This process has been cited as the reason that AED effects last longer than anticipated based on drug half-life, allowing twice daily administration. In contrast to bromide, the short half-life of gabapentin (in humans) results in steady-state concentrations within 24 to 48 hours. The drug is eliminated in people entirely by renal elimination, thus avoiding some of the risks of hepatotoxicity and drug interaction. The drug is sufficiently safe that TDM is not necessary; rather, the dose is increased as needed to control seizures. Mild dizziness, nausea, and vomiting have occurred in a small percentage of human patients.

Gabapentin studies with animals are limited.160,161 Gabapentin has been studied in the dog following oral administration of 50 mg/kg. Oral bioavailability was 80% and plasma protein binding was < 3%. Mean intravenous elimination half-life of 2.9 hr has been reported in dogs. Repeated administration did not alter gabapentin pharmacokinetics nor did gabapentin induce hepatic drug metabolizing enzymes. In the dog, 34% of the dose was metabolized to the N-methyl form. The principal route of excretion was the urine.162,163 A more recent study compared gabapentin (600 mg to Beagles) after oral administration of either an immediate or slow release product.164 The sustained release product did not disintegrate, but the release kinetics were not substantially different from the immediate release product.

The addition of gabapentin (35 to 50 mg/kg divided every 8 to 12 hr) to either phenobarbital or bromide was studied in epileptic dogs (n=17) using an open, uncontrolled design. The interictal period increased, but the number of seizures did not. However, seizures were eradicated in 3 dogs. Side effects evolving with the addition of gabapentin included sedation, which resolved within several days, and hind limb ataxia which resolved with a reduction in the bromide dose.165 One of the major disadvantages of this drug is its expense. Gabapentin may not be effective for the control of epilepsy when used at doses extrapolated from human patients. Clinical trials are indicated to establish the most appropriate dosing regimen for dogs or cats. Gabapentin is among the drugs for which status epilepticus may occur during withdrawal.166


Levetiracetam is a single (S)- enantiomer acetamide derivative AED drug. Its mechanism of action is novel and does not appear to involve any known neurotransmitter, ion channel protein or receptors. Levetiracetam was most useful experimentally in blocking seizures caused by pilocarpine and kainic acid, and in the kindling model of rats, both models for complex partial seizures with secondary generalization. Food does not impair the extent, but does impair the rate, of oral absorption. In humans, close to 70% of the drug is renally excreted; hepatic metabolism of the remainder reflect acetamide hydrolysis, which is not CYP 450 dependent. Levetiracetam is metabolized by plasma B- esterases which will continue once blood is drawn. As such, serum rather than plasma or whole blood is the desired test tissue of choice.167 The elimination half-life in humans is approximately 7 hrs. Drug interactions appear to be minimal; competition for renal tubular secretory proteins may occur.

The disposition of levetiracetam has been described in mongrel dogs (n=6).168,169 Following IV administration of 20 mg/kg yielded, a maximum concentration of approximately 44 mcg/ml, a VD of 0.45 ± 0.13 l/kg, a clearance of 1.5 ml/min/kg, and elimination half-life of 3.6 ± 0.8 hr and MRT of 5 ± 1 hr. In a separate study170 (n=2 to 8 male and female dogs), oral administration of 54 mg/kg yielded a Cmax of 50 to 65 mcg/ml (mean of 53 and 55 mcg/ml in male and female dogs, respectively), and an elimination half-life of 2 to 3 hrs. Dewey et al171 reported the disposition of levetiracetam in dogs (n=6) after IV single dose (60 mg/kg over 2 min) administration. The extrapolated Co was 254 ± 81 mg/ml, Vdss was 0.48 L/kg and CL was 1.4 ml*min/kg. The elimination half-life was 4.0 ± 0.82 hr and MRT was 6.0 ± 0.9 hr. The dose was well tolerated. Patterson172 studied levetiracetam after IV, IM and PO (19.5 - 22.6 mg/kg) administration in Hound dogs (n=6). Peak drug concentrations were 37 ± 5, 30.3 ± 3 and 30 ± 4 mcg/ml after IV (Co), IM and PO administration, respectively. The volume of distribution (beta) was 0.55 L/kg and clearance was 55 ml/min (not standardized to kg). Elimination half-life was 3 ± 0.3 hr. Bioavailability after IM and oral administration were 113+13% and 100+7%, respectively. No pain was detected with intentional perivascular injection.

In dogs, 1200 mg/kg IV or 2000 mg/kg PO were not lethal but were associated with salivation, vomiting, tachycardia and restlessness. Long-term administration (≥ 6 months) in some species was associated with enzyme induction (centrilobular hypertrophy) at 50 mg/kg/day. In the dog, 1200 mg/kg/day for 13 and 52 - weeks resulted in transient restlessness and tremor, and centrally mediated salivation and vomiting. Liver weight increased, although histopathological changes did not appear in the liver.

The disposition of levetiracetam has been described in cats (n=10) receiving 20 mg/kg either IV or PO as a single dose.173 Cats tolerated dosing well, with no significant adverse effects noted. However transient mild to moderate degree of hypersalivation occurred with oral dosing. Median peak concentration after IV dosing was 37.52 µg/mL (range, 28.05 - 51.86 µg/mL), with a median half-life of 2.86 hours (range 2.07 - 4.08 hours) and MRT of 4.57 hours (range, 3.09 - 6.0). Clearance (ml/kg/min) was 2.0 ml/kg/min (range 1.5 - 3.4 ml/kg/min) and Vdss was 0.52 L/kg (range, 0.33 - 0.64 L/kg). After oral dosing, in 7 of 10 cats, therapeutic plasma concentrations were achieved within 10 minutes and remained within the therapeutic range for at least 9 hours. Median peak concentration (Cmax) was 25.54 µg/mL (range, 13.22 - 37.11 µg/mL), Tmax was 1.67 hours (range 0.33 - 4.0 hours), T1/2 was 2.95 hours (range 1.86 - 4.63 hours) and MRT was 5.65 hours (range, 4.23 - 7.86). Mean oral bioavailablilty was 100%.

Response to levetiracetam of dogs (n=14) with refractory epilepsy was described prospectively.174 Refractoriness was based on monitoring; eligibility required that drug concentrations be in the upper quartile of the recommended range (mean 32 ± 4.6 mcg/ml). Levetiracetam was administered at 10 mg/kg orally every 8 hr; the dose was increased to 20 mg/kg tid if seizures did not decline by at least 50%. At two months, 8/10 dogs responded with seizure number reduced by 73% and number of days/month reduced by 67%. At 6 months, 6/11 dogs remained classified as responders. However, with long term follow-up, only 3 animals remained responders, suggesting that efficacy of levetiracetam declined. Drug concentrations were not measured; as such, the cause of therapeutic failure due to tolerance or worsening disease was not discriminated from declining drug concentrations.174

The use of levetiracetam in cats has been reported.173,175 Four cats with seizure disorders that were poorly controlled with PB alone were treated with oral LEV as an add-on drug at a dose regimen of 20 mg/kg body weight, q 8 h. LEV serum concentrations were within the reported therapeutic range for people (545 µg/ml) for all samples in all cats. The overall average serum LEV level for all cats was 16.5 µg/ml (range: 6.9 - 24.3 µg/ml). The median serum half-life (t ½) of elimination for LEV in cats was 5.3 h. Seizure frequency was reduced by an average of 30.5% in 3 cats, and increased by 33.3% in one cat. The results of this pilot study suggest that levetiracetam is a safe drug for cats that may provide some therapeutic benefit when used as an add-on to phenobarbital.


Developed in Japan, zonisamide (zonisamide), 3-sulfamoylmethyl- 1, 2 benzisoxazole, is a synthetic sulfonamide-based anticonvulsant approved for use in the USA in 1998 for treatment of seizures related to human epilepsy.202 The efficacy of zonisamide for treatment of human epilepsy is similar to phenobarbital and superior to other classic drugs including valproic acid and phenytoin. Its mechanism is not clear, but it appears to inhibit neuronal voltage-dependent sodium and T-type calcium channels.203,204 It also modulates the dopaminergic system and accelerates the release of γ-amino butyric acid (GABA) from the hippocampus.205,206 An additional potential advantage of zonisamide is free radical scavenging which protects against the destructive nature of radicals, especially in neuronal membranes.207 Finally, zonisamide blocks the propagation of seizures from cortex to subcortical areas of the brain. Its AED efficacy, has been described similar to phenytoin or valproic acid, thus minimally impacting normal neuronal activity. These multiple mechanisms of action may translate to improved efficacy compared to other anticonvulsant drugs.

The clinical pharmacology of zonisamide has been investigated in humans with similar characteristics in dogs.208-210 Disposition is complicated. Oral absorption tends to be rapid, complete and minimally impaired by food. After 12 hours of dosing, zonisamide concentrations in the brain are two fold that in plasma. The extent of protein binding does not limit the rapid movement into the brain. Binding of zonisamide to erythrocytes (RBC) and plasma proteins contributes to complex kinetics. Erythrocyte concentrations in whole blood tend to be twice as high as plasma and serum in humans, and is characterized by binding that is both saturable and non-saturable;211 the saturable portion may reflect binding to carbonic anhydrase in epileptic patients. Accumulation of drug in RBC is reversible, and the complex relationship between zonisamide and RBC may make therapeutic drug monitoring of plasma or serum advantageous. Metabolism of zonisamide involves both phase I and phase II hepatic metabolism with cytochrome P450 3A4 being the major isozyme and a glucuronidated compound the major metabolite.212 Enzymes CYP3A4, CYP3A5 CYP2C19 contribute to metabolism in humans.213 Renal elimination and recovery of zonisamide indicates parent drug recovery of 35%. Using radio labeled (carbon) zonisamide administered to dogs, 83% of the drug was excreted in 72 hour urine as the either the parent compound or metabolites. The remaining proportion was recovered in feces.208 The terminal half-life of zonisamide in the dog following a single oral dose (20 mg/kg) administration differed depending on the tissue studied, with the shortest being 15 hours for plasma and the longest, 42 hours for RBC.208 The longer elimination half-life allows a convenient dosing interval while minimizing dramatic fluctuations in zonisamide concentrations that might cause recurrence of seizures. Recommended therapeutic concentrations for zonisamide initially were 10-70 µg/ml, with 16.5 to 49.6 µg/ml also suggested when dosed twice daily.214 The author recommends the more commonly accepted 10 to 40 mcg/ml; however, monitoring should be based on patient need. Non-linear pharmacokinetics have been reported in some human patients, particularly with chronic dosing, resulting in disproportionate, and thus unexpected, increases in drug concentrations compared to changes in dose. In dogs undergoing toxicity studies, plasma concentrations, never reached steady state over the course of thirteen weeks of dosing at 75 mg/kg, compared to proportional steady state concentrations by week 13 at 10 to 30 mg/ kg.210

Clinical pharmacokinetics of zonisamide have been described in normal dogs (n=8); 4 male and 4 female) ranging from 3 to 4 years of age using a randomized crossover design following single intravenous (IV) and oral administration, 6.85 and 10.25 mg/kg, respectively.215 Zonisamide concentrations differed among blood compartments after single dosing, with oral maximum concentration (Cmax) being greatest in RBC (28.73 µg/ml) and least (14.36 µg/ml) in plasma. Clearance of zonisamide was 57.55 ml/hr/kg from plasma and 5.06ml/hr/kg from RBC. However, zonisamide concentrations did not differ among blood compartments at the end of multiple dosing, suggesting any blood component can be monitored. The fraction of unbound drug was 60.48 ± 13.4%. Elimination half-life in plasma was 16.4 hr in serum and 57.4 hr in RBC. Volume of distribution also differed, being greater (1L/kg) in plasma and least in (0.4 L/kg) RBC. Bioavailability was 126.8% for RBC and 189.6% for plasma. After multiple dosing (10.17 mg/kg) twice daily for 8 weeks, the accumulation ratio of zonisamide was 3.5 (plasma) and 4.3 (RBC). The resulting mean Cmax at steady-state was 56 ± 12 µg/ml, suggesting a beginning dose of 2 to 3 mg/kg twice daily to target the low end of the therapeutic range. The half-life at 8 weeks was 23 ± 6 hrs. Plasma drug concentrations varied by 17.2% between 12 hour dosing intervals, suggesting a 12 hr dosing interval is appropriate. Differences in clinical pathology data occurred at the end of the 8 week study period, although all results remained within normal limits. Serum alkaline phosphatase and calcium increased above baseline, where as total serum protein and albumin both decreased below baseline.

Zonisamide pharmacokinetics have been described in cats (n=5) following a single 10 mg/kg dose (Table 23.4).216 Safety and adverse reactions were studied during chronic (9 weeks) dosing at 20 mg/kg once daily. Zonisamide was not well tolerated at this dose; 50% of cats exhibited vomiting, diarrhea and anorexia. Mean peak and trough concentration with chronic dosing in all cats were 46 and 59 mcg/ml, respectively, with concentration at 42, 59 and79 in cats with adversities. Zonisamide appears to be minimally involved in drug interactions typical highly protein-bound drugs.208 However, it is involved with interactions involving CYP enzymes. Nakasa213 demonstrated clearance was decreased 31%, 23% and 17% by ketoconazole, cyclosporine A and miconazole, respectively; fluconazole inhibited clearance to a lesser degree but itraconazole appeared to have no effect.

Zonisamide does not appear to affect its own metabolism nor the metabolism of other drugs in animals or humans. Phenobarbital will shorten zonisamide half-life. The impact of 35 days of dosing phenobarbital on the disposition of zonisamide was studied in dogs. Unfortunately, all data was pictorially represented, limiting assessment of changes in disposition.217 After 35 days administration of phenobarbital, concentrations appeared to decrease to about 2.75 mcg/ml, returning to 3.5 only after approximately 12 weeks after phenobarbital was discontinued. The decrease in half-life appeared to approximate 3 hrs or approximately 30%. Phenobarbital shortened the half-life of zonisamide from 27 to 36 hours in humans, resulting in lower plasma drug concentrations.218 The impact of phenobarbital does not appear to be profound, but monitoring is warranted and for some patients, collection of both a peak and trough sample might be warranted in patients receiving phenobarbital with zonisamide.

As a sulfonamide, zonisamide inhibits thyroid synthesis of thyroid hormones. Anticonvulsants (phenytoin) may also have a direct negative effect on TSH response to TRH. Drug-induced changes in T4-binding globulins have also been documented in human patients taking anticonvulsants. Boothe215 demonstrated that zonisamide dosed for 8 weeks was associated with a decrease in total T4 below normal limits. Free T4 and TSH were also decreased from pre treatment concentrations, although both were within normal limits. Zonisamide concentrations were higher than the recommended therapeutic range. Thyroxin and TSH concentrations might facilitate diagnosis of hypothyroidism in animals receiving zonisamide.196 Note that thyroid supplementation suppresses response to TSH, and testing should not be performed until supplementation has been discontinued for 4 to 6 weeks.

Clinical reports of zonisamide use in animals are limited. In one report, zonisamide was effective in reduction of seizures in patients with epilepsy that had not sufficiently responded to one or more anticonvulsants (including phenobarbital and/or bromide) in 7/12 dogs at doses designed to achieve 10 to 40 µg/ml. Dose reduction or discontinuation of concurrent anticonvulsant was possible in 8/12 dogs. Mean concentrations approximated 20 µg/ml; mean dose was 9 mg/kg every 12 hrs.219 A second open clinical trial studied zonisamide for treatment of refractory seizures in dogs (n=13).220 Mean reduction in seizure was 70%, with three dogs relapsing. Drug concentrations were not measured.


Therapeutic drug monitoring and the design of the dosing regimen

The relationship between the dosing interval of an AED and the rate of elimination affects therapeutic success. The rate of elimination is reflected in the drug elimination half-life (t1/2 ), the time necessary for 50% of the drug to be eliminated at steady state. The dosing interval is generally based on an acceptable level of fluctuation between peak and trough PDC during the interval. Large fluctuations are generally undesirable in seizuring patients because PDC are more likely to reach both toxic (peak, leading to sedation) and subtherapeutic (trough, leading to breakthrough seizure) concentrations during the dosing interval. Fluctuation can be minimized if the ratio of the dosing interval to t1/2is small (<0.5) because less drug will be eliminated between doses. An example is PB, which has a t1/2of about 72 hours but is dosed every 12 hours. Peak and trough concentrations vary little during a 12-hour dosing interval. As therapy is begun with such a drug, however, because little of each dose is eliminated before the next dose, PDC accumulate with each subsequent dose until the amount eliminated during each interval is replaced by the dose. At that point, a steady-state equilibrium has been reached (three to five drug half-lives). Peak PB concentrations at steady state are higher than peak concentrations after the first dose because the final amount in the body reflects drug accumulation over several weeks. Thus, each daily dose contributes little to the total amount of drug in the animal. Conversely, if the ratio of the dosing interval to t1/2 is 1 or more, most (at least 50%) of each dose is eliminated between doses, and fluctuation between peak and trough concentrations during a dosing interval can be dramatic. The drug does not markedly accumulate, however, because most of the drug is eliminated before the next dose is given, and peak concentrations at steady state are very similar to trough concentrations after the first dose. With this dosing regimen, the total amount of drug in the body is provided with each dose.

The clinical sequelae of a large ratio of dosing interval to t1/2 are many (e.g., a 12-hour dosing interval for a drug whose half-life is 3 hours). First, adding an "extra dose" may be of benefit when the patient suffers breakthrough seizures. Second, shortening the dosing interval (i.e., from 12 to 8 hours) may be indicated for an animal suffering from breakthrough seizures. Third, both peak and trough samples are recommended for monitoring to detect wide fluctuations in drug concentrations. For such drugs, if only a peak concentration is collected and the peak is too high, lowering the dose may result in subtherapeutic trough concentrations. Collection of a trough sample only may result in dose modifications that cause the peak to exceed the maximum recommended. Collection of both a peak and a trough sample will allow calculation of the drug half-life and thus design of a proper dose and interval (see Chapter 4). Fourth, response to therapy can be evaluated following one seizure interval, since drug accumulation does not occur and steady-state equilibrium is never truly reached. In contrast, if a drug is administered at a dosing interval that is substantially shorter than the t1/2 , adding an extra dose should not change PDC substantially. Rather, several doses (a "mini" loading dose) will have to be given in order to change PDC. Likewise, administering the dose at a more frequent (shorter) interval is not likely to improve response to therapy. Because PDC do not change substantially during a dosing interval, a single sample (trough) should be sufficient for monitoring. Finally, response to therapy can be evaluated only at steady state (3 to 5 drug half-lives, regardless of the drug) plus one seizure interval after the new dosing regimen has been started.

Regardless of the ratio between dosing interval and t1/2 , 87% of steady-state concentrations occur at 3 t1/2 , and 97% occur at 5 t1/2. Drug efficacy and safety of an AED should not be evaluated until at least 87% of steady-state concentrations have been reached . The duration of the seizuring interval for the animal must be added to the time to steady state before efficacy of a selected dosing regimen or drug can be evaluated. Obviously, with a drug that accumulates, the time to reach steady-state concentration can be several days to several weeks, depending on drug half-life.

Loading dose.

The time to steady state and therapeutic response may be unacceptably long for patients suffering from severe, life-threatening seizures. For such patients, a loading dose can be administered to achieve therapeutic AED concentrations immediately or within the first few days of therapy. In brief, the loading dose is a sum of all the daily doses that would have been administered before steady state minus any drug that would have been eliminated from the body during that time period. The major disadvantage of a loading dose is the sudden effect of therapeutic concentrations in the CNS; there is no time for adaptation to occur, and adverse effects (sedation, ataxia) are more likely than with gradual increases in drug concentrations. The maintenance AED dose that follows the loading dose is designed to replace drug eliminated during a dosing interval, thus maintaining therapeutic concentrations achieved by the loading dose. Both the loading and maintenance doses are, however, based on population disposition parameters. Yet individual differences in drug elimination may result in therapeutic failure with either dosing regimen.

Therapeutic drug monitoring (TDM).

TDM can and should be used to document changes in drug disposition and to guide modification of the maintenance dose. TDM can be a powerful tool for controlling the difficult epileptic animal. Unfortunately, not all AED drugs can be monitored. A therapeutic range must be established for the drug, and response must correlate with PDC. An easy, cost-effective assay that requires minimal sample handling must be available. Among the AEDs to be discussed, automated assays are available for PB (also used for primidone) and the benzodiazepines. KBR can also be assayed, although the tedium of the assay limits the number of laboratories that offer this service.

Monitoring with a loading dose.

Generally, monitoring is recommended after the loading dose (eg, KBR) is completed, and to ensure that the maintenance dose is correct, at one drug elimination half-life; and, to establish baseline for the patient, at steady-state. For a patient receiving a loading dose, the one half-life sample is selected because, if the maintenance dose does not maintain what is achieved with the loading dose, the majority of the change as a new steady state is reached will occur during the first half-life of the drug. If the patient is loaded, the one month sample can be compared to the immediate post-load sample; if the two are comparable, the maintenance dose is maintaining what the loading dose achieved. If the 1 month-post load sample is not within 10% of the immediate post-load sample, the dose should be increased accordingly. Otherwise, drug concentrations will continue to decline until steady-state. In such cases, an patient previously controlled (after loading) may start to seizure 1 to 2 half-lives into maintenance therapy. For KBR, with a half-life of 21 days in dogs, the patient may start seizuring at 2 months into the maintenance dose. The seizures may be interpreted as refractory epilepsy, when in fact, the failure reflects declining drug concentrations.

Monitoring with a maintenance dose.

For most drugs, a baseline can be established at steady-state. However, for drugs with a very long half-life, such as KBR, the patient does not reach steady-state for 3 months. However, a proactive approach can be taken in evaluating the appropriateness of the maintenance dose by measuring drug concentrations at 1 half-life (3 weeks). PDC will be approximately 50% of what they will be at steady-state and the dose can be modified at 1 month accordingly, and checked again 1 month later and at 3 months (new steady-state). For PB, monitoring might occur at 1 month and again at 3 months in order to detect the impact of induction on drug concentrations.

Peak versus trough samples.

Monitoring is indicated whenever the dose is changed (at the new steady state) or when the patient seizures. If drug half-life is substantially longer than the dosing interval, drug concentrations will not change much during the dosing interval, peak and trough concentrations will be very similar, and a single trough sample collected just before a dose generally is sufficient for monitoring. Thus, for KBR and in patients for whom PB is characterized by a long half-life (i.e., induction has been minimal), a single trough sample is sufficient. We recommend a trough rather than a peak sample because of less variability at this time compared with peak times. If a patient is on both PB and KBR, both can be measured in the same sample. On the other hand, if the dosing interval is equal to or greater than 50% of the drug half-life, drug concentrations may fluctuate 25% or more during the dosing interval and both peak and trough samples might be collected, particularly in those animals for which control is difficult. Note that induction by PB cannot be detected without a peak and a trough sample (a sort of catch 22: one must collect both a peak and a trough sample to make sure that subsequent monitoring can consist of a trough alone). For peak samples, we recommend collection at 4 to 5 hours after the dose, followed by a trough sample just before the next dose. A less accurate but more convenient schedule is collection of a trough sample before the morning dose and a peak sample 5 hours after the dose. The time of sample collection in relationship to dose administered before both peak and trough sample collection must be known if the half-life is to be calculated. We recommend that animals fast before samples are collected.

Interpretation of TDM results can be facilitated by a clinical pharmacologist. Therapeutic failure should not be considered if a patient is seizuring simply because the therapeutic range has been achieved. Monitoring should be used to identify the therapeutic range for the individual patient. Thus, for a patient that is not sufficiently controlled, doses can be increased gradually until the maximum range has been reached and the risk of adverse affects becomes too great. We use a "stair-step" approach to dose modification as delineated in the algorithms. For example, we increase PB by 5 µg/mL in patients that have not sufficiently responded. Generally, this requires an approximate increase in dose by 25%. After steady-state concentrations have been reached plus one seizure interval, the patient is reassessed. The incremental increase is continued until either control is sufficient or the patient becomes unacceptably groggy. For KBR, "mini" loading doses are administered if rapid control is indicated. We increase drug concentrations in 0.5-mg/mL increments by administering 250 mg/kg over a 2- to 3-day period. This is added to the maintenance dose, which also should be increased by 20% to 30%. Monitoring is recommended after administration of the loading dose and 3 to 4 weeks later (the latter sample ensures that the maintenance dose is maintaining what the loading dose accomplished). When KBR is added to PB and the patient becomes groggy, the PB dose is decreased by 25% once appropriate KBR concentrations have been documented.

Special handling preparation is not generally necessary for TDM, although the laboratory should be called to confirm special handling procedures. Serum separator tubes should, however, be avoided. Serum separator tubes contain silicon, which may bind AED drugs. Either these tubes should not be used or serum should be withdrawn from the tube immediately after centrifugation.

Acute Seizures & Status Epilepticus

The leading cause of status epilepticus (SE) in humans is inappropriate low concentrations, with non-compliance with prescribed medications among the most common reasons. Accordingly, the application of the principles of pharmacology to anticonvulsant therapy should facilitate effective control and the avoidance of SE. A number of pro-epileptic drugs are also cited as cause for SE (see below).221

Early therapeutic intervention is more important than drug choice; in humans, intervention within the first 30 minutes was associated with an 80% response to first choice drugs whereas 60% of patients in SE for more than 2 hrs did not respond to first-line therapy. Physiologic responses of concern include fever, cardiac arrhythmias, changes in systemic (initially hypertension and later hypotension) and pulmonary blood pressures and altered blood chemistries.221

In humans,221 the traditional definition of SE as continuous or repetitive seizures lasting 20 minutes or more has been challenged; therapy for SE is recommended for as two or more generalized convulsions without full recovery of consciousness between seizures, or continuous convulsive activity for more than 10 minutes. However, recommendations have been reduced to 5 minutes because few seizures last this long,

Benzodiazepines tend to be more effective early but not later, whereas NMDA receptor antagonists (e.g., ketamine) tend to be effective later but not early. In humans, drug of choice and duration of SE include 6-10 min (lorazepam IV, followed by diazepam or midazolam (IN, trans basally or IM); 10-20 min (fospheyntoin); or 10-60 min (midazolam CRI or phenobarbital IV; if no response, propofol CRI or valproate, followed by the addition or alternative of pentobarbital or midazolam if no response; > 60 min, pentobarbital. Despite ketamine's potential efficacy as an anticonvulsant drug, antagonism of NMDA may result in severe neurotoxicity and its use in SE should be reserved until science supports its use.

Treatment for more than one seizure per hour is a medical emergency.49 The use of drugs for acute management of seizures is addressed in detail with individual drugs. In general, however, acute therapy of seizures (e.g., status epilepticus) is preferentially implemented with diazepam (IV bolus to effect). Diazepam has a short half-life and may have to be repeated once or twice during the first 2 hours to stabilize the dog.49 To terminate the seizures, various methods of administration have been recommended. Diazepam is recommended in an IV dose of 5 to 20 mg. Frey and Loscher42 recommended an IV dose of 0.5 to 1 mg/kg. Control by diazepam can be prolonged by continued administration as a constant IV infusion (2 to 5 mg/h of 5% dextrose; the infusion line should first be flushed with the diazepam solution to allow diazepam binding to the polyvinyl), or co-administration of phenobarbital (2 to 6 mg/kg intramuscularly to avoid respiratory or cardiac depression). Clonazepam (0.05 to 0.2 mg/kg IV) may provide AED efficacy that lasts longer (but is not necessarily any more efficacious) than diazepam. Unfortunately, an IV preparation is not available in the United States.

Alternatively, phenobarbital can be administered as the first choice (IV bolus to effect, as a loading dose). Note that for each 3 mg/kg of phenobarbital given IV, serum concentration increases approximately 5 µg/mL. For a patient not receiving phenobarbital at the time that therapy is begun, up to 18-mg/kg total dose (given in 3- to 6-mg/kg increments at 15- to 30-minute intervals) may be necessary to achieve the mid therapeutic range (30 µg/mL). Drug distribution of phenobarbital into the CNS may take 15 to 30 minutes. Failure to control seizures may indicate the need for pentobarbital, which is a general anesthetic, not an anticonvulsant. As such, the risk of cardiovascular or respiratory depression is great. An advantage to the use of pentobarbital, however, is its protective effects on the brain during periods of hypoxia induced by the seizure.

Alternative routes of anticonvulsant therapy might be considered for clients attempting to control life-threatening seizures without immediate access to veterinary medical assistance. Phenobarbital (5 mg/kg), diazepam, and lorazepam are partially bioavailable after rectal administration.94,95 Bromide also can be given rectally. The risk of potassium overload can be minimized by administration of the loading dose over a 12- to 24-hour period in 5- to 15-mL increments.

A retrospective study in a small number (20) of human patients with refractory epilepsy were treated with either propofol (14) or midazolam (6). For each drug, seizures were eradicated in about 65% of patients. However, overall mortality, although not statistically significant, was higher with propofol (57%) compared to midazolam (17%).222

General gas anesthesia generally should be avoided in the patient with status epilepticus because of the risk of hepatotoxicity induced by the anesthetic that may occur with prolonged therapy. If pursued, anesthetics that are minimally hepatotoxic should be selected. Discontinuation of therapy should be undertaken cautiously to minimize the risk of seizures. Propofol and etomidate are two chemical restraining agents that, although expensive, are characterized by anticonvulsant effects. Of the two, etomidate (a human drug only) may be characterized by CNS protective effects. These drugs can be administered as IV infusions to effect (see later discussion under brain trauma).

Diazepam is used in the cat to control epileptic disorders regardless of etiology.117 Generally, an IV dose (5 to 10 mg) is given to effect. A dose as high as 20 mg may be necessary; if high dosages are used, they must be injected slowly. The procedure commonly followed is to administer 2 to 10 mg intravenously and then wait 10 minutes. In the event seizures persist, Kay117 has recommended IV administration of phenobarbital sodium (5 to 60 mg). Caution must be taken not to over sedate or depress the animal when these drugs are administered close together. Should the animal manifest refractoriness to diazepam and phenobarbital as in status epilepticus, pentobarbital anesthesia is then carefully administered to effect.

Status epilepticus may require management of cerebral edema (see later discussion of brain trauma and injury).

Chronic Control of Seizures

Use of selected AED drugs to control seizures has been discussed with individual drugs. What constitutes successful anticonvulsant therapy will vary among clinicians and may be defined as client satisfaction. Eradication of seizures may be an unachievable goal; decreased frequency, severity, or duration of the seizure episode may be considered a success for many animals. Indeed, close counseling of clients and reorientation to what constitutes successful control may be important techniques to successfully treating an epileptic dog. Chronic control of epilepsy is likely to be a balancing act for many patients: controlling seizures without putting patient health at excessive risk. Because patients are likely to received drug therapy life long, establishing a minimum effective dose for any anticonvulsant drug is prudent. Monitoring clearly is important to avoid "toxic" concentrations. However, monitoring, even for newer, safe drugs, is equally important. Monitoring should be used to determine the minimum effective dose for each patient. In the case of break-through seizures, monitoring can determine whether or not the seizures reflect an decrease in drug concentrations (leading to confirming owner compliance, addressing potential drug or diet interactions etc), or a change in the underlying pathophysiology (leading to further diagnostics and potential treatment).

If manipulation of a dosing regimen is the focus of successful control, a chosen therapeutic regimen should not be abandoned until steady-state plasma drug concentrations have been reached. Thus, an animal should not be considered refractory to a drug simply because it is receiving more than the recommended dose or its serum concentrations are within the therapeutic range. A drug should not be abandoned until serum concentrations in the maximum therapeutic range (and, in some circumstances, exceeded if the drug is sufficient safe) have been documented or unacceptable adverse side effects occur. Regardless of the anticonvulsant used, therapy should never be stopped suddenly, and drug concentrations should not be allowed to drop precipitously during a dosing interval. Status epilepticus may occur. Cautious exceptions might be made for drugs with a very long half-life (e.g., bromide) which naturally gradually decline (as long as chloride content in the animal has not increased).

The thyroid and liver status of patients should be determined prior to initiation of therapy. Anticonvulsant-induced liver disease should be distinguished from hepatic induction induced by several anticonvulsant drugs; unnecessary discontinuation of a drug that is controlling seizures might thus be avoided. Moderate elevations in the serum transaminases and serum alkaline phosphatase activity and abnormalities (more than 50 mol/L) in fasting bile acids and serum albumin are indicative of hepatic pathology. The incidence of serious liver toxicity can be reduced by avoiding combination therapy with more than one drug metabolized by the liver, using TDM223 (see Chapter 4) to achieve adequate serum concentrations at the smallest dose possible, and evaluating hepatic function every 6 months or more, depending on the magnitude of phenobarbital serum concentrations. The higher the plasma drug concentration, the more important hepatic monitoring becomes. Seizure-induced hypoxia can result in liver damage; thus, evaluation of the liver should not occur in association with a seizure episode. Hepatotoxicity induced by anticonvulsants is often reversible if the drug dose is sufficiently decreased before cirrhotic changes occur.

Phenobarbital has remained the first-choice anticonvulsant for chronic control of seizures in both dogs and cats due to its efficacy and, as long as drug concentrations do not approach the maximum therapeutic concentration, safety. Therapeutic drug monitoring should be used to ensure that adequate serum drug concentrations have been achieved before the patient is considered refractory. As concentrations of phenobarbital approach the maximum end of the therapeutic range, an alternative regimen should be considered. The addition of a second anticonvulsant is the most likely next step.

Combination Therapy.

Use of combination therapy appears to be popular in veterinary medicine, based on therapeutic drug monitoring information in the author's laboratory. Although combination therapy is a reasonable approach for control of seizures in patients that fail to reasonably respond to first choice anticonvulsants (e.g., plasma drug concentrations approach or enter the high end of the therapeutic range, or unacceptable side effects emerge), many of these patients are on two or more, drugs each of which is in the sub to low therapeutic range, The American Epilepsy Society notes that most humans can be controlled with single drug therapy and that higher concentrations of a single drug is preferred to lower concentrations of multiple drugs. Single therapy should be considered prudent for several reasons. The most obvious is avoidance of side effects (the combined side effects of a drug might, like efficacy, be worse than either drug by itself), fewer drug interactions, better owner compliance and reduced cost due to the need for more than one prescription and monitoring more than one drug. Other reasons to limit combination therapy to patients with proven need are likely to be less obvious. However, no drug therapy is likely to be innocuous. Drugs that affect the CNS may be problematic because of the sophisticated mechanisms which exist to minimize the effects of CNS drugs. This includes efflux proteins, receptor down regulation or desensitization. In the author's opinion, because the CNS does not want drugs in the CNS, an attempt should be made to respect the body's attempt to limit exposure of the brain to drugs. Accordingly, the author recommends that single drug therapy be targeted and combination therapy be instituted only in patients that have failed initial therapy.

Bromide increasingly is recommended as a first choice AED drug. Although Boothe28 demonstrated that it is not as efficacious as phenobarbital for control of seizures in dogs, its lack of drug interactions and improved safety compared to phenobarbital warrant consideration as first choice particularly in older dogs for which the risk of drug interactions and liver disease might be decreased.

Bromide also has been recommended as the first combination drug of choice for dogs should phenobarbital therapy fail. However, as newer renally-excreted drugs become available in generic preparations, their use should be considered such that the gastrointestinal side effects of bromide might be avoided. Bromide increasingly is being used as the sole anticonvulsant, although a severe seizure history may warrant using a more accepted and predictable first-choice drug (i.e., phenobarbital).

Controlling refractory seizures with any anticonvulsant drug might be facilitated using a stair-step approach. Using bromide as an example, increase concentrations in 0.5-mg/mL increments. If the patient develops seizures at one concentration, the concentration is increased to the next level. This is continued until the patient is acceptably controlled or sedation becomes untenable. In the latter case, decreasing phenobarbital concentrations by 25% may help resolve grogginess. If the goal of bromide therapy is to wean the patient off an anticonvulsant (using Phenobarbital as an example), that concentrations should be confirmed prior to a stair-step decrease. For example, bromide concentrations ideally are at least 1.5 mg/mL before phenobarbital is decreased by 25%. Every time the dose of an anticonvulsant is changed, at least 3 drug half-lives, plus one seizure interval, must lapse before the impact of the dose change can be fully assessed. For Phenobarbital, at least 2 to 4 weeks should lapse before a second decrease is implemented. Ideally, the anticonvulsant is monitored at each decrease in dosage so that a target has been identified, should seizures return. Deciding what concentration to target with the second anticonvulsant can be difficult. If the goal is to simply add a second anticonvulsant, targeting in the lower therapeutic range of the new drug is reasonable. If the goal is to reduce the first anticonvulsant, the target of the second drug might be a little higher. If the goal is to eradicate the original anticonvulsant, at the very least, the second drug concentrations should be at the same level of the therapeutic range (if not higher) than the first. For example, if a patient Phenobarbital is 25 mcg/ml (mid therapeutic range), bromide should be at least 2 mg/ml before a decrease is considered. Some animals require will require higher concentrations: e.g., for bromide, higher than 2.5 mg/mL before phenobarbital can be lowered to less than 20 µg/mL (our standard goal). The first anticonvulsant may not be "decreasable" in some patients, despite the addition of a second drug at concentrations that are in the high therapeutic range. On the other hand, some patients can be totally "weaned" off the first anticonvulsant (e.g., Phenobarbital).

Diazepam has been the second drug of choice for chronic control of seizures in cats. Bromide as either the sodium or potassium salt also has been useful in cats. However, bronchial asthma may be an undesirable complication. For patients for whom bromide and phenobarbital cannot control seizures (after maximum therapeutic concentrations of both drugs have been documented) or for patients that are unacceptably groggy, a third anticonvulsant can be added. The third drug should be anticipated to provide better control or allow a decrease in one of the anticonvulsants (preferably phenobarbital) to minimize grogginess. Ideally, the third drug should be selected to minimize sedative effects and drug interactions. Generally, the addition of a renally excreted drug should be considered if the first drug is metabolized by the liver.

Discontinuing Therapy

Whether or not anticonvulsant therapy facilitates remission of spontaneous seizures is not clear, although a tendency for contemporary anticonvulsant therapy to be associated with epileptic cure has been described in humans.224 In human medicine, AED drugs can be withdrawn in 60% of patients that remain seizure free for 2 to 4 years. A similar statistic is not available in veterinary medicine. The likelihood of success can be somewhat correlated with the underlying cause or type of seizure, with the best being the patient with idiopathic generalized epilepsy in non juveniles, a normal neurologic exam, and the absence of a structural brain lesions. We have recommended that therapy might be discontinued in those patients whose drug concentrations are below the recommended therapeutic range. Should the decision be made to discontinue therapy, we recommend that concentrations first be monitored (to provide a target to which concentrations can be returned if the patient seizures) and then the antiepileptic drug be slowly discontinued over several months (e.g., 25% each month). Note that with each decrease, the response should be assessed after the drug has reached steady-state plus one seizure interval (i.e., assure, as much as possible) that the patient is challenged by a seizure before the next decrease is implemented).

Alternative Therapies

Melatonin is described as having demonstrated anticonvulsant activity in many animal models.225 A study in gerbils demonstrated greater survival in animals treated with melatonin (25 mcg SC daily). Anticonvulsant activity of melatonin may reflect antioxidant activity and subsequent free radical scavenging.226 In an open, uncontrolled study in epileptic children refractory to standard therapy, melatonin (3 mg at bedtime), seizure activity decreased and sleep improved in 5/6 patients.226

Deprenyl was associated with reduction in experimentally-induced seizures in a rat kindling seizure model rats following multiple IP dosing.227 Ldeprenyl, and to a lesser degree, D-deprenyl was effective in controlling electroshock-induced seizures in mice when administered at 1 to 40 mg/kg IP, with the highest reduction (44%) occurring at the highest dose. Although less potent, D-deprenyl also reduced seizures, but was toxic at doses higher than 10 mg/kg. PTZ induced clonic and myclonic, but not tonic seizures also were decreased by L-deprenyl at 5 mg/kg IP (but not SC); 10 mg/kg did not provide further control. The spectrum of anticonvulsant activity was described as comparable to phenobarbital (and levetiracetam) including breadth of seizure type controlled. A proposed mechanism is inhibition of norepinephrine and dopamine metabolism. Other proposed mechanisms include modulation of NMDA receptor activity and stimulation of melatonin synthesis in the pineal gland. In his review, Loscher notes that tricyclic antidepressants provide some anticonvulsant effects through inhibition of norepinephrine uptake.

The role of phenothiazines, and acepromazine in particular, in epileptic dogs is controversial. Although phenothiazines are connected with seizures in humans,228 the contraindication for phenothiazines, and specifically acepromazine, in epileptic animals is less clear. The author has personally (presumably) induced seizures in a dog suffering from lead poisoning that developed acute respiratory distress syndrome. Treatment with acepromazine was immediately followed by severe seizures. Tobias229 retrospectively studied the use of aceprompazine in epileptic dogs (n=47; 15 with idiopathic epilepsy) with seizures that received acepromazine for diagnostic testing, anesthetic premedication, to facilitate postoperative recovery, or to decrease excitatory behavior; acepromazine was given with the intent of reducing seizures in 11 of the dogs. Acepromazine also was administered to 11 of the 47 dogs in order to decrease seizure activity, with either seizures stopping for 1.5 to 8 hrs (n=8) or not recurring (n=2). Clorpromazine has been used (2 -4 mg/kg PO every 8 to 12 hr) to treat status-epilepticus, with 9/10 dogs responding (as reviewed by Tobias).229 McConnell230 also retrospectively studied the medical records of 31 dogs experiencing no seizures (n=3) or a history of acute or chronic seizures, including status epilepticus (n=3) or cluster seizures (n=22). Fifteen of 22 dogs with a history of seizures were receiving AED medication. Dogs were treated with a median of 2 but up to 5 doses of acepromazine during hospitalization; the IV dose ranged from 0.008 to 0.057 mg/kg. Seizures (n=23) occurred in 11/31 dogs during hospitalization; 15 before and 8 (n=4 dogs) after treatment with acepromazine. The dose of acepromazine in dogs that seizured after administration ranged from 0.019 to 0.036 mg/kg, with seizures occurring at 18 minutes to 10 hr after administration. It is the author's opinion that these retrospective studies support the potential use of acepromazine in epileptic dogs as adjuvant therapy - and potentially as anticonvulsant therapy but only after well designed, placebo or other controlled randomized clinical trials in epileptic patients have determined the impact in epileptic dogs or cats.


A number of medications are associated with decreased seizure threshold and an increased risk of SEL antidepressants in humans. Several anticonvulsant drugs may be proepileptic at suprapharmacologic doses (e.g., phenytoin).221 The impact of phenothiazines was discussed with alternative drugs. Drugs inducing seizures in selected patients include fluorinated quinolones, lidocaine, and possibly metoclopramide. Seizures induced by lidocaine should be treated with a benzodiazepine (e.g., diazepam).231 Morphine sulfate and related compounds as well as CNS stimulants such as the methylxanthines and behavior-modifying drugs should be avoided. Chloramphenicol also activates the CNS and should not be used in dogs known to have epileptiform seizures. Glucocorticoids may also decrease seizure threshold, although they stabilize neuronal membranes. Long-term effects on the neuronal membrane, however, may reflect down-regulation of glucocorticoid receptors and thus loss of the stabilizing effect. Long-term use of glucocorticoids might be minimized for epileptic patients. Behavior modifying drug are CNS stimulants, and, accordingly, might be associated with an increased risk of seizures. Drugs associated with seizures in non-human animal models have included but not limited to the tricyclic antidepressants, bupriopion, and doxipen. However, increased concentration of serotonin or norepinephrine by transport inhibitors are likely to be associated with anti- rather than pro-convulsant effects in epileptic patients, despite the fact that overdose may be associated with seizures.10 Finally, general, drugs for which CNS derangements or seizures are a listed side effect should be avoided in epileptic animals.

Related Videos
NAVC Gives
NAVC CEO Gene O'Neill
Senior Bernese Mountain dog
© 2024 MJH Life Sciences

All rights reserved.