RENAL CLEARANCE OF CARPROFEN IN THE ISOLATED PERFUSED RAT KIDNEY

9-9556/81/91-25$2./ DRG MrrABLI5M AND DlsPosmoN Copyright 1981 by The American Society for Pharmacology and Experimental Therapeutics Vol. 9, No. I Printed in. S. A. RENAL CLEARANCE OF CARPROFEN IN THE ISOLATED PERFSED RAT KIDNEY IHOR BEKERSKY AND WAYNE A. COLBRN Department of Pharmacokinetics and Biopharmaceutics Hoffmann-La Roche Inc. (Received September 29, 198) ABSTRACT: The renal clearance of the anti-inflammatory agent, carprofen, was calculation of the tubular transit rate at each carprofen concentration indicated the following net mechanism(s) of renal carprofen studied in the isolated perfused rat kidney (IPK). The dosing range used (.5-25 mg) produced perfusate concentrations comparable clearance: filtration at low concentration, reabsorption at intermediate concentrations, and secretion at the highest concentration. At to and greater than therapeutic plasma concentrations expected in man. In vivo studies in the rat were done as a basis for comparison low urinary ph and flow rates, reabsorption effectively counteracted to the In vitro IPK parameters. Because of its extensive binding secretion. The present in vitro studies suggest that carprofen is (>99%) to the protein fraction of the perfusate, the urinary excretion excreted by the kidney by the pathway common to a variety of of carprofen was low and perfusate The organic acids. The isolated perfused kidney (IPK) is an established model for studying the renal disposition of drugs and other compounds (1). Experiments utilizing this in vitro preparation provide data on renal excretion mechanisms, i.e., filtration, secretion, and reabsorption, that are consistent with those of in vivo experiments (2, 3). For drugs that are not extensively bound to plasma/perfusate protein, filtration of the free (unbound) drug contributes significantly to the elimination process. Nonfiltered (bound) drug may be cleared by secretion, which is a function of total drug concentration rather than of the free drug concentration in plasma! perfusate. Filtered and secreted drug is subject to reabsorption, which is a function of urine ph and flow rate. Carprofen, racemic 6-chloro-a-methylcarbazole-2-acetic acid is a nonsteroidal anti-inflammatory agent. It is a weak acid (pka 4.7), has low aqueous solubility, and is bound in excess of 99% to plasma proteins ofrat, dog, and man. The compound is efficacious in the treatment of rheumatoid arthritis and gout and is undergoing further investigation in the clinic (4). Carprofen is eliminated primarily by metabolism in the rat, dog, and man (5). Biliary secretion followed by excretion in the feces accounts for approximately 7% of the total radioactivity following an iv dose of 14C-labeled compound in the rat and dog, whereas 1-3% is excreted in the urine. In man, elimination occurs primarily through direct conjugation of carprofen to form the glucuronide ester. Approximately 4% of an oral 4C dose is secreted in human bile; however, only 2% is excreted in the feces because of enterohepatic recycling and subsequent excretion in the urine (5). The major component in human urine is carprofen glucuronide. Although direct renal excretion of intact carprofen plays only a minor role in the overall elimination of this compound in rat, I Abbreviations used are: IPK, isolated perfused kidney; F,, free fraction; CB, buffer concentration; C,., perfusate concentration; GFR, glomerular filtration rate; T, tubular transit rate; C, urinary concentration; Q, urine flow rate; W, water fraction of perfusate; K/P. kidney-to-perfusate concentration ratio. Send reprint requests to: Dr. lhor Bekersky, Department of Pharmacokinetics and Biopharmaceutics, Hoflmann-La Roche, Inc., Nutley. N. J. 71 1. dog, and man (<5%), the mechanisms that influence the renal excretion of carprofen as well as other highly plasma-bound substances are of interest. The purpose of the present experiment was to utilize the IPK to study the mechanism(s) of the renal excretion of carprofen as affected by escalating doses. Methods Perfusate binding of carprofen was determined in quadruplicate in the Spectrum#{174} equilibrium dialyzer with 1-ml cells. Carprofen was dissolved in the perfusate at concentrations of 1, 1, and 3 g/m1 and the perfusate was dialyzed for 4 hr at 37#{176}Cagainst an equal volume of Krebs- Henseleit buffer (ph 7.45). Aliquots of perfusate and buffer were then assayed for carprofen by HPLC. The free fraction (Fr) was calculated according to the equation Ff = CB/CP where CB is the concentration of carprofen in buffer and C is its concentration in perfusate. Male Sprague-Dawley rats (3 g) were used to determine the in vivo renal clearance of carprofen. Both ureters were cannulated with PE- 1 tubing, and urine was collected at 15-mm intervals. The jugular vein was cannulated for dose administration and subsequent blood sampling (6). Carprofen (5 mg/kg) was administered iv along with inulin (mixed with trace amounts of 1H3]inulin) as a short infusion over I mm to four rats. Blood samples (.3 ml) were collected in heparin-treated tubes. Blood was replaced by normal saline immediately after each sampling to maintain hydration and consistent urine flow. Plasma was separated and aliquots of plasma and urine were assayed for inulin by liquid scintillation and for carprofen by HPLC. Male Sprague-Dawley rats (325-375 g) were used in the IPK experiments, which were performed as previously reported (7) by a modification of a method described by Bowman (1). The initial volume of perfusate used in the recirculating system of each experiment was 65 ml. Perfusion experiments were carried out for eleven 1-mm periods. The first two periods (control) were used to establish normal values for renal function. Carprofen (.5-25 mg) was dissolved in perfusate (.5-1. ml) and added as a 1-min short infusion into the circulating perfusate at 2 mm. The ph of each urine was determined immediately upon collection and the volume determined gravimetrically. Perfusate was sampled (.3 ml) at the midpoint of each lo-min urine collection interval. rine and perfusate samples were assayed for glucose, inulin, Na, and K to monitor the viability of the preparation. Glucose was determined in a Beckman glucose analyzer. Inulin was mixed with trace amounts of (3H]inulin (1. mci/7.3 mg, New England Nuclear, Boston, Mass.) to determine the GFR by liquid scintillation in a Searle Mark IV counter. Concentrations of Na and K were determined with a model 143 flame photometer (lnstrumen- Downloaded from dmd.aspetjournals.org at ASPET Journals on March 5, 216 25

26 BEKERSKY AND COLBRN tation Laboratories, Inc., Boston, Mass.). Suitable aliquots of urine and the corresponding perfusate were assayed for carprofen by HPLC with fluorescence detection (8). At the completion of each IPK experiment a final perfusate sample was taken, and the perfused kidney was removed from the system, blotted, weighed, and frozen. The tissue and final perfusate samples were assayed for carprofen to calculate the kidney-to-perfusate (K/P) ratios. The renal clearance of carprofen was calculated according to the conventional equation, Cl = A.Xu/Cp(mid), 11] where the amount excreted in urine, X,,, during the collection interval is divided by the concentration in plasma at the midpoint of the urinecollection interval Cp(mid). This calculation does not take into account the fraction of carprofen which is bound to perfusate protein. Because only free (unbound) drug is subject to filtration, the clearance values of carprofen were also corrected for the free fraction of carprofen in perfusate to determine if mechanisms other than filtration influence the net excretion of carprofen. To calculate whether clearance was the result of simple filtration, net tubular secretion, or net reabsorption, the tubular transit rate (T) was calculated according to the formula, T = CQ, - CF1W(GFR), [2] where C is the concentration of carprofen in urine, Q is the urine flow rate, C is the concentration ofcarprofen in perfusate, F is the free fraction of carprofen in perfusate, W is the fraction of perfusate which is water, and GFR is the glomerular filtration rate. A positive value for T indicates net tubular secretion, a negative value indicates net tubular reabsorption, and a value of zero would indicate that net excretion is equivalent to filtration. Results Perfusate binding of carprofen was determined by equilibrium dialysis after the in vitro addition ofcarprofen to control perfusate. The results of these studies are reported in table 1. Carprofen is very highly bound in perfusate with the free fraction increasing with increasing carprofen concentrations. High binding resulted in low clearance, as the contribution of filtration to elimination was sharply curtailed. Mean carprofen concentrations in plasma following a 5-mg!kg iv dose ranged from 5.4 to 19. sg,/ml. These concentrations are similar to those observed in perfusate following the 5- and.5-mg doses to the IPK (83 and 8.3 zg, ml, respectively). The results from the in vivo experiments (table 2) indicate a low urinary excretion rate and clearance of carprofen, and are presented to provide a comparison for subsequent IPK results. The functionality parameters of the IPK preparation and carprofen transit in the IPK system for each of the four doses of carprofen are presented in table 3. The functionality parameters are similar to those previously presented for systems using a 5% BSA perfusate (4-6). The carprofen!inulin clearance ratio for all doses is considerably less than unity and is consistent with the low TABLE 1 Perfusate binding of carprofen Data represent the means ± SD of quadruplicate determinatio each concentration. The F1 of carprofen in rat plasma at a concenir of 22 g/m1 was.14 ±.2 (N = 4). Concentration Perfusate Free Fraction big/mi 1.26 ±.8 1.54 ±.3 3.62 ±.5 as at ation TABLE 2 Clearance of carprofen in the ureter-cannulated rat after iv administration of5 mg ofcarprofen per kg Data are reported as means ± SD of four experiments. Plasma rinary Time Concentration Excretion mmn pg/mi g/15 mm d/min 5 5.4 ± 7.4 1 38.4 ± 4.7 15 33.9 ± 4.5.43 ±.48 3 26.8 ± 3..53 ±.56.91 ±.96.26 ±.29 45 22.5 ± 3.6.76 ±.76 1.77 ± 1.95.42 ±.47 6 21.2 ± 1.8.9 ±.82 2.69 ± 2.64.64 ±.62 75 19.2 ± 2.8.93 ±.83 2.94 ± 2.74.7 ±.65 9 19. ± 2.1 1.27 ± 1.32 4.25 ± 4.13.12 ±.98 Mean 2.55.61 Total (jig) 4.82 ii Mean of 2 determinations. percentage of dose excreted (table 2) and low urinary excretion rates shown in fig. 1. The net T for carprofen varied with dose from essentially zero (.5 mg) and negative (5 and 12.5 mg) to strongly positive (25 mg). This indicates that, dependent on dose, carprofen was undergoing filtration, net reabsorption, or net secretion. The rate ofurinary excretion ofcarprofen from the IPK (plotted as cumulative amounts excreted) after.5- to 25-mg doses are presented in fig. 1. Inasmuch as the urinary excretion of carprofen is low, perfusate concentrations did not decrease significantly during the experiment, resulting in apparent linear excretion of carprofen during the study period. Carprofen clearance was dependent upon the urinary ph (fig. 2) and the urinary flow rate (fig.3). At the highest dose of carprofen (25 mg), urinary ph was consistently higher than that observed after the lower doses (.5-12.5 mg), causing a shift in the clearance, urinary flow-rate relationship. The clearance ofcarprofen during the IPK experiments for each dose is shown in table 4. The renal clearance of carprofen was much less than the GFR for all doses of carprofen administered. However, when the renal clearance values for carprofen were corrected for the free fraction in perfusate, carprofen clearance was greater than the GFR for the.5- and 25-mg doses but substantially below the GFR for the 5- and 12.5-mg doses. These differences presumably reflect the dependence of carprofen elmination on the interrelationship of urinary ph and flow rate. At the end of each experiment, a fmal perfusate concentration (P) and the kidney tissue concentrations (K) of carprofen were measured (table 5). The K/P ratio for carprofen was below unity, indicating that movement of carprofen into kidney tissue was limited. This is consistent with low tissue build-up due to extensive perfusate binding. However when the perfusate carprofen concentrations are corrected for Ff, K/P ratios greater than unity are found, indicating uptake by kidney tissue. Discussion The urinary excretion rate of carprofen in vivo (table 2) and the IPK (fig. 1) was very low. Renal clearance calculations showed that carprofen clearance was only a fraction of the GFR (tables 2 and 4). Carprofen is extensively bound in plasma and in the BSA perfusate, and F is dependent on the concentration; i.e., Ff increased with increasing concentration. High binding makes carprofen largely unavailable for filtration. The fraction of drug in Downloaded from dmd.aspetjournals.org at ASPET Journals on March 5, 216

RENAL CLEARANCE OF CARPROFEN 27 TABLE 3 Functionality and tubular transit rate of carprofen in the isolatedperfused kidney Data are reported as means ± SD. N, number of experiments, with nine determinations of each parameter per experiment. - Value at Carp rofen Dose (mg) Parameter.5 5 12.5 25 Q. gsl/min 79. 1 ± 33.7 45. 1 ± 16.3 8.2 ± 44.3 47.5 ± 2. ph 6.68 ±.2 6.2 ±.6 6.5 ±.3 7.15 ±.9 Cij/Cp, inulin 5.2 ± 2.3 8.9 ± 1.9 1 1.7 ± 3.3 8. ± 2.3 Cu/Cr, carprofen.23 ±.12.2 1 ±.1 1.14 ±.5.5 1 ±.44 Tcarprofen,ng/min 6±1-158±93-618±261 458±336 GFR,sl/min 411±19 495±176 69±13 362± 19 Perfusionflow,ml/min 31 ±4 33±3 33±6 34±1 Perfusion pressure, mm Hg 88 ± 9 95 ± 12 87 ± 7 88 ± 2 Ratweight,g 351±28 37±1 347±19 353±13 Kidney weight, g 1.5 1 ±.7 1.45 ±.3 1.41 ±. 13 1.57 ±.12 N 3 3 3 4 % ofdose excreted in urine.2 ±.3.13 ±.7.16 ±.1 1.49 ±.19 z I-. >- z > -j. 4 2 I 8 6 4 2 2 I 5 5.5-6- 25mg OI5mg 5 mg 4. - 3. E 2 2 H.5mg 5.OmQ H 2 5 mg. H 25 mg. #{149} A Downloaded from dmd.aspetjournals.org at ASPET Journals on March 5, 216 3 4 5 6I 7I 8I 9I 1 I 11 I TIME, FIG. 1. rinary excretion of carprofen in the isolated perfused kidney. Each point represents the mean ± SD from three to four experiments at each dose. plasma that is not filtered may be effectively cleared by active tubular secretion, inasmuch as the rate of secretion is a function of total drug concentration rather than free (unbound) drug in perfusate. Although protein binding has been shown to be a factor in clearance by limiting filtration, there are highly bound cornpounds whose clearances exceed the GFR and approach that of p-aminohippuric acid (9). This phenomenon has been demonstrated recently for the transport of 2,4,5-trichlorophenoxy acetate in the IPK (1). When T for each dose of carprofen administered was calculated (table 3), net excretion equivalent to filtration was indicated at the mm I I ii i i3i i i I i I i i i I I II I ii I I III 4.5 5. 5.5 6 6.5 7. 7.5 FIG. 2. Effect ofurinary ph on clearance of carprofen. Each point represents the average of 2 to 14 determinations at each ph. Clearance is not corrected for F1..5-mg dose, net reabsorption at the 5- and 12.5-mg doses, and net secretion at the 25-mg dose. Such an interpretation of the mechanisms involved in tubular transit needs to be approached with caution, inasmuch as small changes in binding can dramatically alter the net T value. For example, as carprofen concentration increased, Ff increased, providing more available drug for clearance by filtration and thus shifting T. More dramatic effects on carprofen clearance (and T) result from the marked effects of urinary ph and flow rate (fig. 2 and 3). Increased clearance of weak acids due to increased urinary ph ph

28 BEKERSKY AND COLBRN H Q 5 mg 5. mq A S 2.5 mg 6-525mg 5. - C 4. - 1. 2 4 3- S 4 -I 2 IC 11.2 4 8 2 RINE FLOW, ml I mm. FIG. 3. Effect ofurineflow on the clearance of carprofen. Each point represents the average of four to ten determinations. Clearance is not corrected for F. TABLE 4 Renal clearance ofcarprofenfollowing doses ofo.5-25 mg in the isolated, perfused kidney Data are reported as means ± SD of N experiments after the addition of carprofen. Time mm -1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 Mean Mean Cl/mean GFR MeanCl (corrected for Fr) Mean Cl(F1)/mean GFR.5mg (N-3) 5mg (N=3) 12.5mg (N=3) GFR Cl GFR Cl GFR Cl GFR Cl 386 ± 26.92 ±.82 385 ± 233.95 ± 1.1 448 ± 273 1.97 ± 1.35 428 ± 22 1.23 ±.67 514 ± 287 1.56 ±.95 394 ± 164 1.26 ±.66 398 ± 135 1.69 ±.4 379 ± 1 16 2.68 ± 1.93 37 ± 149 3.27 ± 1.97 41 1 ± 19 1.87 ± 1.2.45 69±485 1.68 51 1 ± 214.93 ±.94 521 ± 245 1.15 ± 1.24 58 ± 216 1.11 ±.51 525 ± 197.66 ±.41 536 ± 221.73 ±.51 471 ± 176.64 ±.59 435 ± 192.66 ±.51 439 ± 178.84 ±.91 59 ± 24 1.1 ±.64 495 ± 176.85 ±.64.17 158±118.3 19 d/min 763 ± 81 1.33 ±.81 793 ± 156.79 ±.3 665 ± 156.78 ±.41 793 ± 39.91 ±.95 745 ± 78.8 ±.71 73 ± 17 1.5 ±.7 641 ± 47 1.2 ±.83 55 ± 153 1.8 ±.77 559 ± 1 17 1.97 ± 1.45 69 ± 13 1.8 ±.78.16 174±125.25 1 25mg (N=4) 484 ± 18 3.35 ±.93 451 ± 74 3.85 ± 1.6 416 ± 65 3.41 ± 1.21 387 ± 94 3.64 ± 1.5 46 ± 125 3.88 ±.79 319 ± 1 3.7 ±.86 31 ± 65 2.97 ± 1.34 244 ± 47 2.87 ± 1.18 255 ± 7 3.2 ± 1.65 362 ± 19 3.32 ± 1.5.92 535±177 1.48 Downloaded from dmd.aspetjournals.org at ASPET Journals on March 5, 216 and/or urinary flow rate is well established (1 1) and has been shown to affect salicylate clearance in the IPK (2). At higher urinary ph, reabsorption is decreased and net clearance becomes more dependent on tubular secretion; i.e., increased net clearance (25 mg dose, see table 2). Reabsorption by diffusion in the tubule is also decreased by high urine flow rate and could account for the relatively high clearance found with.5 mg. It is also apparent that, for the intermediate doses, the interaction of urinary ph and urine flow rate becomes extremely complex as the critical ph for increased clearance is approached. In contrast, at the highest dose urinary ph seems to be the predominant determinant of carprofen elimination, as urinary flow rate has little effect (fig. 3) on net clearance. The observed K/P ratios (table 4) indicate that transport of carprofen into kidney tissue is limited, presumably as a result of extensive perfusate binding. However, when carprofen perfusate concentrations were corrected for Ff the K/P ratios became much greater than unity (mean = 72.6), indicating that carprofen is bound to kidney tissue. This observation is consistent with the thesis that tubular secretion is a potential mechanism in the disposition of carprofen by the kidney. In summary, the urinary excretion of carprofen is low both in vivo and in the IPK as the result ofseveral interacting mechanisms. The mechanisms that influence the renal clearance of carprofen in vivo can be defined with the in vitro IPK and are as follows: 1) extensive binding of carprofen in IPK perfusate or plasma limits

RENAL CLEARANCE OF CARPROFEN 29 TABLE 5 Pe,fusate and kidney concentrations of carprofen at the co mpletion of each experiment Data are reported as means ± SD for N experiments. Dose N Perfusate Kidney mg tg/mi ncorrected K/P C ed f orrec1 or.5 3 6.1 ± 1.8 1.81 ±.43.34 ±.19 13.6 ± 76.2 5 3 78.4 ± 3.3 1.89 ± 3.38.14 ±.5 25.9 ± 8.7 12.5 3 21.9 ± 5.3. 9.7 ± 27.4.45 ±.14 72.7 ± 23.5 25 4 383.2 ± 7.9 125.76 ± 19.29.33 ±.5 61.4 ± 18.1 the filtration of carprofen in the glomerulus; 2) tubular secretion of carprofen is effectively counteracted by tubular reabsorption at low urinary ph and limited urinary flow; 3) net tubular secretion becomes apparent when higher urinary ph and increased urine flow rates decrease tubular reabsorption; 4) as a result of these interacting mechanisms, the urinary excretion of carprofen is perfusate concentration-dependent. Acknowledgments. The authors wish to thank Mrs. W. Morley for preparing the manuscript. References 1. R. H. Bowman, Methods EnzymoL 39, 3 (1975). 2. I. Bekersky, L. Fishman, S. A. Kaplan, and W. A. Colburn, J. Pharmacot Exp. Ther. 212, 39 (198). 3. I. Bekersky, W. J. Poynor, and W. A. Colburn, Drug Metab. Dispos. 8, 64 (198). 4. J. E. Ray, D. H. Wade, G. G. Graham, and R.. Day, J. Clin. PharmacoL 19, 635 (1979). 5. F. Rubio, S. Seawall, B. DeBarbieri, W. Benz, L. Berger, L. Morgan, J. Pao, T. H. Williams, and B. Koechlin, J. Pharm. Sci.. 69, 1245 (198). 6. J. R. Weeks and J. D. Davis, J. AppL PhysioL 19, 54 (1964). 7. I. Bekersky, W. A. Colburn, L. Fishman and S. A. Kaplan, Drug Metab. Dispos. 8, 3 19 (198). 8. C. V. Puglisi, J. C. Meyer and J. A. F. de Silva, J. Chrom. Sci. 136, 391 (1977). 9. I. M. Weiner, Handb. PhysioL 8, 521 (1973). 1. F. J. Koschier and M. Acara, J. PharmacoL Exp. Ther. 28, 287 (1979). 1 1. I. M. Weiner, Am. J. Med. 36, 743 (1964). Downloaded from dmd.aspetjournals.org at ASPET Journals on March 5, 216