Abstract
Objectives
Recently, the increased cost and decreased availability of doxycycline has sparked an interest in using minocycline as an alternative. The purpose of this study was to determine the pharmacokinetics of minocycline in domestic cats in order to facilitate dosage decisions.
Methods
Purpose-bred, young adult cats were administered a single dose of either intravenous (IV; n = 4; 5 mg/kg) or oral (n = 6; 50 mg/cat) minocycline. Blood was collected from each at intervals up to 24 h afterwards. Minocycline was measured using high performance liquid chromatography with ultraviolet detection. A one-compartment pharmacokinetic model was fit to the oral data and a two-compartment model to the IV data via a computer program. Plasma protein binding was measured by fortifying blank plasma from untreated healthy cats with minocycline at two concentrations and applying an ultracentrifugation method.
Results
Two cats became transiently lethargic and tachypneic during IV drug infusion. One cat vomited 6.0 h after infusion, and two cats vomited either 1.5 h or ~5.0 h after oral drug administration. The mean oral dose administered was 13.9 ± 0.47 mg/kg. Oral bioavailability was approximately 62%. Plasma protein binding was 60% at 5 µg/ml and 46% at 1 μg/ml. After IV administration, elimination half-life (t½), apparent volume of distribution at steady-state, and systemic clearance were 6.7 h (coefficient of variation [CV] 14.4%), 1.5 l/kg (CV 34.5%) and 2.9 ml/kg/min (CV 40.8%), respectively. After oral administration the terminal t½ and peak concentration (Cmax) were 6.3 h (CV 9%) and 4.77 µg/ml (CV 36%), respectively.
Conclusions and relevance
Because most bacteria will have a minimum inhibitory concentration of ⩽0.5 μg/ml, an oral dose of 8.8 mg/kg q24h would be adequate to meet pharmacokinetic–pharmacodynamic targets after adjusting for protein binding. Although some gastrointestinal upset may occur, one 50 mg capsule orally q24h would provide appropriate dosing for most cats.
Introduction
Tetracyclines were the first broad-spectrum antibacterial drugs and are still used frequently in veterinary medicine. These drugs possess efficacy against feline pathogens, including Mycoplasma species, Rickettsia species, Chlamydia species, a variety of gram-negative (eg, Pasteurella species, Bartonella species, Francisella species) and gram-positive (eg, some methicillin-resistant Staphylococcus pseudintermedius) bacteria, and even against some anaerobes (eg, Clostridium species, Actinomyces species).1–4 In addition to their antibacterial actions, tetracyclines also offer anti-inflammatory actions, inhibit matrix metalloproteinase, possess antiapoptotic and antioxidant properties, and may be neuroprotective.5–8
In cats, doxycycline has long been the tetracycline of choice by most veterinarians for treating a large number of infections, particularly hemoplasmosis, upper respiratory infections, chlamydia and vector-borne infections. Recent manufacturer shortages of doxycycline, along with increased costs, have lead clinicians to seek alternative options. 9 Although not currently approved for use in cats, minocycline has a similar but slightly broader antimicrobial spectrum than doxycycline, is less protein bound, is more lipophilic – which may improve intracellular and respiratory tract distribution – and has a longer half-life (t½) in humans.10–13 A recently published study by one of our authors showed some favorable characteristics for its use in dogs. 14 However, no pharmacokinetic data exist to guide use in cats. The objective of this study was to describe the pharmacokinetic profile of minocycline in healthy cats after both oral and intravenous (IV) administration.
Materials and methods
Animals and sample collection
Seven purpose-bred female intact domestic shorthair cats, approximately 10 months of age and ranging in weight from 3.1–4.4 kg, were used. Cats were cared forin accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and all procedures were approved by the Animal Care and Use Committee of the University of Missouri. As part of an unrelated study, each cat was fitted with a subcutaneous jugular venous access port (VAP; CompanionPort CP4 [Access Technologies]) prior to this study. After obtaining informed owner consent, an additional three healthy pet cats were used to obtain 10 ml plasma for use as blanks in drug analysis.
Both IV (n = 4) and oral (n = 6) studies were performed. For the IV dose study, minocycline (Triax Pharmaceuticals) was administered at a dose of 5 mg/kg intravenously to each of four cats via a cephalic venous catheter placed just prior to drug infusion. Minocycline was diluted to 10 mg/ml stock solution in sterile water, and the calculated dose for each cat was further diluted to make up 10 ml in sterile 0.9% saline solution. The drug was infused by means of syringe pump over 10 mins. Oral minocycline (Watson Pharmaceuticals) was administered to each of six cats as a 50 mg capsule, regardless of the cat’s weight, followed immediately with 6 ml of water PO. Three of the four cats receiving the IV dose were also given the oral medication; for these three, a 1 week washout period occurred between IV and oral administration. Cats were monitored for adverse reactions continuously for the first 2 h after dose administration and periodically thereafter.
Blood samples were obtained via the VAP at 10, 20, 40 and 60 mins, then at 2, 4, 6, 12, 15 and 24 h postadministration of minocycline; for the IV study, the first sample was collected 10 mins after infusion was completed. Prior to sampling, topical anesthetic cream (lidocaine/prilocaine emulsion; APP Pharmaceuticals) was applied to the skin over the venous access site. An initial aliquot of blood (0.5 ml) containing heparin solution from the vascular access port was discarded; then, 2 ml of blood was aspirated for use in analysis. Immediately after sample collection, each VAP was flushed with heparinized saline.
Blood samples were transferred into lithium heparin-containing collection tubes and immediately placed on ice and protected from light. Samples were centrifuged at 750 g for 15 mins at 4°C, and plasma was collected and stored at −80°C until it was shipped overnight on dry ice for analysis.
Sample analysis and calculations
Plasma samples were analyzed by high performance liquid chromatography (HPLC) to determine drug concentrations. Separation was achieved at 40°C with a Zorbax SB-C8 4.6 × 150 mm reverse phase column (Agilent Technologies). The system consisted of a quaternary solvent delivery system (Agilent Technologies 1100 Series), an autosampler providing a 20 µl injection (Agilent Series 1100), an ultraviolet (UV) detector with UV absorbance set at 350 nm (Agilent Technologies 1200 Series) and a computer for data collection and analysis (Agilent 1200 Series ChemStation 2D software version B.03.02). The mobile phase consisted of 5% acetonitrile, 7% methanol and 88% oxalic acid buffer. The flow rate was 1 ml/min. Plasma samples were prepared via a drug releasing protein precipitation method validated and used in our laboratory.14–16 The lower limit of quantification for this assay was defined as the lowest concentration of analyte that could be quantified with acceptable precision and accuracy using guidelines on method validation from International Conference on Harmonisation (ICH). 17
Plasma drug concentrations after administration of minocycline were analyzed, with the aid of a computer program (Phoenix WinNonlin version 6.1.0.173; Pharsight), using compartmental pharmacokinetic methods to determine drug disposition for each cat. Initial selection of appropriate models for analysis was made after plasma drug concentrations were plotted on linear and semi-logarithmic graphs for visual inspection. Plasma drug concentrations were weighted by a factor of 1/(predicted Y) 2 for pharmacokinetic analysis, where Y is the plasma concentration. The specific model (eg, one compartment, two compartments) was determined for best fit on the basis of a smaller value for the Akaike’s information criterion. 18 For the two-compartment, bi-exponential analysis, the following model was fit to the data:
where C is the plasma drug concentration at time t; A1 and A2 are the y-axis intercepts for the distribution and elimination phases of the curve, respectively; λ1 and λ2 are the slopes of the distribution and elimination phases of the curve, respectively; and e is the base of the natural logarithm. The elimination t½ was estimated from the relationship: t½ = Ln 0.5/λ2. Other compartmental pharmacokinetic parameters are calculated according to published formulae. 19
For the oral dose, a one-compartment model with first order elimination was used. The general formula for the compartmental analysis was:
where C is the plasma concentration, t is time, Ka is the non-intravenous absorption rate (assuming first-order absorption), K is the elimination rate constant, V is the apparent volume of distribution, F is the fraction of drug absorbed and D is the non-intravenous dose. In this model, it is assumed that Ka was much greater (at least5 ×) than K, or that there is no ‘flip-flop’ effect caused by slow gastrointestinal absorption. To account for dissolution of capsule and stomach emptying time, a lag-time was added to the model. Secondary parameters calculated from the model included peak concentration (Cmax), time to peak, area under the drug concentration time curve (AUC), systemic clearance per fraction absorbed, and the respective absorption and terminal t½.
Oral bioavailability was calculated based on the AUC from all four cats used in the IV dosing study and all six cats given oral dosing. The formula used for this estimation, where F = bioavailability, was
Determination of minocycline plasma protein binding was performed with a micropartition device (Centrifree Micropartition system; Amicon Millipore). Aliquots of pooled feline plasma from untreated cats were spiked with minocycline to generate concentrations of 1 and 5 µg/ml, representing a range of concentrations anticipated for this study. The spiked samples were each divided into three replicates of 1 ml, which were subsequently added to the micropartition system to obtain a protein-free ultrafiltrate for HPLC analysis. A second set of three replicates of spiked plasma at the same concentrations were processed by normal plasma protein precipitation procedure and analyzed by HPLC for comparison. Plasma protein binding was determined using the following equation:
Results
Drug dosing and tolerance
Although each cat in the IV dosing study was to receive 5 mg/kg, cat 3 had minor initial drug loss due to improperly connected fluid line. Although the exact volume of fluid/drug lost could not be quantified, it was estimated to be approximately 1.5 mg of a planned 18.3 mg dose. The entire 5 mg/kg dose was given successfully intravenously to the other three cats. Administration of a single 50 mg minocycline capsule to each of six cats resulted in a mean dose of 13.94 ± 0.47 mg/kg (range 11.36–16.13 mg/kg) for the oral dosing study.
Minor adverse reactions were observed after drug administration. During the IV infusion, two of the cats became mildly tachypneic and demonstrated reduced activity levels compared with behavior prior to the infusion. Signs resolved within 10–20 mins of completion of the infusion. One cat vomited approximately 6 h after IV drug administration. After oral drug administration, two cats vomited. Cat 3 vomited 1.5 h postdose; vomitus appeared yellow and likely contained at least some portion of the drug. Cat 6 vomited sometime between 4 and 6 h after oral dosing but there was no evidence of the yellow drug in the vomitus.
Pharmacokinetic measurements
Mean ± SD plasma concentrations overlaid on the fitted line from the model are shown in Figure 1. Calculated pharmacokinetic parameters determined from the fitted model of the plasma minocycline concentrations over time are presented in Tables 1 and 2. After IV administration, the elimination t½, apparent volume of distribution at steady-state and systemic clearance were 6.7 h (coefficient of variation [CV] 14.4%), 1.5 l/kg (CV 34.5%) and 2.9 ml/kg/min (CV 40.8%), respectively. After oral administration, the terminal t½ and Cmax were 6.3 h (CV 9%) and 4.77 µg/ml (CV 36%), respectively. Mean oral bioavailability was 61.9%. For oral dose, the AUC was 52.10 ± 17.20 h*μg/ml. Protein binding at plasma concentration of 5 μg/ml was 60.6 ± 0.71%, and 46.5 ± 2.91% at 1 μg/ml.
Plasma concentrations of minocycline in cats after intravenous (IV) injection of 5 mg/kg (n = 4) and oral administration of 13.94 mg/kg (n = 6). The points represent mean ± SD concentrations, with lines overlaid representing the fitted lines from the pharmacokinetic model
One-compartment model with first order input for oral administration of minocycline in six healthy cats
VD/F = volume of distribution per fraction absorbed; K01 = absorption rate and corresponding half-life (t½); K10 = elimination rate constant and corresponding t½; Tlag = lag time for drug to appear in systemic circulation; AUC = area under the drug concentration time curve; CL/F = systemic clearance per fraction absorbed; Tmax = time to peak concentration; Cmax = peak concentration; CV = coefficient of variation
Two-compartment model with bolus input after intravenous administration of 5 mg/kg minocycline in four healthy cats
*Because there was a minor, but undefined, loss of dose for cat 3, mean ± SD excluding this cat are provided in parentheses
A = y-axis intercept for distribution phase; α (also called λ1) = distribution rate constant, and corresponding half-life (t½); AUC = area under the drug concentration curve; B = y-axis intercept for elimination phase; β (also called λ2) elimination rate constant, and corresponding t½; Cl = systemic clearance; C0 = initial plasma concentration extrapolated to y-axis intercept; CV = coefficient of variation; K12, K21 = microdistribution rate constants; MRT = mean residence time; Vc = volume of distribution of the central compartment; Vss = apparent volume of distribution at steady-state
Discussion
The results of this study demonstrate that minocycline can practically achieve what are likely useful concentrations for treatment of susceptible bacterial infections in cats. Clinical antimicrobial efficacy depends on the ratio of the free drug concentration to the minimum inhibitory concentration (MIC) for the pathogen in question (ie, AUC/MIC). 12 The dosage and efficacy will therefore depend on the MIC for a specific pathogen. Breakpoints for human bacterial isolates are available from standards published by the Clinical and Laboratory Standards Institute (CLSI). However, the Veterinary Antimicrobial Testing Subcommittee of CLSI has not yet established interpretive criteria for setting minocycline breakpoint values for feline pathogens. Nonetheless, it is likely that for most common pathogens that might be treated with tetracyclines, the MIC will be ⩽0.5 μg/ml based on wild-type distributions of common bacteria. Based on an MIC of ⩽0.5 μg/ml, a dose can be calculated from data obtained in this study to estimate a dose that could be used in future clinical studies. With an AUC/MIC of unbound (free) drug plasma concentration as a target, protein binding and clearance values measured from this study, the dose can be calculated to attain the AUC/MIC target of 25 suggested for tetracyclines. 20 Using this approach, an approximate minocycline dose of 8.7 mg/kg once daily would attain this target in an average cat. Half this dose (4.3 mg/kg) twice daily would produce the same exposure assuming linear pharmacokinetics (dose proportional). As minocycline is commercially available as 50 mg capsules, a single 50 mg dose once daily for a 5 kg cat would meet this target. Minocycline is also available in tablet formulation but it is undetermined if tablets would produce higher or lower concentrations that those measured in this study.
Only the unbound fraction of drug can penetrate tissues to reach extravascular sites of infection. In our study, approximately 40% of the drug was unbound. This is slightly less than the protein binding of doxycycline in cats but is very similar to the protein binding reported for minocycline in dogs.14,21 Compared with tetracycline and oxytetracycline, minocycline is more lipophilic. Minocycline passes more effectively through blood–brain and blood–ocular barriers than other tetracycline congeners, and minocycline accumulation in the brain and cerebrospinal fluid is greater than that of doxycycline.16,22
Oral and IV minocycline were reasonably tolerated in all cats. In dogs, rapid IV administration of minocycline has been associated with adverse effects, including severe hypotension, cardiovascular depression and hemolysis.14,22,23 However, in a more recent study in dogs, a rapid IV injection was well tolerated except for some hypercoagulability in one dog. 14 We are unware of previous reports of IV administration of minocycline to cats. Although we did not measure blood pressure in treated cats, two did become transiently lethargic, which may or may not have been related to hypotension. In these cats, behavior normalized very quickly after the infusion was halted. While hematocrit was not measured, hemolysis was not detected in centrifuged blood samples. The only other adverse event after IV dosing was vomiting in one cat. An additional two cats vomited after oral minocycline administration. Gastrointestinal signs such as vomiting, diarrhea and anorexia are reportedly the most common adverse reactions in cats to the closely related drug doxycycline. 24 In humans, gastrointestinal tract disturbances are commonly observed following oral administration of tetracycline antibiotics; however, doxycycline shows the fewest gastrointestinal adverse effects among the tetracycline antibiotics. However, of all the tetracycline antibiotics, doxycycline hyclate, especially in capsule form, is most likely to cause esophageal ulceration. 25 It is undetermined how minocycline compares in these properties.
Vomiting after oral administration of the drug may have affected oral drug absorption, particularly in the cat that vomited less than 2 h after receiving the medication. The absorption t½ for this cat was 0.23 h compared with a mean for all six cats of 0.32 h. The Cmax for this cat was actually higher than the mean Cmax for all cats given oral minocycline, suggesting that drug absorption was adequate for interpretation of pharmacokinetics, despite a potential loss of some drug.
There were limitations to this study. The first limitation relates to the number of animals used, with only six cats included in the oral dosing study and four for IV dosing. Further, all of the cats evaluated were domestic shorthair cats of similar age and of the same sex. The variation (CV %) we observed in the oral pharmacokinetic parameters in this group was somewhat less than observed in other studies. Data from additional cats of various breeds, ages and of either sex might provide more information on the variability among cats in a treatment population. All of the cats used in this study were healthy. Illness and inflammation can affect drug pharmacokinetics. Organ dysfunction can also produce changes in drug distribution and clearance. In the case of minocycline, the drug is eliminated by non-renal routes in dogs but is undetermined in cats. 22 The effect of age, breed, organ dysfunction, body condition, infection or sex on pharmacokinetics of minocycline is undetermined and requires further study. In our study, oral dosing was performed in fasted cats, and thus we cannot judge how feeding might affect the pharmacokinetics of the drug. In dogs, feeding decreases oral absorption of minocycline. 26 As the study did not use a crossover design, calculated oral bioavailability should be viewed as an estimate. Finally, oral dosing used a capsule formulation; it is possible that bioavailability may be less were a tablet formulation used and that, in turn, might influence dosing. Dosing recommendations derived from this study are presented only to help guide further studies, and we make no claims of efficacy at these doses. These dosages should be confirmed with clinical studies in cats with infections caused by susceptible bacteria.
Conclusions
Oral minocycline was reasonably well tolerated in cats. Pharmacokinetic analysis suggests that minocycline could be used to treat susceptible infections in cats at a dose of 8.7 mg/kg once daily. Practically, a single 50 mg capsule given once daily would be useful for most adult cats. If vomiting occurs at this dose, treatment with 4.3 mg/kg twice daily may be implemented instead.
Footnotes
Acknowledgements
We wish to thank Mr Matt Haight and Mrs Delta Dise for their valuable technical assistance.
Conflict of interest
The authors have no potential conflicts of interest to declare.
Funding
This work was funded, in part, by the Oliver Kronke Fund for Internal Medicine Investigation at the University of Missouri; in part by a 5013c charitable foundation – the ALSAM Foundation; and, in part, by a contribution from the North Carolina State Clinical Pharmacology Laboratory.