Hepatic lipidosis in camelids is a relatively uncommon clinical entity that can be difficult to treat.1–4 It is a multifactorial disease characterized by excessive fat mobilization secondary to negative energy balance, and clinical signs are often not apparent until later stages of the disease.5 A negative energy balance may result from concurrent disease conditions; inadequate nutrition; and stress associated with transport, parturition, or lactation.5 Excessive fat mobilization leads to hepatic lipidosis.5 The hepatic lipidosis syndrome in South American camelids is characterized by inappetence; weight loss; high serum concentrations of glucose, bilirubin, bile acids, and nonesterified fatty acids; high serum alkaline phosphatase, γ-glutamyltransferase, and sorbitol dehydrogenase activities; and decreased responses to insulin.5,6 Lactating and pregnant camelids appear to be at greater risk for developing hepatic lipidosis because of increased energy demands imposed by the mammary gland and developing fetus.1,2,5
Serum biochemical markers used to assess hepatocyte integrity (eg, sorbitol dehydrogenase and aspartate aminotransferase activities), biliary function (eg, γ-glutamyltransferase activity), and liver excretion (eg, concentration of bile acids) increase earlier in feed-restricted llamas that develop hepatic lipidosis, compared with llamas that do not develop hepatic lipidosis.6 Those findings suggest that feed restriction during periods of increased energy demand may result in subclinical liver dysfunction.6 Definitive diagnosis of lipidosis is typically made on the basis of results of histologic examination of liver biopsy specimens.1,2,5 However, less invasive tests of hepatic function may prove useful in assessing hepatopathy and hepatocyte damage in animals that are at risk.
Reference values for caffeine clearance have been determined for multiple species. Caffeine clearance is a useful measure of hepatic function because caffeine has low toxicity, is simple to measure, and has an elimination half-life that is highly correlated with hepatic function.7–13 The purpose of the study reported here was to determine total plasma clearance and elimination half-life of caffeine in clinically normal alpacas and llamas.
Materials and Methods
Animals—Four castrated adult male alpacas and 4 adult female llamas were used. Animals were housed on fescue pasture and received supplemental grass hay. All were determined to be healthy on the basis of complete physical examination and results of a CBC and serum biochemical analysis. Animals were placed in stalls in groups of 2 and allowed to acclimate to confinement for 1 week prior to initiation of the study. All llamas and alpacas were observed to be consuming grass hay and water and urinating and defecating normally. On the day before the start of the study, each animal was weighed and an IV catheter was placed in each jugular vein; 1 catheter was used for drug administration, and the other was used for blood sampling. Prior to drug administration, after drug administration, and after sample collection, catheters were flushed with heparinized saline (0.9% NaCl) solution. Procedures were reviewed and approved by the institutional animal care and use committee of each institution (ie, University of Missouri and The Ohio State University).
Experimental design—Caffeine was administered (dose, 3 mg/kg IV) as caffeine and sodium benzoatea in a sterile pyrogen-free normal saline solution (concentration, 50 mg of caffeine/mL of solution) through 1 of the catheters.8 Heparinized blood samples (7 mL) were withdrawn from the contralateral catheter at the time of administration and 3, 6, 12, 30, and 45 minutes and 1, 1.25, 1.5, 1.75, 2, 2.25, 2.5, 3, 4, 5, 6, 8, 12, 24, 36, 48, 72, 96, and 120 hours after administration. Sampling time points were similar to those previously reported.8 Plasma was harvested by centrifugation of heparinized whole blood at 1,000 × g for 10 minutes, placed in individual cryovials, and stored at −70°C until analyzed. Four sets of alpaca blood samples and 2 sets of llama blood samples were analyzed via HPLC at the University of Missouri. The 2 remaining sets of llama samples were analyzed via LC-MS at The Ohio State University.
Preparation of caffeine stock solutions for use in HPLC determination of plasma caffeine concentrations—Stock solutions (A) of caffeine were prepared by dissolving 100 mg of caffeine in 10 mL of methanol-water (50:50, vol/vol; end concentration, 10 mg of caffeine/mL).b Stock solution A was stored in darkness at 4°C. Working standard solutions of caffeine were prepared daily by dilution of stock solution A in methanol-water (50:50, vol/vol) to 1 mg/mL (B), 0.1 mg/mL (C), and 0.01 mg/mL (D). Appropriate volumes of stock solutions A, B, C, and D were added to 0.25 mL of blank alpaca plasma for preparation of plasma caffeine calibrators ranging in concentration from 0.031 μg of caffeine/mL to 64 μg of caffeine/mL. Plasma caffeine calibrators were prepared, extracted, and analyzed with each set of experimental samples. In addition, negative control samples (ie, blank alpaca plasma with no detectable caffeine) were prepared and extracted with each batch of experimental samples.
Sample preparation for HPLC determinations—Plasma samples were thawed at ambient temperature and mixed to assure homogeneity. Three 250-μL aliquots of each plasma sample were pipetted into individual tubes. Samples were prepared according to a described protocol14 involving acid precipitation of protein and injection of the protein-free supernatant solutions onto the chromatograph.14 In brief, to each 250-μL aliquot of plasma, an equal volume of 0.8M perchloric acid was added; resulting samples were mixed vigorously and centrifuged at 14,000 × g for 20 minutes at 21°C.14 A 200-μL aliquot of clarified supernatant solution was transferred from each tube to an autosampler vial containing 10 μL of 4M NaOH.14 Vials were capped and mixed, and 50 μL was injected onto the liquid chromatograph.
HPLC system—An HPLC pump and controller, autosampler, and dual-wavelength UV detector were used for analyses.c The mobile phase was modified from a previously used mixture13 to optimize retention time and peak shape with the columnd (25 × 0.46 cm; 5-μm particle size) used in the present study. The mobile phase was 20mM phosphate buffer (pH, 3.0)-acetonitrile, 85:15 (v/v), at a flow rate of 1 mL/min. Caffeine in test samples was quantified on the basis of characteristic retention time and UV absorbance at 273 nM, in comparison with caffeine standards.13,b Data were recorded by use of a computer workstation and chromatography software.c Daily sample analysis included the running controls, calibrators, and experimental samples, all of which were prepared, extracted, and analyzed together.
Preparation of caffeine and 8-chlorocaffeine stock solutions for use in LC-MS determination of plasma caffeine concentrations—Caffeine reference standard solutions (1.0 mg/mL) were obtained from commercial sources.e,f Working standard solutions (10 ng/μL; A and B) were prepared by diluting 0.1 mL of each of the reference standard solutions with methanol to a volume of 10.0 mL. Further 10-fold dilutions of A and B were made to prepare 1.0 ng/μL standard solutions (C and D). The gravimetric reference standard solution of 8-chlorocaffeine (E; internal standard with a concentration of 1.0 mg/mL) was prepared by dissolving 50 mg of 8-chlorocaffeine in 50 mL of methanol.g Working internal standard solution (F; 10 ng/μL) was prepared by diluting 0.1 mL of the gravimetric standard solution to a volume of 10 mL with methanol. Stock solutions were stored in darkness at 4°C.
Sample preparation for use in LC-MS determinations—Appropriate volumes of working standard caffeine solutions (A and C) were added to 0.5 mL of blank alpaca plasma for preparation of 2 sets of caffeine calibrators (0.025 to 1.0 μg/mL and 1.0 to 20.0 μg/mL) in blank alpaca plasma. Control samples were prepared from negative control plasma by adding appropriate volumes of working standard solutions (B and D). To each experimental sample, calibrator, and control sample, 0.25 μg of 8-chlorocaffeineg was added, samples were vortexed, and 2 mL of 1M sodium acetate buffer (pH, 5) was added. After mixing, 4 mL of methylene chloride-isopropanol (9:1, vol/vol) was added, tube contents were vortexed for 10 minutes, and tubes were centrifuged at 1,000 × g for 5 minutes. Bottom (solvent) layers were transferred to clean tubes and evaporated under nitrogen at 40 ± 5°C. Dried residues were dissolved in 35 μL of methanol, and that mixture was added to 215 μL of aqueous 30mM ammonium formate (pH, 6.5). Sample extracts were transferred to autosampler vials and injected onto the LC-MS system. Control samples included negative control (ie, no detectable caffeine), positive controls (0.03, 0.8, 1.2, and 16.0 μg of caffeine/mL; 0.25 μg of 8-chlorocaffeine), system washes (methanol-30mM ammonium formate; pH, 6.5), and solvent blanks (0.5 mL of water and 0.25 μg of internal standard).
LC-MS system—Plasma caffeine concentrations were determined by use of LC-MSh in 2 llamas. Briefly, caffeine was quantified by use of an LC-MSD trapi equipped with a guard column (2.1 × 12.5 mm; 5 μm) and C18 (2.1 × 150 mm; 5 μm) analytical column.j The mobile phase (solvent A, 30mM ammonium formate with pH of 6.5; solvent B, methanol) was delivered at a flow rate of 0.3 mL/min. The gradient developed was 16% solvent B from time 0 to time 1.0 minute, with ramp from 16% to 85% solvent B from time 1.0 minute to time 6.5 minutes; hold from time 6.5 minutes to time 8.5 minutes; and 16% solvent B from time 8.51 minutes to time 15.0 minutes. Column temperature was maintained at 35°C, and 10 μL of each sample extract was injected. The mass spectrometer was operated in positive-ion mode at normal resolution with nitrogen flow at 9 L/min (dry gas temperature, 365°C). The mass spectrometer was tuned to –3,500 V at 20 V (skimmer), 110-V capillary exit, octapole 1 (14 V), octapole 2 (2.5 V), octapole RF (80 V), lens 1 voltage (–5 V), and lens 2 voltage (–60 V). Mass spectrometer parameters were as follows: caffeine transition: m/z 195 → 110, 138, 151; caffeine isolation: m/z 195; isolation width, 0.7; fragmentation cutoff, 125; fragmentation amplitude, 2.0; with 4 scan averages. Parameters used to identify 8-chlorocaffeine were as follows: transition: m/z 229 → 144, 172; isolation: 229.
HPLC method validation: standard curves, detection limit, and recovery—The limit of quantitation for determining caffeine concentration in plasma via HPLC was estimated by injection of calibrators in triplicate on 4 days (n = 4 samples; range, 0.031 μg/mL to 64 μg/mL) and determining the concentration at which the relative SD for difference between caffeine detected and caffeine added approximated 20%.15 The corresponding concentration was the limit of quantitation. The limit of detection was determined on the basis of the lowest concentration with a signal-to-noise ratio of 3:1.15
Known concentrations of caffeine in plasma were used to determine recovery of caffeine from acid-precipitated plasma samples. Recovery was calculated by use of this equation:
HPLC accuracy and precision—Blank alpaca plasma was used to prepare control samples at 4 fortification levels (0.0625, 0.5, 8.0, and 64 μg of caffeine/mL), and a minimum of 3 replicates were analyzed at each fortification level. Each replicate was injected in triplicate, and the mean value of the peak area of caffeine in each replicate was used to determine concentration from the intercept and slope of the line of that day's calibration curve. Those plasma fortifications were prepared and analyzed on 4 separate days.16 Within-day accuracy was determined each day for each fortification level as the difference between caffeine detected and caffeine added and by determining the relative SD for each fortification level on that day:
Precision of the HPLC assay was determined by running the caffeine-fortified plasma samples in triplicate on 4 separate days. Within- and between-day precisions were determined by subjecting the daily replicate values (detected) to ANOVA (1-way design) and determining the source of variance between between-day and within-day values (reported as relative SD of the mean [SD%])16:
Method validation: LC-MS calibration curves, detection limit, and recovery—The ratio of peak area of caffeineto-peak area of 8-chlorocaffeine was calculated for each calibrator, control sample, and experimental sample extract. Concentrations of caffeine in experimental and control samples were calculated from the slope and intercept of the equation for the calibration line obtained by linear regression analysis of the peak area ratio versus the caffeine concentration in the corresponding calibrator. Plasma caffeine concentrations determined from analyses of positive control samples (nominally 0.03, 0.8, 1.2, and 16.0 μg of caffeine/mL) by use of LC-MS analysis were subjected to similar calculations to determine assay accuracy and precision.16 Positive control samples from 2 sourcese,f were used to independently verify results.
Pharmacokinetic analysis—Plasma concentration-versus-time data were summated, and mean ± 1 SD was determined for alpacas (n = 4) and llamas (4). Log plasma concentration-versus-time data were plotted.
Individual plasma concentration-versus-time data underwent compartmental analysis by use of commercially available software.k Zero-time plasma drug concentration intercepts of the biphasic IV disposition curve (A, B) were determined via the computer program by extrapolation of the plasma caffeine concentration to t = 0 for the distribution phase (A) and terminal exponential phase (B). The hybrid rate constants of the biphasic IV disposition curve (α and β) are related to the slopes of the distribution and elimination phase, respectively. The first-order elimination rate constant for disappearance of the drug from the central compartment is K10, whereas the first-order transfer rate constants for drug distribution between the central and peripheral compartments are K12 and K21, respectively. The total plasma clearance was estimated according to this equation:
The apparent volume of the central (V1) and peripheral (V2) compartments and volume of distribution at steady state (Vss) were estimated from this equation:
This equation was used to calculate Vss:
This equation was used to calculate V2:
Distribution half-life (h) was calculated by use of the following equation:
Elimination half-life was calculated by use of the following equation:
The maximum plasma concentration for each data set is represented by Cpmax.
Area under the curve from time of dosing to infinity was calculated by use of this equation:
The AUMC from time of dosing to infinity was calculated by use of the following equation:
Mean residence time was determined from the ratio of AUMC to AUC according to the following equation:
The compartmental model that best described the data for each animal was chosen on the basis of the Akaike information criterion.17 Individual animal coefficients of the compartmental models and variables derived from those coefficients were reported as median and range. Pharmacokinetic variables obtained from alpaca and llama data were compared by use of the rank sum test.
Results
Plasma caffeine concentrations were readily determined in alpaca and llama plasma by use of the HPLC-UV method. Retention time of caffeine in the system was 9.5 to 10.4 minutes. Recovery of caffeine from camelid plasma was good (> 85%) over the range of concentrations examined (Table 1). When caffeine-fortified samples prepared in alpaca plasma (at concentrations of 0.0625, 0.5, 8.0, and 64 μg of caffeine/mL) were assayed via HPLC, recoveries were 106%, 87%, 94%, and 89%, respectively.
Summary of data from HPLC and LC-MS assays for between- and within-day accuracy and precision and percentage recovery of drug after extraction over the range of concentrations used to prepare the standard curves. Data represent percentages and are the means of 3 replicates each of 4 concentrations (for HPLC, 0.0625, 0.5, 8.0, and 64 g of caffeine/mL) run on 4 different days or represent positive control calibrators (for LC-MS, 0.03, 0.8, 1.2, and 16 m g of caffeine/mL). On each day, a standard curve was prepared for analysis of the calibrators run on that day.
| Analytic method | Concentration (μg/mL) | Within-day accuracy | Within-day precision | Between-day accuracy | Between-day precision | Recovery |
|---|---|---|---|---|---|---|
| HPLC | 64 | 11.9 | 3.4 | 11.2 | 5.0 | 89 |
| 8 | 3.6 | 1.9 | 5.6 | 4.6 | 94 | |
| 0.5 | 13.5 | 1.8 | 13.4 | 2.8 | 87 | |
| 0.0625 | 4.5 | 8.2 | 5.9 | 18.4 | 106 | |
| LC-MS | 16 | 0.03 | 7.2 | 5.2 | 12.1 | 105 |
| 1.2 | 1.9 | 5.2 | 5.3 | 6.3 | 95 | |
| 0.8 | 3.9 | 1.0 | 5.1 | 2.5 | 95 | |
| 0.03 | 3.0 | 9.2 | 7.7 | 3.5 | 92 |
In test samples, retention times of caffeine (7.4 to 7.5 minutes) and the internal standard (8.9 to 9.0 minutes) as determined by use of LC-MS were within ± 2% of values observed in the corresponding standards, and no interference from other substances was observed. Caffeine recoveries from positive-control plasma samples (at concentrations of 0.03, 0.8, 1.2, and 16 μg/mL) as assayed by use of LC-MS were 92%, 95%, 95%, and 105%, respectively (Table 1).
Regression analysis of 6 calibration curves generated by use of HPLC indicated that the relationship between peak area and caffeine concentration was linear over the range examined (ie, 0.0625 μg of caffeine/mL to 64 μg of caffeine/mL; r2 > 0.994). The limit of quantitation for the HPLC assay for caffeine in plasma was 0.0625 μg/mL. Regression analysis of the 5 calibration curves generated by use of LC-MS indicated that the relationship between the ratio of peak area (caffeine/internal standard) and caffeine concentration was linear over the range examined (0.025 μg/mL to 20 μg/mL; r2 > 0.995). The limit of quantitation for the LC-MS method was approximately 0.025 μg/mL.
Within- and between-day accuracy for HPLC analysis as determined by the relationship between concentration added and concentration detected indicated that there was < 15% variation in replicate determinations at all fortification levels on the 4 days (Table 1). Assay precision was < 15% for within-day and < 15% for between-day determinations at all concentrations except 0.0625 μg/mL, for which the relative SD for between-day precision was < 20%. Both between- and within-day accuracies for the LC-MS assay were < 10% for concentrations from 0.030 μg/mL to 16 μg/mL.
Plasma caffeine concentrations remained above the HPLC-method limit of quantitation for 72 hours in all alpaca samples and in the 2 samples from llamas. Plasma caffeine concentrations remained above the HPLC-method limit of quantitation for 120 hours in 1 llama and above the LC-MS limit of quantification for 120 hours in the 2 llamas.
Mean plasma concentrations of caffeine in alpacas and llamas were summarized (Figure 1). Concentration-time data were best described by a biexponential equation corresponding to a 2-compartment open-pharmacokinetic model. Pharmacokinetic variables for alpacas and llamas derived from 2-compartment analysis of concentration-time data were summarized (Table 2). No significant differences between pharmacokinetic variables for alpacas and llamas were detected (P < 0.05).
Median (range) values for pharmacokinetic variables from a 2-compartment model, first-order elimination for caffeine administered IV to healthy adult alpacas and llamas (n = 4 each).
| Pharmacokinetic variable | Units | Alpaca | Llama |
|---|---|---|---|
| A | g/mL | 8.84 (6.7–13.4) | 3.3 (2.8–3.5) |
| α | 1/h | 7.1 (0.2–10.3) | 1.0 (0.41–2.1) |
| B | g/mL | 4.9 (0.4–6.6) | 3.6 (1.9–4.2) |
| b | 1/h | 0.06 (0.02–0.074) | 0.43 (0.04–4.15) |
| Cpmax | g/mL | 13.73 (7.0–20.0) | 6.7 (5.5–7.4) |
| K10 | 1/h | 0.017 (0.13–0.20) | 0.09 (0.07–0.24) |
| K12 | 1/h | 4.4 (0.04–6.7) | 0.43 (0.13–0.89) |
| K21 | 1/h | 2.6 (0.03–3.4) | 0.61 (0.18–1.1) |
| Cltot | mL/h/kg | 44 (24–56) | 42 (30–109) |
| V1 | mL/kg | 0.22 (0.15–0.43) | 0.45 (0.41–0.55) |
| V2 | mL/kg | 0.39 (0.30–0.50) | 0.32 (0.23–0.60) |
| Vss | mL/kg | 0.60 (0.45–0.93) | 0.75 (0.68–1.15) |
| T1/2α | h | 0.10 (0.07–3.95) | 1.00 (0.33–1.7) |
| T1/2β | h | 11.0 (9.3–29.8) | 16 (5.4–17) |
| AUC∞ | g•h/mL | 68.6 (54.0–124) | 76 (27–99) |
| AUMC∞ | h•h• g/mL | 971 (808–2,285) | 1,666 (203–2,290) |
| MRT | h | 15.0 (13.0–18.5) | 21 (7.4–23) |
A = Zero-time plasma drug concentration intercept of initial distribution phase of the biphasic IV disposition curve. a = Hybrid rate constant of the distribution phase of the biphasic IV disposition curve. B = Zero-time plasma drug concentration intercept of the terminal exponential phase of the biphasic IV disposition curve. b = Hybrid rate constant of the elimination phase of the biphasic disposition curve. Cpmax = Maximum plasma concentration. K10 = First-order elimination rate constant. K12, K21 = First-order rate transfer constants for drug distribution between the central and peripheral compartments, respectively. Cltot = Total body clearance or plasma clearance of drug (sum of all clearance processes). V1, V2 = Apparent volume of central and peripheral or tissue compartments, respectively. Vss = Apparent volume of distribution at steady state. T1/2α = Half-life of the distribution phase. T1/2β = Half-life of the elimination phase. AUC∞ = Total area under the plasma concentration-versus-time curve to infinity. AUMC∞ = Area under the moment curve to infinity. MRT = Mean residence time.
Relationship between log10 plasma concentration and time for pharmacokinetic disposition of caffeine in 8 healthy adult alpacas and llamas (n = 4 each). Animals received 3 mg of caffeine/kg IV, and samples were collected from time = 0 to time = 120 hours after administration. Datum points represent mean values. Error bars represent SD of data at each time point. Inset—plasma concentration-time curves for the interval between time = 0 and time = 12 hours after administration.
Citation: American Journal of Veterinary Research 67, 6; 10.2460/ajvr.67.6.1063
Discussion
The HPLC method used in the present study for determination of plasma caffeine concentration was modified from methods used in previous studies by changing the length of the column (to 250 mm from 150 mm) and by increasing the buffer strength (to 20mM from 7.3mM phosphate at pH 3.0) and acetonitrile concentration (to 15% from 13%) in the mobile phase.13 Those modifications improved peak shape, retention time, and separation of the caffeine peak from interfering peaks.13 To reduce the effects of a strong base on column packing material, the volume of 4M NaOH solution that was added to each sample prior to injection was reduced (to 10 μL from 27 μL), compared with the volume of solution reported previously.14 Those changes were not associated with alterations in peak shape or retention time (9.5 to 10.4 minutes) of caffeine in alpaca and llama plasma, and there was no interference from early peaks associated with injection events. The accuracy and precision of the results of these determinations indicate that both HPLC and LC-MS analytic methods are suitable for analysis of caffeine in alpaca and llama plasma.
The pharmacokinetic variables determined by compartmental analysis of concentration-versus-time data for caffeine administered IV at a dose of 3 mg/kg were not significantly different between alpacas and llamas. Results also revealed that the pharmacokinetic properties of caffeine in alpacas and llamas were similar to disposition of the drug in other species.7–13 Values for total plasma caffeine clearance, elimination half-life, and mean residence time in adult alpacas and llamas were similar to those reported for horses,7,18 donkeys,18 sheep,13 cattle,9,13 and dogs.10 Values for volumes of distribution were similar to those reported for horses,7,18 donkeys,18 sheep,13 cattle,9,11 dogs,10 and camels.8 Those results suggest that rates of caffeine disposition and elimination were similar to those reported for other species with the exception of the camel. Values for total plasma caffeine clearance, elimination half-life, and mean retention time in camels differ substantially from values in alpacas, llamas, other domesticated livestock species, and dogs.8 In camels, the combination of lower total plasma clearance and a comparable volume of distribution results in longer elimination half-life.8
Variation among individual animals and in method may have limited detection of differences in caffeine clearance between alpacas and llamas. The small number of animals limited the power of the study for detection of differences between species.13,18 Variability among individuals in caffeine clearance is common in trials involving humans.19 It is also possible that variability was introduced by combining results obtained by use of different assays. The LC-MS assay is superior to the HPLC assay in terms of accuracy, precision, and limit of quantification, and accuracy and precision affect pharmacokinetic variables; however, because of multi-institutional participation, use of the same assay for all animals was not possible. Assessment of the contributions of animals and assay to the observed variance was beyond the scope of this study.
Similarity in caffeine clearance among healthy alpacas, llamas, and other species suggests that plasma clearance is dependent on functional hepatic cytochrome P450 enzyme systems. Caffeine is a capacity-limited drug in alpacas and llamas, as is true in other species. As such, the rate of caffeine elimination from plasma represents an index of functional liver metabolic capacity.11,13,18,20,21 This differs from elimination of drugs with blood flow–limited metabolic clearance (eg, indocyanine green), for which plasma clearance reflects the rate at which the compound is delivered to the liver.18,21
Caffeine is metabolized in hepatocytes via demethylation to related compounds (ie, theophylline, theobromine, and paraxanthine) that may be detected under similar conditions as those described13,18 for caffeine. The extent of metabolism (and therefore of plasma caffeine clearance) is similar among species, whereas the metabolic fate of caffeine may differ,13 possibly because of interspecies differences in expression of the cytochrome P450 enzymes responsible for caffeine metabolism.13,22 Caffeine clearance in sheep and cattle differs in the formation of paraxanthine and theobromine, both of which are metabolites of caffeine.13 Little information regarding drug metabolism pathways for South American camelids is available at present. Because expression of hepatic cytochrome P450 isoenzymes is species specific and the various isoforms of P450 are substrate selective, determination of caffeine clearance and metabolic fate of caffeine in camelids may improve our understanding of hepatic metabolism of other drugs in those species.13 For example, the metabolic fate of phenylbutazone in cattle is markedly different from that in a number of other species. The elimination half-life of phenylbutazone in mature dairy cows (40 hours) and bulls (62 hours) is longer than that in llamas (2 hours).23–26 Because phenylbutazone is metabolized primarily in the liver, species differences in metabolic pathways or phenylbutazone disposition may reflect differences in expression or activity of hepatic drug–metabolizing enzymes. Assessment of caffeine metabolite profiles after caffeine administration may prove useful in understanding which metabolic pathways are used for caffeine clearance by alpacas and llamas, information that has been determined in other species.13,18,19 Reduction in caffeine metabolite-caffeine ratios may be a more sensitive indicator of reduced hepatocyte function than caffeine clearance alone.19
Caffeine clearance is substantially reduced in humans with chronic liver disease19 and in dogs with experimental liver disease.20 Development of liver function tests such as caffeine clearance may prove useful for determining severity of hepatic injury, for modification of drug dosages, and in assessing response to treatment.20,27,28 Those assertions are supported by work in calves and goats, in which reduction in plasma clearance and prolongation of elimination half-life of caffeine, antipyrine, sulfadimidine, and paracetamol were observed after feed and water restriction.12,29,30 The decreased plasma clearance and increased elimination half-life observed in those capacity-limited drugs were proposed to result from changes in hepatocyte oxidative and conjugative metabolism.12,29,30 Less severe hepatic dysfunction, as is observed in camelids with transient inappetence, would be associated with less marked reductions in caffeine clearance.
ABBREVIATIONS
| HPLC | High-performance liquid chromatography |
| LC-MS | Liquid chromatography-mass spectrometry |
Caffeine sodium benzoate (C4144), Sigma-Aldrich Co, St Louis, Mo.
Caffeine (C8960), Sigma-Aldrich Co, St Louis, Mo.
Waters 626 LC System, Waters Corp, Milford, Mass.
Alltech Altima RP C18 column (25 × 0.46 cm; 5 μm), Alltech Associates, Deerfield, Ill.
Caffeine reference standard solution, 1.0 mg/mL, C6035, Sigma Chemical Co, St Louis, Mo.
Caffeine reference standard solution, 1.0 mg/mL, 017283, Alltech Associates, Deerfield, Ill.
8-chlorocaffeine (S80,332-4), Sigma Chemical Co, St Louis, Mo.
The Ohio State University Analytical Toxicology Laboratory, College of Veterinary Medicine, The Ohio State University, Columbus, Ohio.
Agilent, high-performance liquid chromatography/MSD-trap, Agilent Technologies, Palo Alto, Calif.
Zorbax Eclipse XDB-C18 (2.1 × 150 mm; 5 μm) analytic column, Agilent Technologies, Palo Alto, Calif.
WinNonlin Professional, version 4.1, Pharsight Corp, Mountain View, Calif.
References
- 1
Van Saun RJ, Callihan BR, Tornquist SJ. Nutritional support for treatment of hepatic lipidosis in a llama. J Am Vet Med Assoc 2000;217: 1531–1535.
- 2
Anderson DE, Constable PD & Yvorchuk KE, et al. Hyperlipemia and ketonuria in an alpaca and a llama. J Vet Intern Med 1994;8: 207–211.
- 3
Jonsson NN, Rozmanec M. Tick paralysis and hepatic lipidosis in a llama. Aust Vet J 1997;75: 250–253.
- 4
Semrad SD. Septicemic listeriosis, thrombocytopenia, blood parasitism, and hepatopathy in a llama. J Am Vet Med Assoc 1994;204: 213–215.
- 5↑
Tornquist SJ, Van Saun RJ & Smith BB, et al. Hepatic lipidosis in llamas and alpacas: 31 cases (1991–1997). J Am Vet Med Assoc 1999;214: 1368–1372.
- 6↑
Tornquist SJ, Cebra CK & van Saun RJ, et al. Metabolic changes and induction of hepatic lipidosis during feed restriction in llamas. Am J Vet Res 2001;62: 1081–1087.
- 7
Schumacher J, Spano JS & Wilson RC, et al. Caffeine clearance in the horse. Vet Res Commun 1994;18: 367–372.
- 8↑
Wasfi IA, Boni NS & Elghazali M, et al. The pharmacokinetics, metabolism and urinary detection time of caffeine in camels. Res Vet Sci 2000;69: 69–74.
- 9
DeGraves FJ, Ruffin DC & Duran SH, et al. Pharmacokinetics of caffeine in lactating dairy cows. Am J Vet Res 1995;56: 619–622.
- 10↑
Golden DL, Spano JS & Wilson RC, et al. Application of an enzyme-multiplied immunoassay technique for determination of caffeine elimination kinetics as a test of liver function in clinically normal dogs. Am J Vet Res 1994;55: 790–794.
- 11
Janus K, Antoszek J. The effect of sex and age on caffeine pharmacokinetics in cattle. Res Vet Sci 2000;69: 33–37.
- 12
Janus K, Antoszek J, Suszycki S. The effect of short-term starvation or water deprivation on caffeine pharmacokinetics in calves. Res Vet Sci 2001;70: 109–113.
- 13↑
Danielson TJ, Golsteyn LR. Systemic clearance and demethylation of caffeine in sheep and cattle. Drug Metab Dispos 1996;24: 1058–1061.
- 14↑
Holland DT, Godfredsen KA & Page T, et al. Simple high-performance liquid chromatography method for the simultaneous determination of serum caffeine and paraxanthine following rapid sample preparation. J Chromatogr B Biomed Sci Appl 1998;707: 105–110
- 15↑
Ermer J. Validation in pharmaceutical analysis. Part I: an integrated approach. J Pharm Biomed Anal 2001;24: 755–767.
- 16↑
Peck K, Mealey KL, Matthews NS, et al. In:Wernimont GT, Spendley W. eds.Use of statistics to develop and evaluate analytical methods. Gaithersburg, Md: AOAC International, 1985;11–30.
- 17↑
Akaike H. Information theory and an extension of the maximum likelihood principle, inProceedings. 2nd Int Symp Information Theory1973;267–281.
- 18↑
Peck K, Mealey KL & Matthews NS, et al. Comparative pharmacokinetics of caffeine and three metabolites in clinically normal horses and donkeys. Am J Vet Res 1997;58: 881–884.
- 19↑
Jodynis-Liebert J, Flieger J & Matuszewska A, et al. Serum metabolite/caffeine ratios as a test for liver function. J Clin Pharmacol 2004;44: 338–347.
- 20↑
Boothe DM, Cullen JM & Calvin JA, et al. Antipyrine and caffeine dispositions in clinically normal dogs and dogs with progressive liver disease. Am J Vet Res 1994;55: 254–261.
- 21
Greene EW, Woods WE, Tobin T. Pharmacology, pharmacokinetics, and behavioral effects of caffeine in horses. Am J Vet Res 1983;44: 57–63.
- 22
Chauret N, Gauthier A & Martin J, et al. In vitro comparison of cytochrome P450-mediated metabolic activities in human, dog, cat, and horse. Drug Metab Dispos 1997;25: 1130–1136.
- 23
Eberhardson B, Olsson G & Applegrew L, et al. Pharmacokinetic studies of phenylbutazone in cattle. J Vet Pharmacol Ther 1979;2: 31–37.
- 24
Williams RJ, Boudinot FD & Smith JA, et al. Pharmacokinetics of phenylbutazone in mature Holstein bulls: steady-state kinetics after multiple oral dosing. Am J Vet Res 1990;51: 371–375.
- 25
Williams RJ, Smith JA & Boudinot FD, et al. Pharmacokinetics of phenylbutazone given intravenously or orally in mature Holstein bulls. Am J Vet Res 1990;51: 367–370.
- 26
Navarre CB, Ravis WR & Nagilla R, et al. Pharmacokinetics of phenylbutazone in llamas following single intravenous and oral doses. J Vet Pharmacol Ther 2001;24: 227–231.
- 27
Cornelius CE. A review of new approaches to assessing hepatic function in animals. Vet Res Commun 1987;11: 423–441.
- 28
Tennant B. Liver function. In:Kaneko JJ, Harvey JW, Bruss ML, eds.Clinical biochemistry of domestic animals. San Diego: Academic Press Inc, 1997;243–261.
- 29
Elsheikh HA, Osman IA, Abdullah AS. The effect of water deprivation on the pharmacokinetics of antipyrine and sulphadimidine following intravenous administration in nubian goats. Vet Res Commun 1997;21: 587–597.
- 30
Janus K, Grochowina B & Antoszek J, et al. The effect of food or water deprivation on paracetamol pharmacokinetics in calves. J Vet Pharmacol Ther 2003;26: 291–296.
