Effect of Injection Site on Carbenicillin Pharmacokinetics in the Carpet Python, Morelia spilota

Transcription

1 Effect of Injection Site on Carbenicillin Pharmacokinetics in the Carpet Python, Morelia spilota Peter H. Holz1, BVSc, DVSc, MACVSc, DACZM, John P. Burger2, BA, BSc, Kirby Pasloske DVM3, DVSc, DACVCP, Rupert Baker4, BVSc, MACVS, MRCVS, Sam Young5, BVSc (Hons), BSc (VetHons), 1. Healesville Sanctuary, PO Box 248, Healesville, Victoria 3777, Australia 2. Department of Biomedical Sciences, Ontario Veterinary College, University of Guelph, Guelph, Ontario N IG 2W1, Canada S. Havens Drive, New Palestine, ID 46163, USA 4. Healesville Sanctuary, PO Box 248, Healesville, Victoria 3777, Australia. Current address: Clockhouse Veterinary Hospital, Wallbridge, Stroud GL5 3JD, United Kingdom 5. Healesville Sanctuary, PO Box 248, Healesville, Victoria 3777, Australia. Current address: Royal Melbourne Zoological Gardens, Elliott Avenue, Parkville, Victoria 3052, Australia A bstract: The premise that drugs should not be injected into the caudal body of reptiles because they will be carried by the renal portal system to the kidneys and rapidly excreted was tested by comparing the pharmacokinetics of carbenicillin in carpet pythons, Morelia spilota, injected anteriorly with those injected posteriorly. Seven carpet pythons were treated with 200 mg/kg carbenicillin administered intramuscularly. Three were injected anteriorly and four posteriorly. Serial blood samples were withdrawn over a period of 120 hr. Carbenicillin blood levels were measured over each time period and results used to calculate maximum carbenicillin plasma level, time of maximum plasma level, terminal half-life, area under the curve, volume of distribution and total body clearance. The study was repeated five months later with those snakes that were previously injected anteriorly now being injected posteriorly and vice versa. Blood was taken and the same parameters calculated. Following all calculations no significant differences in pharmacologic parameters were found based on injection site. Dissection of three dead carpet pythons determined that renal portal anatomy was the same as that described for other python species. Radio-opaque dye was injected in the caudal vein of one python and observed radiographically to pass via the afferent renal portal veins through the kidneys and enter the efferent renal portal veins. K ey W ords: carpet python, Morelia spilota, carbenicillin, renal portal system. INTRODUCTION The renal portal system (RPS) is present in all birds, reptiles, amphibians, and most fish. Its structure and function was reviewed recently (Holz, 1999). Reptiles do not have a Loop of Henle and cannot produce hypertonic urine. In order to conserve water during times of water deprivation glomerular filtration ceases. Renal portal blood continues to perfuse the tubule cells during this period and prevents ischemic necrosis. The RPS conveys venous blood from the caudal regions of the body to the renal tubule cells. From there the blood continues to the heart. It is possible for blood to flow through the kidneys or bypass them. Controversy has surrounded the RPS, with a number of authors stating that drugs should not be injected into the caudal region of reptiles as the RPS will convey them directly to the kidneys resulting in their premature excretion (Gauvin, 1993, Mautino and Page, 1993, McCracken, 1994). However, more recent work has shed doubt on this theory, with studies using gentam icin and carbenicillin in red-eared sliders, Trachemys scripta elegans, (Holz, et al, 1997b) and gentamicin in box turtles, Terrapene Carolina Carolina, (Beck, et al, 1995) showing no significant effect of injection site on drug kinetics. Furthermore, a recent radiographic study in green iguanas, Iguana iguana, has demonstrated that most of the blood returning from the hind leg bypasses the kidney and enters the general circulation (Benson and Forrest, 1999). Drug studies thus far have only examined the effect of injection site on drug pharmacokinetics in chelonians. The current study was devised to provide further information from a different family of reptiles. Carbenicillin was selected to enable direct comparisons to be made between the red-eared slider and the carpet python studies and because it is excreted, at least partially, by tubular secretion (Kucers and Bennett, 1979), making any renal portal effects more likely. METHODS The study used seven long term captive adult carpet pythons (four male and three female) of unknown age weighing 1.43 to 7.90 kg. T hey w ere housed in enclosures containing a heat pad. Ambient temperature varied from 22 C to 30 C (72 to 86 F) in the enclosures, and up to 35 C (95 F) on the heat pad. Pythons were maintained on a diet of thawed rats and mice. Three pythons were randomly assigned to receive an intramuscular injection of 200 mg/kg carbenicillin disodium (Agri-Bio, North Miami, FL, USA) in the anterior epaxial muscles approximately 15 cm posterior to the head, Journal of Herpetological Medicine and Surgery Volume 12, No. 4,2002

2 while the remaining four received the same dose in the epaxial muscles approximately 8 cm anterior to the cloaca. The study was repeated on the same pythons five months later, with the snakes that had previously been treated in the anteri or part of the body now being injected in the posterior part of the body, and vice versa. This way each python served as its own control. A previous study in carpet pythons indicated that the kidneys would be situated anterior to the posterior injec tion site (McCracken, 1999). Blood (1 ml) was withdrawn from the caudal vein using a one-inch, 22 ga needle and 1 ml syringe prior to injecting the antibiotic. Repeat blood samples were collected 0.5, 1, 2, 4, 8, 12, 24, 48, 72, 96, and 120 hr post injection and placed into 2 ml EDTA vacutainers (Greiner Labortechnik, Kremsmiinster, Austria). Blood was centrifuged, the plasm a removed and stored in 3 ml plastic vials (Sarstedt, Australia) at approxi mately -20 C until assay. Plasma carbenicillin levels were determined by a microbio logical diffusion assay (Arret, et al, 1971). Antibiotic base medium 9 and seed m edium 10 was obtained from Difco Laboratories, Detroit, Michigan, USA. Pseudomonas aerugi nosa was purchased from A m erican Type Culture Collection, Manassas, VA, USA. For the standard curve serial dilutions of carbenicillin dis odium were prepared ranging from 1000 g/ml down to g/ml. The median drug concentration, 125 g/ml, was used as the reference standard. Due to the difficulty in obtaining large volumes of python plasma, the drug standards were prepared in both pooled python and bovine drug free plasma and the standard curves compared. The dose responses were similar and consistent with the findings of Lawrence, et al (1984a). Bovine drug free plasma was subsequently used to prepare all drug standards. The inter-assay standard curve was linear from g/ml to 1000 g/ml on a semi-log plot (r>0.9954) and dilutions of plasma samples were not needed. On the day of drug analysis, 13 :1 of sample or drug stan dard were transferred to 6.35 mm ( 1 / 4 inch) diam eter filter paper discs (Schleichef & Schuell, Keene, NH, USA) in a Petri dish containing seed and base media. Samples and stan dards were incubated in triplicate for an average of 20 hr (+/30 minutes) at 37 C (99 F). Inhibition zone diameters were measured immediately after the incubation period. Standard curve inhibition zone diameters ranged from approximately 8 mm to 25 mm. Inter-assay variation, as measured by differ en ces in z o n e d ia m e te r in h ib ite d b y th e sam e drug concentration, did not exceed 10%. Noncompartmental pharmacokinetic parameters were cal c u la te d u sin g W in N o n lin (V ersio n 3.1, P h a rs ig h t Corporation, M ountain View, CA) computer software. The param eters calculated and their associated formulas were Cmax (maximum plasm a level achieved), Tmax (time when maximum plasma level was achieved), 9 and ti/2 lambda (ter minal half-life; ti/2 lambda = In(2)/9), AUCo-~ (integration of the under the plasma concentration curve from the time of dosing to infinity;auco-~ = AUClast + Clast/9), Vd (volume of distribution based on the terminal phase; Vd = dose/9* AUCo-~) and Cl (total body clearance for extravascular administration; Cl = dose/ AUC0-8). An ANOVA test was used to test for sig nificant differences (p < 0.05) in these pharm acokinetic parameters following injection of the drugs in the fore- or hind-limb. After the determination of pharmacokinetic para Volume 12, No. 4,2002 meters, doses for clinical use were established using the fol lowing equations: Cmaxss = (D/Vd) x (l/l-e -^ ) and Cminss = Cmax x e--^ (where CmaxSs is the desired maximum steady state concentration in pg/ml, Cminss is the desired minimum steady-state concentration in pg/ml, D is the dose in mg/kg, Vd is the volume of distribution in ml/kg, X is the elimination rate constant in hr ~l, and x is the dosing interval in hours). All parameters and equations were calculated assuming 100% bioavailability of drug from both injection sites and in all ani mals treated. Three dead carpet pythons, which had not been involved in the pharmacokinetic study, were dissected and the anatomy of the RPS mapped. In addition, one live carpet python was anesthetised w ith isoflurane (ISO, V eterinary C om panies of Australia Pty. Ltd., Artarmon, NSW, Australia), and then its caudal vein was injected with a radiopaque dye (meglumine iothalamate 280 mg/ml, Conray 280, Mallinckrodt Medical Pty Ltd, Nottinghill, Victoria, Australia) at a dose of 560 mg/kg. A ventro-dorsal radiograph was taken immediately fol lowing injection to observe the distribution route of the dye. RESULTS The average plasma carbenicillin levels over time for the two injection sites are shown in Figure 1. Forty-eight hours after one dose of carbenicillin mean plasma drug levels were below the limit of detection (15.62 g/ml). The average maxi mum drug concentration and the time when it was achieved, the plasma elimination half-life, area under the curve, clear ance, and volume of distribution are presented in Table 1. Injection site (anterior versus posterior) was not associated with a significant difference in any of the pharmacologic parameters examined. From steady state calculations it can be shown that a 200 mg/kg intramuscular injection of carbeni cillin every 24 hr will theoretically result in steady state peaks of 505 g/ml and 485 g/ml, and steady state troughs of 50 g/ml and 115 g/ml after anterior and posterior injections respec tively. The latter trough is higher due to an average longer plasma elimination half-life. However, because penicillins are considered time-dependent antibiotics, it is probably more important to keep the plasma concentration above the MIC for the entire dosing period. For com parative purposes, results of a previously pub lished study in red-eared sliders (Holz, et al, 1997b) are also presented in Table 1. The renal portal anatomy of carpet pythons is the same as that described for other snake species (Beddard, 1906, Holz, 1999) (Figure 2). Briefly, the caudal vein bifurcates as it leaves the tail and forms the two afferent renal portal veins, which perfuse the two kidneys. Portal blood re-emerges in the two efferent renal portal veins, which unite to form the post caval vein that transports blood back to the heart. Several branches arise from the afferent renal portal veins. These unite to form a single mesenteric vein, which transports blood around the kidneys to the liver. The abdominal vein has its origin in the fat body. The contrast study showed dye flowing in two streams, in the two afferent renal portal veins, to the kidneys (Figure 3). The dye entered and highlighted both kidneys, the left kidney lying posterior to the right kidney as described previously (McCracken, 1999). It then emerged in the two efferent renal Journal of Herpetological Medicine and Surgery 13

3 portal veins. These two veins united to form the post caval vein. A small quantity of dye entered the mesenteric vein, and this can be seen as a separate faint line adjacent to the post caval vein (Figure 3). DISCUSSION Figure 1. Plasma carbenicillin concentration (mean ± SD; n = 7) after injection (200 mg/kg IM) into anterior and posterior injection sites in carpet pythons. Figure 2. Renal portal system of the Order Squamata, Suborder Serpentes (dorsal view). Arrows indicate direction of blood flow. The oval, stippled organs represent the kidneys. EA = Extremitatenanastomose. 1 = Origin of the abdominal vein from the afferent renal portal vein, found in the African rock python. 2 = Origin of the abdominal vein from the extremitatenanastomose, found in the Indian python. This study demonstrated no significant effect of injection site on pharmacokinetic parameters in carpet pythons injected intram uscularly with 200 mg/kg carbenicillin. When this study was performed in red-eared sliders a statistically significant difference (p<0.05) did occur in the area under the curve (AUC) between injection sites over the first eight hours (Holz, et al, 1997b) with the anterior injection having a higher AUC than the posterior injection site. In the present study, although there was no statistical difference between injection sites and corresponding average AUCs, mean blood levels were higher initially in snakes injected anteriorly compared to snakes injected posteriorly (Figure 1). Regardless of whether the anterior injection site AUC was statistically larger than the posterior injection site AUC in either red-eared sliders or carpet pythons, plasma carbenicillin levels were still adequate to treat susceptible micro-organisms when injected once daily at 200 mg/kg. The present study demonstrated a large range in pharmacokinetics between snakes, independent of injection site. Figure 1 illustrates the high degree of drug concentration variability at each time point while Table 1 demonstrates the large standard d ev iatio n s fo r each p h arm aco k in etic param eter (pharmacokinetic parameters are generated for each animal independently and the means and standard deviations calculated). One exam ple of the large ratige in drug kinetics between snakes is the carbenicillin elimination half-life in the anterior study (3.5 to 34 hours) and the posterior study (5.5 to 32 hr). A second example involves the AUCs that are directly calculated from each animal s drug concentration versus time curve. One snake had a three-fold less AUC than any other snake in the anterior injection site study while another snake had a ten fold lower AUC than any other snake in the posterior injection study. The AUC param eter is also used to calculate volume of distribution (Vd) and clearance (Cl). So it is no surprise that there were no differences found between pharm acokinetic param eters and injection sites. Another caveat is that Vd and Cl assume carbenicillin bioavailability is equal to 100% for every injected animal. If even one animal is dosed incorrectly, AUC, Vd, and Cl for that animal could differ trem endously from the rest because bioavailability could be much less than 100%. Large pharmacokinetic differences between snakes could also result from individual variations in blood flow and absorption from injection sites (Hilf, et al, 1991), as well as fluctuations in ambient temperature. Temperature could vary by as much as 13 C depending on whether or not the python had been lying on the heat pad prior to sampling. Previous studies have shown pharmacokinetics in reptiles to be temperature sensitive (Mader, et al, 1985, Caligiuri, et al, 1990). Pythons in this study were of unknown age. Given the large variation in size it is likely that ages differed widely also. In humans plasma carbenicillin half life increases by about 25% between the ages of 30 and 70 (Katzung, 1998). 14 Journal of Herpetological Medicine and Surgery Volume 12, No. 4,2002

4 A previous reptile study found carbenicillin sensitive organism s to have a M inimum Inhibitory Concentration (MIC) of 50 g/ml or less (Lawrence, et al, 1984a). It is generally regarded that effective therapy requires a peak blood level four times MIC (Conzelman, 1980), ie 200 g/ml or more. Blood levels were maintained above therapeutic concentrations for the same period of time in both anterior and posterior injection groups (Figure 1). Carpet python pharmacokinetic parameters varied from those described in the sliders (Table 1) and also for another study that injected 400 mg/kg carbenicillin (twice the dose used in this study) into a variety of snakes (Lawrence, et al, 1984b). Peak carbenicillin blood levels and total drug absorption tended to be lower in the pythons than the sliders but greater in the pythons than the other snakes. This may be due to species variation with respect to drug liberation, absorption, metabolism, distribution, and elimination and has been similarly described for enrofloxacin kinetics in two chelonian species (Prezant, etal, 1994, Raphael, et al, 1994). The radiographic study demonstrated that much of the contrast medium perfused the kidneys when injected into the caudal vein. Presum ably this would also result in drugs injected caudally passing through the kidneys. However, it has been demonstrated in red-eared sliders that blood flow can be altered to pass through or around the kidneys (Holz, et al, 1997a). As a small amount of contrast bypassed the kidneys in the carpet python shunting of blood is also possible in this species. The m ethod of controlling blood flow is unknown at this stage and requires further investigation. In conclusion, injection site does not significantly affect carbenicillin pharmacokinetics in carpet pythons. A dose of 200 mg/kg carbenicillin every 24 hr will provide therapeutic blood levels regardless of site of administration. For the RPS to have a major effect on drug systemic exposure, a large portion of the drug would have to be eliminated by the kidneys Figure 3. Ventro-dorsal radiograph of carpet python injected with radiopaque dye (meglumine iothalamate, 280 mg/ml; 2 ml/kg ) into the caudal vein. A = Afferent renal portal veins. E = Efferent renal portal veins. K = Kidney. M = Mesenteric vein. P = Post caval vein. by first pass effects. Once past this physiological barrier the drug is in the systemic circulation and injection site is irrelevant. This study failed to demonstrate significant carbenicillin elimination after a first pass through the kidneys and further supports the hypothesis that drug kinetics are unaffected by injection site in reptiles. ACKNOWLEDGEMENTS This project was m ade possible by a grant from the Association of Reptilian and Amphibian Veterinarians. The authors acknowledge the assistance of Janine Ferguson, Rod Hare, Byron Manning, David Middleton, Marcia Standfield and Michael Taylor, without whose involvement this project could not have been completed. We would also like to thank M ark Papich for advice on pharm acokinetics and Anne Valliant for assistance with statistical analysis. REFERENCES Arret B, Johnson DP, Kirshbaum A Outline of details for microbiological assays of antibiotics: second revision. J Pharmacol Sci, 60: Beck K, Loomis M, Lewbart G, Spelman L, Papich M Preliminary comparison of plasma concentrations of gentamicin injected into the cranial and caudal limb musculature of the eastern box turtle (Terrapene Carolina Carolina). J Zoo Wildl Med, 26: Beddard FE Contributions to the anatomy of the Ophidia. Proc Zool Soc, Benson KG, Forrest L Characterization of the renal portal system of the common green iguana (Iguana iguana) by digital subtraction imaging. J Zoo Wildl Med, 30: Calilgiuri R, Kollias GV, Jacobson E, McNab B, Clark CH, Wilson RC The effects of ambient temperature on amikacin pharmacokinetics in gopher tortoises. J Vet Pharmacol Therap, 13: Conzelman GM Pharmacotherapeutics of aminoglycoside antibiotics. JAVMA, 176: Gauvin J Drug therapy in reptiles. Semin Avian Exotic Pet Med, 2: H ilf M, Swanson D, W agner R, Yu VL Pharmacokinetics of piperacilllin in blood pythons (Python curtus) and in vitro evaluation of efficacy against aerobic gram-negative bacteria. J Zoo Wildl Med, 22: Holz P, Barker IK, Crawshaw GJ, Dobson H. 1997a. The anatomy and perfusion of the renal portal system in the red-eared slider (Trachemys scripta elegans). J Zoo Wildl Med, 28: Holz P, Barker IK, Burger JP, Crawshaw GJ, Conlon PD. 1997b. The effect of the renal portal system on pharmacokinetic parameters in the red-eared slider (Trachemys scripta elegans). J Zoo Wildl Med, 28: Holz PH The reptilian renal portal system - a review. Bull ARAV, 9:4-9. Katzung BG Special aspects of geriatric pharmacology. In Katzung BG (ed): Basic and Clinical Pharmacology, 7th ed. Appleton and Lange, Stamford, CT, USA Kucers A, Bennett NM The Use of Antibiotics, 3rd ed. William Heinemann Medical Books, London, UK. Volume 12, No. 4,2002 Journal of Herpetological Medicine and Surgery 15

5 Law rence K, M uggleton PW, Needham JR. 1984a. Preliminary study on the use of ceftazidime, a broad spectrum cephalosporin antibiotic, in snakes. Res Vet Sci, 36: Lawrence K, Needham JR, Palmer GH, Lewis JC. 1984b. A preliminary study on the use of carbenicillin in snakes. J Vet Pharmacol Therap, 7: Mader DR, Conzelman Jr GM, Baggot JD Effects of ambient temperature on the half-life and dosage regimen of amikacin in the gopher snake. JAVMA, 187: Mautino M, Page CD Biology and medicine of turtles and tortoises. Vet Clin North Am Sm Anim Pract, 23: McCracken H Husbandry and diseases of captive reptiles. Proc 233 Post Graduate Committee in Vet Sci, Univ Sydney McCracken HE Organ location in snakes for diagnostic and surgical evaluation. In Fowler ME, Miller RE (eds): Zoo and Wild Animal Medicine: Current Therapy 4. WB Saunders Co, Philadelphia, PA: Prezant RM, Isaza R, Jacobson ER Plasma concentrations and disposition kinetics of enrofloxacin in gopher tortoises (Gopheruspolyphemus). J Zoo Wildl Med, 25: Raphael BL, Papich M, Cook RA Pharmacokinetics of enrofloxacin after a single intramuscular injection in Indian star tortoises (Geochelone elegans). J Zoo Wildl Med, 25: Table 1. Mean pharmacokinetic parameters (+/- SD) calculated following intramuscular injection of carbenicillin at 200 mg/kg in either the anterior or posterior part of the body of carpet pythons and red-eared sliders (Holz, et al, 1997b). Anterior Injection Site Parameter Carpet Python, Morelia spilota Red-eared Slider, Trachemys scripta elegans Tmax (hr) / /-1.28 Cmax (g/ml) / / Ti/2 (hr) / /-7.65 AUCo-i hr (g/hr/ml) / / AUCq.2 hr (g/hr/ml) / / AUC0.4 hr (g/hr/ml) / / AUCo.g Hr (g/hr/ml) / / AUCq^, (g/hr/ml) / / Cl (ml/min/kg) / /-0.22 Vd (l/kg) 0.44+/ /-0.05 Posterior Injection Site Parameter Carpet Python, Morelia spilota Red-eared Slider, Trachemys scripta elegans Tmax (hr) / / Cmax (g/ml) / / Ti/2 (hr) / /-8.07 AUCo-1 hr (g/hr/ml) / / AUCq.2 hr (g/hr/ml) / / AUC0.4hr (g/hr/ml) / / AUC0.g hr (g/hr/ml) / / AUC0., (g/hr/ml) / / Cl (ml/min/kg) / / Vd (l/kg) 0.54+/ / Tmax = Time, after injection, when maximum plasma drug levels were reached. Cmax = Maximum plasma drug level achieved. Ti/2 = Drug elimination half-life. AUC0.n hr = Area under the curve calculated 1,2,4, or 8 hr postinjection. Cl = Total drug clearance. Vd = Volume of distribution. Journal of Herpetological Medicine and Surgery Volume 12, No. 4,2002