Variations in serum concentration of phenobarbitone in dogs receiving regular twice daily doses in relation to the times of administration
- R. Monteiro, DVM, MRCVS,
- T. J. Anderson, BVM&S, MVM,PhD, DSAO, DipECVN, MRCVS,
- G. Innocent, BVM&S, MSc, PhD,
- N. P. Evans, BSc, PhD and
- J. Penderis, BVSc, MVM, PhD,1
- 1 Institute of Comparative Medicine, Faculty of Veterinary Medicine, University of Glasgow, Glasgow G61 1QH
The laboratory records of 1427 client-owned dogs on chronic phenobarbitone treatment were analysed. They were divided into two groups: the 918 dogs from which blood samples were collected at the trough, that is, within two hours before the next dose of phenobarbitone, and the 509 dogs from which samples were taken during the non-trough period. There were no significant differences between the mean serum concentrations of phenobarbitone in the trough and non-trough samples from dogs receiving doses ranging from 2 mg/kg per day to more than 10 mg/kg per day. However, the higher doses of phenobarbitone were associated with progressively lower phenobarbitone concentrations in the trough group relative to the non-trough group, and this difference was significant at doses of more than 10 mg/kg per day.
EPILEPSY is the most common naturally occurring chronic neurological condition in dogs and it has been estimated to affect 0·5 to 5·7 per cent of them (Bielfelt and others 1971, O’Brien 2003, Patterson and others 2005). The most common form is idiopathic epilepsy (defined as recurrent seizures with no identifiable underlying cause), but it may be caused by a variety of conditions. Irrespective of the underlying cause (with the exception of reactive seizures to certain metabolic and toxic diseases), recurrent epileptic seizures in dogs are usually treated with phenobarbitone. It is accepted that the control of seizures and the minimisation of side effects are correlated more closely with the serum concentra- tion of phenobarbitone than with the dose administered (Levitski and Trepanier 2000).
In practice the serum concentration is usually measured within two hours before the next scheduled dose, that is, when it would be expected to be at its lowest or ‘trough’ value. However, this practice causes practical difficulties for owners and veterinary surgeons. In a small study of 33 epileptic dogs on a range of doses (1·0 to 10·9 mg/kg per day, mean [sd] 5·8 [2·4] mg/kg per day), Levitski and Trepannier (2000) observed no therapeutically relevant change in the serum concentrations in most of the dogs during the 12-hour interval between doses, suggesting that the time when blood samples were taken would not be clinically relevant.
Clinically, the two most important categories of dogs in which to measure serum phenobarbitone concentrations are dogs starting treatment or in which there has recently been a change in dose, and dogs whose epilepsy is poorly controlled on the current dose or which require high doses to control seizures. In the first category the serum concentrations of phenobarbitone are measured to ensure that they are receiving adequate doses, and in the second category they are measured to ensure either that they do not have sub-therapeutic or detrimentally high concentrations or to investigate whether they may be refractory to treatment with phenobarbitone. Before seizures can be defined as refractory to treatment, it is necessary to show that the dogs have adequate serum phenobarbitone concentrations. A comparison of peak and trough concentrations may also be useful when evaluating dogs suspected to be refractory to treatment, to demonstrate significant differences between the two.
Treatment with phenobarbitone is known to induce the proliferation of smooth endoplasmatic reticulum in hepatocytes, increase the weight of the liver and induce the production of hepatic cytochrome P450 in a time-dependent manner, changes that can accelerate the metabolism of endogenous components and various drugs, including phenobarbitone, thus reducing its half-life and leading to the requirement for larger doses to maintain its serum concentration (Trepanier and others 1998, Müller and others 2000, Hojo and others 2002, Gaskill and others 2005). Phenobarbitone also induces the production of plasma α1-acid glycoprotein in a dose-dependent manner, which can potentially affect the pharmacokinetics of drugs that bind to the protein (Hojo and others 2002). One possible explanation for why some dogs require higher doses of phenobarbitone is that the half-life of the drug may have been substantially reduced by these processes. In these dogs there are likely to be wider variations in its serum concentration between successive doses.
The aims of this study were first, to assess whether the time at which blood samples are collected in relation to the administration of the drug is clinically relevant, using a large population of dogs on clinically relevant doses, and secondly, to determine whether the requirement for higher doses of phenobarbitone to maintain its serum concentration influences this relationship.
Materials and methods
The laboratory records of 1436 client-owned dogs on chronic phenobarbitone treatment for epileptic seizures were analysed. All of them had their serum phenobarbitone concentrations monitored as part of the Vétoquinol UK phenobarbitone monitoring voucher scheme, which provides a voucher for one free determination of a dog’s serum phenobarbitone concentration for every batch of the drug purchased from the manufacturer. Only the 1427 dogs that had been on continuous treatment for at least four weeks were included in the study. Data recorded from each dog included the time when a blood sample had been collected in relation to the administration of the drug, the animal’s weight, its total daily oral dose of phenobarbitone and its serum concentration as determined by solid-phase competitive chemiluminescent enzyme immunoassay (Immulite Phenobarbital Immunoassay; Siemens Medical Solutions Diagnostics). The dogs were divided into two groups: the 918 dogs from which the blood sample was collected at the trough, and the 509 dogs from which the blood sample was collected during the non-trough period. This classification was limited by the data recorded on the phenobarbitone monitoring vouchers, which did not record specific times of drug administration and blood sample collection, but rather whether the relationship between the two was known, and if so, whether the sample was collected during the two hours before the next dose (trough sample) or not during that period (non-trough sample). Dogs for which the time relationship between the two was not known were not included in the study. In order to assess whether the requirement for higher doses of phenobarbitone influences whether the timing of the blood sample in relation to the administration of the drug is clinically relevant, the dogs were further subdivided into four groups on the basis of their total daily dose of phenobarbitone: group 1, 2 to 5 mg/kg/day (the manufacturer’s recommended dose; 478 trough samples and 271 non-trough samples); group 2, more than 5 to 8 mg/kg/day (an arbitrary interval the same as the previous group; 277 trough samples and 154 non-trough samples); group 3, more than 8 to 10 mg/kg/day (77 trough samples and 44 non-trough samples); and group 4, more than 10 mg/kg/day (86 trough samples and 40 nontrough samples); the distinction between the last two groups was made on the basis of the median dose for the dogs on a dose of more than 8 mg/kg/day.
The data were analysed using the unpaired t test with Welch’s correction to account for lack of homoscedasticity. An analysis of variance was used to determine whether the four dosage groups had significantly different serum phenobarbitone concentrations. P values of less than 0·05 were considered statistically significant.
The serum phenobarbitone concentrations in the dogs in group 1 ranged between 17·9 and 181 µmol/l, in group 2 between 24·2 and 185 µmol/l, in group 3 between 30 and 176 µmol/l, and in group 4 between 24·5 and 190 µmol/l. In group 1, the mean dose of phenobarbitone for the dogs in which the sample was collected during the trough period was 3·61 mg/kg/day and for the dogs in which the sample was collected during the non-trough period it was 3·60 mg/kg/day; in group 2, the mean concentrations were 6·34 mg/kg/day and 6·38 mg/kg/day; in group 3, they were 9·02 mg/kg/day and 9·04 mg/kg/day; and in group 4, they were 13·9 mg/kg/day and 13·4 mg/kg/day. In group 1, the mean serum phenobarbitone concentrations for the trough and non-trough groups were 65·0 and 65·6 µmol/l, respectively; in group 2, they were 85·5 and 87·2 µmol/l; in group 3, they were 98·2 and 101·9 µmol/l; and in group 4, they were 114·8 and 128·7 µmol/l. The increased oral doses of phenobarbitone were positively correlated with increased serum concentrations (P<0·0001). Fig 1 shows the mean serum concentrations (relative to dose) for the dogs in which the sample was collected during the trough period and for the dogs in which the sample was collected during the non-trough period. In the dogs receiving the lower doses of phenobarbitone there was no significant difference between the serum concentrations of the trough and non-trough groups. However, as the dose of phenobarbitone increased there was a trend for the trough group to have lower serum concentrations than the non-trough group, and the difference was significant (P=0·0228) at doses of more than 10 mg/kg/day.
Phenobarbitone is considered to be an effective anticonvulsant medication in 60 to 80 per cent of dogs with epilepsy (Farnbach 1984, Schwartz-Porsche and others 1985). When it fails, the failure is more often due to the inappropriate use of the drug rather than to pharmacoresistance. In this study, two large populations of dogs were compared: one in which the blood sample for the determination of the serum concentration was collected during the trough period, and one in which the sample was collected outside this period (non-trough). By correcting for the dose of phenobarbitone, it was possible to compare the serum concentrations in the two groups at different oral doses. Levitski and Trepanier (2000) reported that in 33 epileptic dogs on a range of doses of phenobarbitone there was no therapeutically relevant change in the serum concentration in most of them during the 12 hours between doses. They compared trough and non-trough blood samples in the same dog, whereas in this study two different populations of dogs have been compared, making it possible to compare large numbers of dogs giving sufficient statistical power to investigate the effect of a wide range of dose levels.
There was no significant difference between the serum phenobarbitone concentrations of the trough and non-trough animals that were treated with the lower doses, in agreement with the study by Levitski and Trepanier (2000). However, in the dogs on the highest doses of phenobarbitone the concentration in the non-trough animals was significantly higher than in the trough animals.
One reason why some dogs require higher doses of phenobarbitone to maintain their serum concentrations is that there may have been substantial phenobarbitone-induced induction of hepatic microsomal enzyme systems and a reduction in the half-life of phenobarbitone. This reduction could result in wider variations in the serum concentrations of phenobarbitone between doses and explain the difference between the trough and non-trough animals. However, the effects of dose on absorption and protein binding, genetic differences and other factors, cannot be excluded.
The nature of the Vétoquinol phenobarbitone monitoring voucher scheme introduces several limitations into the study. The data included the length of treatment, concurrent anti-epileptic agents and the treatment regimen, but details of the underlying diagnosis, concurrent disease, and information about the doses of other anti-epileptic agents and other drugs that were administered and might have affected the expression of cytochrome P450 and thereby the half-life of phenobarbitone were not recorded. However, the large numbers of dogs in the study would have reduced the influence of these factors. The use of two populations of dogs also introduced between-dog variation; using paired serum samples from the same dogs would have been an alternative approach, but one which could not be justified ethically or on a cost basis, in the light of the study by Levitski and Trepanier (2000). The recording of the time when the blood sample was taken relative to the time when the dose was administered was also limited. The inclusion of only those cases where this relationship was known gives precise information for the trough group, but the blood samples could have been collected from the non-trough group between 0 and 10 hours after the drug was administered. Any of these samples collected within a few minutes of the drug being administered could be essentially trough samples. However, the effect of this lack of precision would have been to underestimate, rather than overestimate, any difference between the serum phenobarbitone concentrations of the two groups. The difference in serum concentrations during the 12-hour interval, between the trough and non-trough groups on their higher doses might therefore have been larger than that observed, with the potential that the effect could become clinically significant at lower doses of phenobarbitone.
In the dogs receiving regular total daily doses of more than 10 mg/ kg phenobarbitone, the mean serum concentration of the dogs from which a trough sample was collected was significantly lower than that in the dogs from which a non-trough sample was collected. Accuracy in measuring its serum concentration is essential for assessing the therapeutic effect of phenobarbitone and for making accurate adjustments to the dose. In the literature, trough samples are commonly used as the basis for determining the therapeutic ranges for serum phenobarbitone concentrations. However, it may not be necessary to collect a trough sample from every dog on high doses of phenobarbitone. When large oral doses of phenobarbitone are administered, consistency in the time when the blood samples are collected in relation to when the drug is administered is more important, so that the concentration of phenobarbitone can be assessed accurately and the results of serial samples can be compared.
The authors are grateful to Vétoquinol UK for access to the phenobarbitone voucher scheme data.
- British Veterinary Association. All rights reserved.