Effects of probenecid on furosemide kinetics and natriuresis in man Furosemide kinetics were studied in 4 normal subjects after single intravenous injections (I mg/kg). One experiment was done after pretreatment with probenecid. The apparent
volume offurosemide distribution was unchanged after probenecid (10.9 L). The mean plasma clearance fell from 155 to 85 mllmin and the mean plasma tVz rose from 36 to 61 min. Renal clearance of furosemide fell below 50% of control after probenecid, but the kidney remained the main route of its excretion (75% of the dose appeared in the urine). In another experiment in 4 subjects an infusion of furosemide was sustained following a loading dose to maintain a constant plasma level. After a control period, probenecid was given orally. This resulted in a decrease in renal excretion of furosemide with a simultaneous rise in its plasma concentration. Despite the rising plasma furosemide concentration, however, there was a diminution in both urine flow and the excreted fraction of filtered sodium, which suggested some reduction of diuretic action. In doses commonly used, probenecid reduces renal elimination of furosemide in man with only a mild impairment of its diuretic activity. This suggests that furosemide is eliminated predominantly by way of proximal tubular secretion and that tubular rather than plasma concentration is the main determinant of its diuretic effect.
Jamshid Honari, M.D., Andrew D. Blair, Ph.D., and Ralph E. Cutler, M.D. Seattle, Wash. Department of Medicine, School of Medicine, University of Washington at Harborview Medical Center
Furosemide is a potent natriuretic drug which acts on chloride and sodium transport in the loop of Henle. 3 • 8. 14, 17 Its elimination from the body involves both renal and nonrenal routes. 7 Renal excretion is greater than glomerular filtraPortions of this work were supported by a grant (RR·133) from the Division of Research Resources, General Clinical Research Centers Branch, National Institutes of Health; Contract No. NIH· NIAMDD 72-2219, Artificial Kidney-Chronic Uremia Program, National Institutes of Health, and by a grant from Hoechst·Roussel Pharmaceuticals, Inc. Received for publication March 15. 1977. Accepted for publication June 7, 1977. Reprint requests to: Dr. Ralph E. Cutler, Department of Medicine, Harborview Medical Center, 325 Ninth Ave., Seattle, Wash. 98104.
tion rate (GFR) and, in view of its extensive binding to plasma proteins, it is thought that tubular secretion is the major type of transport. 5 • 7 Because furosemide is an organic acid, it is possible that similar compounds may interfere with its tubular transport or its diuretic action, or both. This concept has been tested in dogs with probenecid, and antagonism to furosemide natriuresis was found. 12 However, this concept has not been tested in man. We report here our measurements in man of the effects of pretreatment with probenecid on both furosemide elimination kinetics and diuretic action. 395
396
Honari, Blair, and Cutler
Materials and methods
Studies were performed on 6 normal healthy male subjects at the Clinical Research Center, Harborview Medical Center, University of Washington. Volunteers did not eat or drink for 8 hr prior to the study or during the first 6 hr of study on the following day. They remained in a recumbent position throughout the study and stood only to void at the end of each collection period. An indwelling cannula was first inserted in a forearm vein in the morning, and an intravenous infusion of lactated Ringer's solution was begun and continued throughout the study in all but two experiments. In two experiments (Subjects 5 and 6), an electrolyte solution containing sodium, potassium, and chloride (130, 10 and 140 mEq/L, respectively) was used. The rate of infusion was frequently adjusted to assure a volume for volume replacement of voided urine. The state of volume equilibrium was frequently checked by weighing the patient during the experiment. The following experiments were performed. Single intravenous injection of furosemide. Four subjects were studied twice with a "control" and "probenecid pretreatment" protocol. The order of these experiments was randomized. A "control" study consisted of a single intravenous injection of 35S-furosemide (1 mg/kg body weight containing 15 to 20 jLCi). Blood samples for drug assays (furosemide and probenecid) were drawn as follows: 0, 10,20,30,45 min; 1, 1V2, 2, 2!h, 3, and 4 hr. Urine was collected every 15 to 45 min, depending on the intensity of diuresis. The "probenecid pretreatment" protocol differed from the control study only by the administration of 0.5 gm of probenecid orally, 8 and 2 hr before the injection of furosemide. Sustained infusion offurosemide. Four subjects were studied. Two participants had also been studied with the single injection protocol. The plan of this protocol was to give a loading dose of furosemide and then sustain the plasma concentration attained by a constant intravenous infusion for the next 6 hr. Two different loading and sustaining doses were given. Two subjects received a furosemide prime of 0.4 mg/kg and a sustaining infusion of 0.3 mg/min. This resulted in a plasma concentration of approxi-
Clinical Pharmacology and Therapeutics
mately 2 mg/L. A lower sustained plasma concentration between 0.5 and 1.0 mg/L was obtained in 2 additional subjects with a lower loading dose of 0.2 mg/kg and a sustaining infusion of 0.13 mg/min. GFR was measured by inulin clearance as described below and the fraction of filtered sodium which was excreted thus calculated. A loading dose of inulin according to subject's body weight was given intravenously to obtain a plasma concentration of 200 to 250 mg/L. This was immediately followed by a sustaining infusion of inulin, at a rate based upon the estimated GFR, to maintain a relatively constant blood level. About 45 min after the sustaining infusion was begun, three 30-min urine collections were made with midpoint drawing of blood samples. Renal clearances of inulin, furosemide, and uric acid were calculated. Following these control measurements, 2 gm of probenecid was given orally in 0.5-gm doses over the next 2 hr and clearance collections were continued for at least 2 hr beyond the last dose of probenecid. Assays. Urine and serum were analyzed for sodium and potassium by flame photometry. Chloride, creatinine, uric acid, and inulin were measured by standard AutoAnalyzer techniques. Osmolality was obtained by freezing point depression. Furosemide concentrations were calculated from liquid scintillation data after background and quenching corrections were made. 9 During the course of this study, we also developed a high-pressure liquid chromatographic (HPLC) assay for furosemide.! Most of the samples in this study were then reassayed by HPLC. The agreement between the two assays was excellent, which suggests that 98% of the measured radioactivity was in the form of furosemide.! Probenecid was also assayed by HPLC with the use of the conditions previously described for flucytosine.! Kinetic computations and statistical techniques. Kinetic parameters for furosemide were calculated from serum and urine data by computer programs by means of a two-compartment open model. The apparent volume of distribution was calculated as the sum of the central and peripheral compartment. The tlh was calculated from the elimination rate constant of the central compartment. The plasma clearances of furose-
Volume 22 Number 4
397
Effects of probenecid on furosemide
Table I. Kinetics of intravenous furosemide (F) alone or with oral probenecid (F+P) in normal subjects Distribution volume Subject
Dose* (mg)
1 2 3 4 Mean SD
66 70
72
71
t1,6 (min)
(L)
F /F+P
9.9 13.6 13.9 11.9 12.3 1.8
11.4 11.3 9.0 11.7 10.9 1.2
F
36.8 41.9 30.9 33.7 35.8 4.7
IF+P
59.8 67.4 52.8 63.2 60.8t 6.2
Plasma clearance (mllmin)
Renal clearance (mllmin)
Nonrenal clearance (mllmin)
F IF+P
F
F
IF+P
140 146 190 143 155 24
116 125 168 127 134 23
24 21 22 16 21 3
13 21 20 15 17 4
74 83 92 92 85t 9
IF+P 61 62
72
57 63t 6
Creatinine clearance (mllmin)
Dose excreted in 24 hr urine (%)
F /F+P
FIF+P
109 105 122 130 114 132 151 148 124 129 19 18
83 86 88 89 87 3
83 75 78 62 75 9
• Dose = I mg/kg body weight. t Difference between F and F+ P is significant (p < 0.05).
mide were calculated by dividing the dose administered by the area under the serum concentration curve from zero to infinite time. Renal clearances were calculated by dividing the total amount of furosemide excreted in the urine in 24 hr by the area under the serum concentration curve from zero to infinite time. The area under the plasma concentration curve was estimated from the exponential equation generated by a computer routine which gave the best least-squares fit to the serum concentration. Statistical comparisons between studies were made with the use of a paired t test. The null hypothesis was rejected if the significance level was equal to or less than 0.05.
Table II. Initial 6-hour urine volume and sodium excretion following intravenous furosemide (F) alone or with oral probenecid (F +P) in normal subjects Volume (m1l6 hr) Subject
1 2 3 4 Mean SD
4,925 4,906 5,281 5,281 5,098 211
Sodium (mEq/6 hr)
F+P
F
F+P
6,708 4,741 6,740 6,467 6,164* 956
538 510 627 638 578 64
768 526 829 656 694t 133
·No significant difference between F and F+P; p value, 0.09. tNo significant difference between F and F+P; p value, 0.14.
Results
Single injection experiment. (Tables I and II). The apparent volume of furosemide distribution was 12.3 ± 1.8 L (mean ± SD); no
significant change occurred after probenecid administration and no significant change in the mean clearance of creatinine was noted following probenecid. The plasma and renal clearance and the elimination t'h of furosemide were, however, significantly reduced following probenecid. The mean plasma clearance fell from 155 ± 24 to 85 ± 9 ml/min. This reduction of plasma elimination lengthened the plasma tV2 from 35.8 ± 4.7 to 60.8 ± 6.2 min. Except for Subject 1, the change in plasma clearance after probenecid was due entirely to changes in renal clearance. In Subject 1, the nonrenal clearance fell by 11 ml/min after probenecid.
The mean renal excretion of furosemide was 87% ± 3% of the administered dose and fell to a mean of 75% ± 9% after probenecid (which is not a significant difference). The initial 6hr excretion of urine after furosemide was 5,098 ± 211 ml and the sodium was 578 ± 64 mEq (Table II). The group mean for both of these parameters increased following probenecid, but the difference was not statistically significant. Contant furosemide infusion study (Table III, Fig. 1). For data analysis, the study was arbitrarily divided into three periods. Period 1 was the first 90 min of observation before probenecid administration. This portion of the study served as a control for the data obtained after probenecid. Period 2 was the 90-min span
398
Honari, Blair, and Cutler
Clinical Pharmacology and Therapeutics
...J
......
0>
E
./---/-"",-
4
........
I G)
\I
-0
""~" 0:::::--
E
t)
II)
...
o
,,_
O /0_0-0-0 O O D_O_ -
O'..:::o-~-O/
2
::J
--*---*~--
*,
U.
_/
/
/- _/
:*-*--*~ e-e 11-----------------___ e_e_e-e_e-e_e_e_e-e-e_e
0 180
...J 160
...... 0>140
/e-e
E 120
~
/e
100
u 80
-
t)
;60
~~ a..
"e
-/.-------
~ -
/-
//
40 20
0 7/ -
o 60
120
180
/
0
*__ /0, -o-O~ 0-0 240
300
360
Time- min Fig. 1. Top, Plasma furosemide concentrations in 4 subjects following an intravenous loading dose and sustaining infusion. Bottom, Plasma probenecid concentrations in the same subjects after oral dose.
immediately following the first dose of probenecid when plasma concentrations of probenecid were just beginning to rise (Fig. I) . Period 3 was the 90-min span toward the end of the study (between 210 to 300 min; Fig. 1) when plasma probenecid concentration had reached a plateau in most subjects. Period I-control. During this time the plasma furosemide concentration fell slightly as the estimated body clearance exceeded the infusion rate (Fig. 1). The larger furosemide dose in Subjects 1 and 2 resulted in plasma concentrations of about 2 mg/L, the lower dose a mean furosemide plasma concentration slightly under 1 mg/L. The mean values during this period for urine flow (V), the excreted fraction of filtered sodium (EFNa), renal clearances of inulin (CI 1n), furosemide (CI F), and uric acid (Cl uA ), along with plasma probenecid concentration (C p ), are noted for each subject in Table III (along with group statistical analyses). Period 2. During this period the plasma concentration of probenecid rose (group mean of 15 ± 11 mg/L). Small, but significant, changes were noted in the renal clearances of
both furosemide and uric acid. The decrease in furosemide excretion resulted in a rise in plasma concentration in all subjects (Fig. 1). Period 3. The plasma concentration of probenecid had risen further by this time; the group mean was 76 ± 48 mg/L. The plasma concentration of furosemide continued to rise in association with a further depression of furosemide renal clearance (Table III, Fig. 1). Renal excretion of uric acid rose further. By this time, and despite a further rise in plasma furosemide concentration, there was a significant fall in both urine flow and EFNa for the group, which was mainly due to the marked changes in Subjects 2 and 6 who absorbed probenecid better and achieved plasma concentrations about 4 times those of the other subjects. Discussion
The renal tubular site of action of furosemide is still debated, but most of the evidence reported from renal clearance and micropuncture studies points to the thick ascending limb of Henle's loop as the primary area of furosemide action with somewhat conflicting experimental
Volume 22 Number 4
399
Effects of probenecid on furosemide
Table III. Effect of oral probenecid on renal function during a constant intravenous infusion of furosemide V
Subject
Cl 1n
Period I-control 1 32.2 2 27.0 5* 33.7 6* 24.0 Mean 29.2 4.5 SO Period 2-post-probenecid 1 35.6 2 21.7 5 16.5 21.1 6 Mean 23.7 SO 8.3 Period 3-postprobenecid 20.5 1 2 12.6 5 14.3 11.1 6 Mean 14.6t SO 4.1
CIF
ClvA
(mllmin)
(mllmin)
(mllmin)
26.6 23.0 14.3 10.7 18.7 7.4
91 97 183 165 134 47
119 119 117 141 124 11
18 6 18 17 15 6
28.6 23.2 10.7 7.7 17.6 10.0
89 74 126 158 112 38
106 106 100 124 109t 10
34 21 17 24 24t 7
26.3 14.3 12.3 4.7
86 65 113 144 102 34
61 31 46 39
55 26 29 55 41t 16
(mllmin)
14.4t 8.9
44t 13
Cp (mg/L)
o o o o
8
30
8
12 15 11
32 111 38 123 76 48
See text for explanation of symbols and periods. • Subjects receiving lactate-free electrolyte solution. t Difference from period I is significant (p < 0.05).
results regarding its proximal tubular effect. 2-4. Microperfusion studies by Burg, and co-workers 3 have demonstrated that low concentrations (I 0- 6 M) of the drug in the tubular lumen is much more inhibitory of sodium chloride transport than higher concentrations (10- 4 M) bathing the peritubular surface. These results are particularly interesting because they were the first to suggest that furosemide action in the thick ascending limb is directed against an active electrogenic chloride pump. It is presumed that furosemide gains access to the loop of Henle through secretion in more proximal portions of the nephron. Furosemide is a weak organic acid (pKa 3.9) which is primarily excreted by tubular secretion. It would be expected to compete with other organic acids for secretion by the proximal convoluted tubule. This appears to be the primary route of renal excretion since the substantial (>90%) binding to plasma protein9 greatly limits glomerular filtration. Kinetic studies have 6. 8. 14. 17
shown that 66% to 99% of an intravenous dose is excreted in the urine at renal clearance rates which exceed the GFR.5. 7 It might be expected that other organic acids with similar tubular transport systems would inhibit furosemide secretion, reduce the amount in the tubular lumen, and thus reduce its diuretic action. Probenecid is known to be a potent competetive inhibitor of secreted weak organic acids. 18 Senft 16 blocked the diuretic action of furosemide in rats with probenecid, and Hook and Williamson 12 reported similar results in dogs. Portwich and associates 13 published similar results in man with diodrast, but Gayerl0 could not demonstrate any inhibition of diuresis or furosemide transport with para-aminohippurate in dogs. Our study demonstrates that probenecid is a potent inhibitor of furosemide elimination in man. The reduction of furosemide plasma clearance is primarily due to a decrease in renal excretion, presumably secondary to a reduction
400
Honari, Blair, and Cutler
of proximal tubular secretion. An alternative explanation would be an increased backdiffusion of the secreted and filtered furosemide. Once in the proximal tubule with near neutral pH, furosemide is almost completely dissociated, and, hence, an insignificant amount of back-diffusion is expected to take place in this part of the nephron, as shown by Deetjen 8 in micropuncture studies. In the distal tubule with a urine pH near 4.5, it is conceivable that there might be some inorganic backdiffusion. It is unlikely, however, that probenecid treatment would alter urine pH sufficiently to cause a significant change in nonionic back-diffusion of furosemide and thereby reduce its renal clearance. The decrease in furosemide renal clearance (about 3-fold) was similar to the increase in uric acid clearance. Infusion of lactated Ringer's solution as a replacement fluid therapy during most of our experiments may raise the question of the effect this amount of lactate might have had on the renal handling of uric acid as well as of furosemide. Many years ago, Gibson and Doisyll observed that sodium lactate ingestion caused a striking, although transient, decrease in urinary uric acid excretion. Yii and colleagues 19 subsequently expanded this observation. With an intravenous infusion of 0.17 to 0.4 M sodium lactate (much larger amounts than used in this study), they demonstrated a marked reduction in renal uric acid clearance and also an abolition of the uricosuric effect of probenecid. Such effects were not observed by us. Furthermore, our observation in 2 subjects (5 and 6) who received a lactate-free solution during the experiments, suggests that the rate of lactate infusion used in this study did not affect furosemide and uric acid clearances. Although a statistical reduction in natriuresis (volume and sodium excretion) occurred after probenecid during the furosemide infusion studies, no effect was noted when the single injection of furosemide was given. The single, intermittent dosage, orally or intravenously, is a more common practice; thus, it appears unlikely that drug interaction with probenecid is clinically important in affecting diuretic action. It is possible, however, that long-term dosing with probenecid may affect the diuretic re-
Clinical Pharmacology and Therapeutics
sponse, and the current results with the use of acute loading may not necessarily apply to that situation. Our study is consistent with the data of Burg, and co-workers 3 and Rose, and associates 15 in suggesting that the amount of furosemide in the tubular fluid rather than plasma concentration is the main determinant of furosemide diuresis. Our failure to demonstrate an effect on sodium excretion comparable to that noted in animal studies 12 could be due to a species difference or to a lower probenecid/furosemide molar concentration ratio at the renal tubular receptor site. The latter explanation is suggested by our observation that subjects with a high plasma probenecid concentration had the greater reductions in natriuresis.
References I. Blair, A. D., Forrey, A. W., Meijsen, B. T., and Cutler, R. E.: Assay of ftucytosine and furosemide by high-pressure liquid chromatography, J. Pharm. Sci. 64:1334-1339, 1975. 2. Brenner, B. M., Keimonitz, R. I., Wright, F. S., and Berliner, R. W.: An inhibitory effect of furosemide on sodium reabsorption by the proximal tubule of the rat nephron, J. Clin. Invest. 48:290-300, 1969. 3. Burg, M., Stoner, L., Cardinal, J., and Green, N.: Furosemide effect on isolated perfused tubules, Am. J. Physiol. 225:119-124,1973. 4. Burke, T. J., Robinson, R. R., and Clapp, J. R.: Determination of the effects of furosemide on the proximal tubule, Kidney Int. 1:12-18, 1972. 5. Calesnick, B., Christensen, J. A., and Richter, M.: Absorption and excretion of furosemide-S 35 in human subjects, Proc. Soc. Exp. BioI. Med. 123:17-22, 1966. 6. Clapp, J. R., and Robinson, R. R.: Distal sites of action of diuretic drugs in the dog nephron, Am. J. Physiol. 215:228-235, 1968. 7. Cutler, R. E., Forrey, A. W., Christopher, T. G., and Kimpel, B. M.: Pharmacokinetics of furosemide in normal subjects and functionally anephric patients, CUN. PHARMACOL. THER. 15:588-596, 1974. 8. Deetjen, P.: Micropuncture studies on site and mode of diuretic action of furosemide, Ann. N. Y. Acad. Sci. 139:408-415, 1966. 9. Forrey, A. W., Kimpel, B., Blair, A. D., and Cutler, R. E.: Furosemide concentrations in serum and urine, and its binding by serum proteins as measured ftuorometrically, Clin. Chern. 20:152-158, 1974. 10. Gayer, J.: Die renaIe Exkretion des neuen
Effects of probenecid on furosemide
Volume 22 Number 4
II.
12.
13.
14.
15.
diurecticum Furosemid, Klin. Wochenschr. 43:898-902, 1965. Gibson, H. V., and Doisy, E. A.: A note on the effect of some organic acids upon the uric acid excretion of man, 1. BioI. Chern. 55:605-610, 1923. Hook, J. B., and Williamson, H. E.: Influences of probenecid and alterations in acid-base balance on the saluretic activity of furosemide, J. Pharmacol. Exp. Ther. 149:404-408, 1972. Portwich, V. G., Buttner, H., and Schafer, 1.: Simultane Untersuchung der Clearance und renalen Extraktion von Pharmaka, in Ullrich, K. S., and Hierholzer, K., editors: Normale und pathologische Funktionen des Neirentubulus, Bern, 1965, Hans Huber, pp. 173-176. Puschett, J. B., and Goldberg, M.: The acute effects of furosemide on acid and electrolyte excretion in man, 1. Lab. Clin. Med. 71:666667, 1968. Rose, H. 1., Pruitt, A. W., Dayton, P. G., and
16.
17.
18.
19.
401
McNay, 1. L.: Relationship of urinary furosemide excretion rate to natriuretic effect in experimental azotemia, 1. Pharmacol. Exp. Ther. 199:490-497, 1976. Senft, G.: Membrantransport and Pharmaka, in Ullrich, K. S., and Hierholzer, K., editors: Norrnale und pathologische Funktionen des Nierentubulus, Bern, 1965, Hans Huber, pp. 63-81. Suki, W., Rector, F. C., Jr., and Seldin, D. W.: The site of action of furosemide and other sulfonamide diuretics in the dog, J. Clin. Invest. 44:1458-1469, 1965. Weiner,!. M., Washington, J. A., and Mudge, G. H.: On the mechanism of action of probenecid on renal tubular secretion. Bull. Johns Hopkins Hosp. 106:333-346, 1960. Yii, F. T., Sirota, J. H., Berger, L., Halpern, M., and Gutman, A. B.: Effect of sodium lactate infusion on urate clearance in man, Proc. Soc. Exp. BioI. Med. 96:809-813, 1957.
volume offurosemide distribution was unchanged after probenecid (10.9 L). The mean plasma clearance fell from 155 to 85 mllmin and the mean plasma tVz rose from 36 to 61 min. Renal clearance of furosemide fell below 50% of control after probenecid, but the kidney remained the main route of its excretion (75% of the dose appeared in the urine). In another experiment in 4 subjects an infusion of furosemide was sustained following a loading dose to maintain a constant plasma level. After a control period, probenecid was given orally. This resulted in a decrease in renal excretion of furosemide with a simultaneous rise in its plasma concentration. Despite the rising plasma furosemide concentration, however, there was a diminution in both urine flow and the excreted fraction of filtered sodium, which suggested some reduction of diuretic action. In doses commonly used, probenecid reduces renal elimination of furosemide in man with only a mild impairment of its diuretic activity. This suggests that furosemide is eliminated predominantly by way of proximal tubular secretion and that tubular rather than plasma concentration is the main determinant of its diuretic effect.
Jamshid Honari, M.D., Andrew D. Blair, Ph.D., and Ralph E. Cutler, M.D. Seattle, Wash. Department of Medicine, School of Medicine, University of Washington at Harborview Medical Center
Furosemide is a potent natriuretic drug which acts on chloride and sodium transport in the loop of Henle. 3 • 8. 14, 17 Its elimination from the body involves both renal and nonrenal routes. 7 Renal excretion is greater than glomerular filtraPortions of this work were supported by a grant (RR·133) from the Division of Research Resources, General Clinical Research Centers Branch, National Institutes of Health; Contract No. NIH· NIAMDD 72-2219, Artificial Kidney-Chronic Uremia Program, National Institutes of Health, and by a grant from Hoechst·Roussel Pharmaceuticals, Inc. Received for publication March 15. 1977. Accepted for publication June 7, 1977. Reprint requests to: Dr. Ralph E. Cutler, Department of Medicine, Harborview Medical Center, 325 Ninth Ave., Seattle, Wash. 98104.
tion rate (GFR) and, in view of its extensive binding to plasma proteins, it is thought that tubular secretion is the major type of transport. 5 • 7 Because furosemide is an organic acid, it is possible that similar compounds may interfere with its tubular transport or its diuretic action, or both. This concept has been tested in dogs with probenecid, and antagonism to furosemide natriuresis was found. 12 However, this concept has not been tested in man. We report here our measurements in man of the effects of pretreatment with probenecid on both furosemide elimination kinetics and diuretic action. 395
396
Honari, Blair, and Cutler
Materials and methods
Studies were performed on 6 normal healthy male subjects at the Clinical Research Center, Harborview Medical Center, University of Washington. Volunteers did not eat or drink for 8 hr prior to the study or during the first 6 hr of study on the following day. They remained in a recumbent position throughout the study and stood only to void at the end of each collection period. An indwelling cannula was first inserted in a forearm vein in the morning, and an intravenous infusion of lactated Ringer's solution was begun and continued throughout the study in all but two experiments. In two experiments (Subjects 5 and 6), an electrolyte solution containing sodium, potassium, and chloride (130, 10 and 140 mEq/L, respectively) was used. The rate of infusion was frequently adjusted to assure a volume for volume replacement of voided urine. The state of volume equilibrium was frequently checked by weighing the patient during the experiment. The following experiments were performed. Single intravenous injection of furosemide. Four subjects were studied twice with a "control" and "probenecid pretreatment" protocol. The order of these experiments was randomized. A "control" study consisted of a single intravenous injection of 35S-furosemide (1 mg/kg body weight containing 15 to 20 jLCi). Blood samples for drug assays (furosemide and probenecid) were drawn as follows: 0, 10,20,30,45 min; 1, 1V2, 2, 2!h, 3, and 4 hr. Urine was collected every 15 to 45 min, depending on the intensity of diuresis. The "probenecid pretreatment" protocol differed from the control study only by the administration of 0.5 gm of probenecid orally, 8 and 2 hr before the injection of furosemide. Sustained infusion offurosemide. Four subjects were studied. Two participants had also been studied with the single injection protocol. The plan of this protocol was to give a loading dose of furosemide and then sustain the plasma concentration attained by a constant intravenous infusion for the next 6 hr. Two different loading and sustaining doses were given. Two subjects received a furosemide prime of 0.4 mg/kg and a sustaining infusion of 0.3 mg/min. This resulted in a plasma concentration of approxi-
Clinical Pharmacology and Therapeutics
mately 2 mg/L. A lower sustained plasma concentration between 0.5 and 1.0 mg/L was obtained in 2 additional subjects with a lower loading dose of 0.2 mg/kg and a sustaining infusion of 0.13 mg/min. GFR was measured by inulin clearance as described below and the fraction of filtered sodium which was excreted thus calculated. A loading dose of inulin according to subject's body weight was given intravenously to obtain a plasma concentration of 200 to 250 mg/L. This was immediately followed by a sustaining infusion of inulin, at a rate based upon the estimated GFR, to maintain a relatively constant blood level. About 45 min after the sustaining infusion was begun, three 30-min urine collections were made with midpoint drawing of blood samples. Renal clearances of inulin, furosemide, and uric acid were calculated. Following these control measurements, 2 gm of probenecid was given orally in 0.5-gm doses over the next 2 hr and clearance collections were continued for at least 2 hr beyond the last dose of probenecid. Assays. Urine and serum were analyzed for sodium and potassium by flame photometry. Chloride, creatinine, uric acid, and inulin were measured by standard AutoAnalyzer techniques. Osmolality was obtained by freezing point depression. Furosemide concentrations were calculated from liquid scintillation data after background and quenching corrections were made. 9 During the course of this study, we also developed a high-pressure liquid chromatographic (HPLC) assay for furosemide.! Most of the samples in this study were then reassayed by HPLC. The agreement between the two assays was excellent, which suggests that 98% of the measured radioactivity was in the form of furosemide.! Probenecid was also assayed by HPLC with the use of the conditions previously described for flucytosine.! Kinetic computations and statistical techniques. Kinetic parameters for furosemide were calculated from serum and urine data by computer programs by means of a two-compartment open model. The apparent volume of distribution was calculated as the sum of the central and peripheral compartment. The tlh was calculated from the elimination rate constant of the central compartment. The plasma clearances of furose-
Volume 22 Number 4
397
Effects of probenecid on furosemide
Table I. Kinetics of intravenous furosemide (F) alone or with oral probenecid (F+P) in normal subjects Distribution volume Subject
Dose* (mg)
1 2 3 4 Mean SD
66 70
72
71
t1,6 (min)
(L)
F /F+P
9.9 13.6 13.9 11.9 12.3 1.8
11.4 11.3 9.0 11.7 10.9 1.2
F
36.8 41.9 30.9 33.7 35.8 4.7
IF+P
59.8 67.4 52.8 63.2 60.8t 6.2
Plasma clearance (mllmin)
Renal clearance (mllmin)
Nonrenal clearance (mllmin)
F IF+P
F
F
IF+P
140 146 190 143 155 24
116 125 168 127 134 23
24 21 22 16 21 3
13 21 20 15 17 4
74 83 92 92 85t 9
IF+P 61 62
72
57 63t 6
Creatinine clearance (mllmin)
Dose excreted in 24 hr urine (%)
F /F+P
FIF+P
109 105 122 130 114 132 151 148 124 129 19 18
83 86 88 89 87 3
83 75 78 62 75 9
• Dose = I mg/kg body weight. t Difference between F and F+ P is significant (p < 0.05).
mide were calculated by dividing the dose administered by the area under the serum concentration curve from zero to infinite time. Renal clearances were calculated by dividing the total amount of furosemide excreted in the urine in 24 hr by the area under the serum concentration curve from zero to infinite time. The area under the plasma concentration curve was estimated from the exponential equation generated by a computer routine which gave the best least-squares fit to the serum concentration. Statistical comparisons between studies were made with the use of a paired t test. The null hypothesis was rejected if the significance level was equal to or less than 0.05.
Table II. Initial 6-hour urine volume and sodium excretion following intravenous furosemide (F) alone or with oral probenecid (F +P) in normal subjects Volume (m1l6 hr) Subject
1 2 3 4 Mean SD
4,925 4,906 5,281 5,281 5,098 211
Sodium (mEq/6 hr)
F+P
F
F+P
6,708 4,741 6,740 6,467 6,164* 956
538 510 627 638 578 64
768 526 829 656 694t 133
·No significant difference between F and F+P; p value, 0.09. tNo significant difference between F and F+P; p value, 0.14.
Results
Single injection experiment. (Tables I and II). The apparent volume of furosemide distribution was 12.3 ± 1.8 L (mean ± SD); no
significant change occurred after probenecid administration and no significant change in the mean clearance of creatinine was noted following probenecid. The plasma and renal clearance and the elimination t'h of furosemide were, however, significantly reduced following probenecid. The mean plasma clearance fell from 155 ± 24 to 85 ± 9 ml/min. This reduction of plasma elimination lengthened the plasma tV2 from 35.8 ± 4.7 to 60.8 ± 6.2 min. Except for Subject 1, the change in plasma clearance after probenecid was due entirely to changes in renal clearance. In Subject 1, the nonrenal clearance fell by 11 ml/min after probenecid.
The mean renal excretion of furosemide was 87% ± 3% of the administered dose and fell to a mean of 75% ± 9% after probenecid (which is not a significant difference). The initial 6hr excretion of urine after furosemide was 5,098 ± 211 ml and the sodium was 578 ± 64 mEq (Table II). The group mean for both of these parameters increased following probenecid, but the difference was not statistically significant. Contant furosemide infusion study (Table III, Fig. 1). For data analysis, the study was arbitrarily divided into three periods. Period 1 was the first 90 min of observation before probenecid administration. This portion of the study served as a control for the data obtained after probenecid. Period 2 was the 90-min span
398
Honari, Blair, and Cutler
Clinical Pharmacology and Therapeutics
...J
......
0>
E
./---/-"",-
4
........
I G)
\I
-0
""~" 0:::::--
E
t)
II)
...
o
,,_
O /0_0-0-0 O O D_O_ -
O'..:::o-~-O/
2
::J
--*---*~--
*,
U.
_/
/
/- _/
:*-*--*~ e-e 11-----------------___ e_e_e-e_e-e_e_e_e-e-e_e
0 180
...J 160
...... 0>140
/e-e
E 120
~
/e
100
u 80
-
t)
;60
~~ a..
"e
-/.-------
~ -
/-
//
40 20
0 7/ -
o 60
120
180
/
0
*__ /0, -o-O~ 0-0 240
300
360
Time- min Fig. 1. Top, Plasma furosemide concentrations in 4 subjects following an intravenous loading dose and sustaining infusion. Bottom, Plasma probenecid concentrations in the same subjects after oral dose.
immediately following the first dose of probenecid when plasma concentrations of probenecid were just beginning to rise (Fig. I) . Period 3 was the 90-min span toward the end of the study (between 210 to 300 min; Fig. 1) when plasma probenecid concentration had reached a plateau in most subjects. Period I-control. During this time the plasma furosemide concentration fell slightly as the estimated body clearance exceeded the infusion rate (Fig. 1). The larger furosemide dose in Subjects 1 and 2 resulted in plasma concentrations of about 2 mg/L, the lower dose a mean furosemide plasma concentration slightly under 1 mg/L. The mean values during this period for urine flow (V), the excreted fraction of filtered sodium (EFNa), renal clearances of inulin (CI 1n), furosemide (CI F), and uric acid (Cl uA ), along with plasma probenecid concentration (C p ), are noted for each subject in Table III (along with group statistical analyses). Period 2. During this period the plasma concentration of probenecid rose (group mean of 15 ± 11 mg/L). Small, but significant, changes were noted in the renal clearances of
both furosemide and uric acid. The decrease in furosemide excretion resulted in a rise in plasma concentration in all subjects (Fig. 1). Period 3. The plasma concentration of probenecid had risen further by this time; the group mean was 76 ± 48 mg/L. The plasma concentration of furosemide continued to rise in association with a further depression of furosemide renal clearance (Table III, Fig. 1). Renal excretion of uric acid rose further. By this time, and despite a further rise in plasma furosemide concentration, there was a significant fall in both urine flow and EFNa for the group, which was mainly due to the marked changes in Subjects 2 and 6 who absorbed probenecid better and achieved plasma concentrations about 4 times those of the other subjects. Discussion
The renal tubular site of action of furosemide is still debated, but most of the evidence reported from renal clearance and micropuncture studies points to the thick ascending limb of Henle's loop as the primary area of furosemide action with somewhat conflicting experimental
Volume 22 Number 4
399
Effects of probenecid on furosemide
Table III. Effect of oral probenecid on renal function during a constant intravenous infusion of furosemide V
Subject
Cl 1n
Period I-control 1 32.2 2 27.0 5* 33.7 6* 24.0 Mean 29.2 4.5 SO Period 2-post-probenecid 1 35.6 2 21.7 5 16.5 21.1 6 Mean 23.7 SO 8.3 Period 3-postprobenecid 20.5 1 2 12.6 5 14.3 11.1 6 Mean 14.6t SO 4.1
CIF
ClvA
(mllmin)
(mllmin)
(mllmin)
26.6 23.0 14.3 10.7 18.7 7.4
91 97 183 165 134 47
119 119 117 141 124 11
18 6 18 17 15 6
28.6 23.2 10.7 7.7 17.6 10.0
89 74 126 158 112 38
106 106 100 124 109t 10
34 21 17 24 24t 7
26.3 14.3 12.3 4.7
86 65 113 144 102 34
61 31 46 39
55 26 29 55 41t 16
(mllmin)
14.4t 8.9
44t 13
Cp (mg/L)
o o o o
8
30
8
12 15 11
32 111 38 123 76 48
See text for explanation of symbols and periods. • Subjects receiving lactate-free electrolyte solution. t Difference from period I is significant (p < 0.05).
results regarding its proximal tubular effect. 2-4. Microperfusion studies by Burg, and co-workers 3 have demonstrated that low concentrations (I 0- 6 M) of the drug in the tubular lumen is much more inhibitory of sodium chloride transport than higher concentrations (10- 4 M) bathing the peritubular surface. These results are particularly interesting because they were the first to suggest that furosemide action in the thick ascending limb is directed against an active electrogenic chloride pump. It is presumed that furosemide gains access to the loop of Henle through secretion in more proximal portions of the nephron. Furosemide is a weak organic acid (pKa 3.9) which is primarily excreted by tubular secretion. It would be expected to compete with other organic acids for secretion by the proximal convoluted tubule. This appears to be the primary route of renal excretion since the substantial (>90%) binding to plasma protein9 greatly limits glomerular filtration. Kinetic studies have 6. 8. 14. 17
shown that 66% to 99% of an intravenous dose is excreted in the urine at renal clearance rates which exceed the GFR.5. 7 It might be expected that other organic acids with similar tubular transport systems would inhibit furosemide secretion, reduce the amount in the tubular lumen, and thus reduce its diuretic action. Probenecid is known to be a potent competetive inhibitor of secreted weak organic acids. 18 Senft 16 blocked the diuretic action of furosemide in rats with probenecid, and Hook and Williamson 12 reported similar results in dogs. Portwich and associates 13 published similar results in man with diodrast, but Gayerl0 could not demonstrate any inhibition of diuresis or furosemide transport with para-aminohippurate in dogs. Our study demonstrates that probenecid is a potent inhibitor of furosemide elimination in man. The reduction of furosemide plasma clearance is primarily due to a decrease in renal excretion, presumably secondary to a reduction
400
Honari, Blair, and Cutler
of proximal tubular secretion. An alternative explanation would be an increased backdiffusion of the secreted and filtered furosemide. Once in the proximal tubule with near neutral pH, furosemide is almost completely dissociated, and, hence, an insignificant amount of back-diffusion is expected to take place in this part of the nephron, as shown by Deetjen 8 in micropuncture studies. In the distal tubule with a urine pH near 4.5, it is conceivable that there might be some inorganic backdiffusion. It is unlikely, however, that probenecid treatment would alter urine pH sufficiently to cause a significant change in nonionic back-diffusion of furosemide and thereby reduce its renal clearance. The decrease in furosemide renal clearance (about 3-fold) was similar to the increase in uric acid clearance. Infusion of lactated Ringer's solution as a replacement fluid therapy during most of our experiments may raise the question of the effect this amount of lactate might have had on the renal handling of uric acid as well as of furosemide. Many years ago, Gibson and Doisyll observed that sodium lactate ingestion caused a striking, although transient, decrease in urinary uric acid excretion. Yii and colleagues 19 subsequently expanded this observation. With an intravenous infusion of 0.17 to 0.4 M sodium lactate (much larger amounts than used in this study), they demonstrated a marked reduction in renal uric acid clearance and also an abolition of the uricosuric effect of probenecid. Such effects were not observed by us. Furthermore, our observation in 2 subjects (5 and 6) who received a lactate-free solution during the experiments, suggests that the rate of lactate infusion used in this study did not affect furosemide and uric acid clearances. Although a statistical reduction in natriuresis (volume and sodium excretion) occurred after probenecid during the furosemide infusion studies, no effect was noted when the single injection of furosemide was given. The single, intermittent dosage, orally or intravenously, is a more common practice; thus, it appears unlikely that drug interaction with probenecid is clinically important in affecting diuretic action. It is possible, however, that long-term dosing with probenecid may affect the diuretic re-
Clinical Pharmacology and Therapeutics
sponse, and the current results with the use of acute loading may not necessarily apply to that situation. Our study is consistent with the data of Burg, and co-workers 3 and Rose, and associates 15 in suggesting that the amount of furosemide in the tubular fluid rather than plasma concentration is the main determinant of furosemide diuresis. Our failure to demonstrate an effect on sodium excretion comparable to that noted in animal studies 12 could be due to a species difference or to a lower probenecid/furosemide molar concentration ratio at the renal tubular receptor site. The latter explanation is suggested by our observation that subjects with a high plasma probenecid concentration had the greater reductions in natriuresis.
References I. Blair, A. D., Forrey, A. W., Meijsen, B. T., and Cutler, R. E.: Assay of ftucytosine and furosemide by high-pressure liquid chromatography, J. Pharm. Sci. 64:1334-1339, 1975. 2. Brenner, B. M., Keimonitz, R. I., Wright, F. S., and Berliner, R. W.: An inhibitory effect of furosemide on sodium reabsorption by the proximal tubule of the rat nephron, J. Clin. Invest. 48:290-300, 1969. 3. Burg, M., Stoner, L., Cardinal, J., and Green, N.: Furosemide effect on isolated perfused tubules, Am. J. Physiol. 225:119-124,1973. 4. Burke, T. J., Robinson, R. R., and Clapp, J. R.: Determination of the effects of furosemide on the proximal tubule, Kidney Int. 1:12-18, 1972. 5. Calesnick, B., Christensen, J. A., and Richter, M.: Absorption and excretion of furosemide-S 35 in human subjects, Proc. Soc. Exp. BioI. Med. 123:17-22, 1966. 6. Clapp, J. R., and Robinson, R. R.: Distal sites of action of diuretic drugs in the dog nephron, Am. J. Physiol. 215:228-235, 1968. 7. Cutler, R. E., Forrey, A. W., Christopher, T. G., and Kimpel, B. M.: Pharmacokinetics of furosemide in normal subjects and functionally anephric patients, CUN. PHARMACOL. THER. 15:588-596, 1974. 8. Deetjen, P.: Micropuncture studies on site and mode of diuretic action of furosemide, Ann. N. Y. Acad. Sci. 139:408-415, 1966. 9. Forrey, A. W., Kimpel, B., Blair, A. D., and Cutler, R. E.: Furosemide concentrations in serum and urine, and its binding by serum proteins as measured ftuorometrically, Clin. Chern. 20:152-158, 1974. 10. Gayer, J.: Die renaIe Exkretion des neuen
Effects of probenecid on furosemide
Volume 22 Number 4
II.
12.
13.
14.
15.
diurecticum Furosemid, Klin. Wochenschr. 43:898-902, 1965. Gibson, H. V., and Doisy, E. A.: A note on the effect of some organic acids upon the uric acid excretion of man, 1. BioI. Chern. 55:605-610, 1923. Hook, J. B., and Williamson, H. E.: Influences of probenecid and alterations in acid-base balance on the saluretic activity of furosemide, J. Pharmacol. Exp. Ther. 149:404-408, 1972. Portwich, V. G., Buttner, H., and Schafer, 1.: Simultane Untersuchung der Clearance und renalen Extraktion von Pharmaka, in Ullrich, K. S., and Hierholzer, K., editors: Normale und pathologische Funktionen des Neirentubulus, Bern, 1965, Hans Huber, pp. 173-176. Puschett, J. B., and Goldberg, M.: The acute effects of furosemide on acid and electrolyte excretion in man, 1. Lab. Clin. Med. 71:666667, 1968. Rose, H. 1., Pruitt, A. W., Dayton, P. G., and
16.
17.
18.
19.
401
McNay, 1. L.: Relationship of urinary furosemide excretion rate to natriuretic effect in experimental azotemia, 1. Pharmacol. Exp. Ther. 199:490-497, 1976. Senft, G.: Membrantransport and Pharmaka, in Ullrich, K. S., and Hierholzer, K., editors: Norrnale und pathologische Funktionen des Nierentubulus, Bern, 1965, Hans Huber, pp. 63-81. Suki, W., Rector, F. C., Jr., and Seldin, D. W.: The site of action of furosemide and other sulfonamide diuretics in the dog, J. Clin. Invest. 44:1458-1469, 1965. Weiner,!. M., Washington, J. A., and Mudge, G. H.: On the mechanism of action of probenecid on renal tubular secretion. Bull. Johns Hopkins Hosp. 106:333-346, 1960. Yii, F. T., Sirota, J. H., Berger, L., Halpern, M., and Gutman, A. B.: Effect of sodium lactate infusion on urate clearance in man, Proc. Soc. Exp. BioI. Med. 96:809-813, 1957.
