June 1991
Volume 80, Number 6
JOURNAL OF PHARMACEUTICAL SCIENCES A publication of the American Pharmaceutical Association
A RTlCLES
Effects of Ethanol on the Pharmacokinetics of Cephalexin and Cefadroxil in the Rat J. P. BARRIO LERA’, A. I. ALVAREZ,AND J. G.PRlETO Received January 23, 1989, from the Laboratorio de Fisiologia Animal, Universidad de Ledn, 24071 Ledn, Spain.
Accepted for publication
July 26, 1990.
Abstract 0 Data are presented on the effect of ethanol on the intestinal
absorption and excretion in rats of two p-lactam antibiotics, cephalexin (CFX)and cefadroxil (CFD). A recirculating perfusion technique within an antibiotic concentration range of 0.5 to 50 mM was used. Ethanol was administered either in an acute form into the intestine or in a chronic form as a 15% drinking solution for 2 months. The results are normalized in relation to the metabolic body weight, intestinal length, and osmotic conditions. Acute ethanol treatment decreases the antibiotic absorption; biliary excretion of CFD is increased, while urinary excretion of CFX is lowered. Chronic treatment shows slight negative effects on the absorption of CFX and CFD. Results are interpreted on the basis of the effect of ethanol on biological membranes. Enhanced urinary excretion after acute ethanol treatment, as well as differences between transport mechanisms, are invoked to explain these effects.
All experimental work on the absorption of pharmacologically active and exogenous substances should be carried out with some estimation of the fate of such compounds in the organism as well as their routes of elimination. In the case of p-lactam antibiotics, the urine is known to be the major excretory route.’ The elimination rates of the active principle (in most cephalosporins) or the degradation products (as with penicillins) a t a given time in urine are also known.2 Tissue distribution of these substances is less well understood; such studies require great experimental effort and complicated mathematical calculations. A number of techniques using either whole animals in situ (“single-pass perfusion,”3 “recirculating perfusion”4.5) or isolated intestinal segments (“everted sacs”6) have been developed to investigate the intestinal absorption of substances. Advantages and disadvantages of these and related approaches have been reviewed elsewhere.7 There is an increasing number of papers dealing with more cell-directed (or even membrane-directed) techniques, such as the use of isolated enterocytes,s membrane vesicles, or liposomes.9 To allow a full understanding of the absorption of a drug or nutrient into the intestinal wall cells (assessing the importance of the brush border region or the ion cotransport phenomena), it is necessary to carry out these kinds of studies. 0022-3549/91/0600-05 1 1$01 .0010 0 1991,American Pharmaceutical Association
But there are still a number of experimental situations where a whole-animal approach is needed, at least as an a priori, first-time study. The transport of p-lactam antibiotics such as cephalexin (CFX) and cefadroxil (CFD) from the intestinal lumen to the blood stream has been extensively studied.5+13 However, no study has been reported on the effects of ethanol on this process. A few workers have studied the intestinal absorption of drugs under the influence of ethanol.lG17 Similarly, the excretion of antibiotics into both the urinary and biliary fluids in the rat is not well documented. While some work has been published on the extension of diuresis in the presence of ethano1,lBno studies have been found that determine whether or not this agent affects antibiotic excretion through bile or urine. This paper compares the effects of ethanol (administered in acute and chronic forms to Wistar rats) on the intestinal absorption and biliary and urinary excretion characteristics of two structurally similar p-lactam antibiotics, CFX and CFD, using variable initial drug concentrations perfused through cannulated, in situ rat duodenum. Since other authors14,15 have not found effects of low ethanol concentrations (0.5%, 2%) on the absorption of several drugs, a higher concentration (6%)of ethanol (similar to that of mild spirits) was chosen for the acute treatment, while the combination of a larger dose (15%)and 2 months of oral administration were the conditions for the chronic treatment.19
Experimental Section Experimental Animals-Female Wistar rats (Rattus noruegicus) were used. A mating schedule was planned to match the different litters to the presumed starting date of the laboratory experiments. Animals from the same litter were grouped from the time of weaning in order to provide enough replicates for the experiments. The animals were offered commercial chow and tap water ad libitum until they reached a body weight of 250 ( 2 30)g. ReagentsCephalexin [CFX;7-(D-cu-amino-a-phenyl acetamidol3-methyl-3-cephem-4-carboxylic acid] and cefadroxil [CFD; 74D-aamino-~-hydroxyphenylacetamido~-3-methyl-3-cephem-4-carboxylic acid] were kindly supplied by the Research Department of Antibi6ticos, S.A. (Le6n, Spain). The initial antibiotic concentrations that were perfused ranged between 0.5 and 50 mM. The upper limit was Journal of Pharmaceutical Sciences I 51 1 Vol. 80, No. 6, June 1991
determined by the solubility of CFX in the buffer solution. Ethanol concentrations were determined spectrophotometrically using alcohol dehydrogenase (Boehringer Mannheim, S.A., Barcelona, Spain) and nicotinamide-adenine-dinucleotide(NAD’; Sigma Chemical Company, St. Louis, MO) according to the procedure of Bergmeyer.2D Ethanol Administration-Ethanol was administered either in acute or chronic form. Acute ethanol treatment involved the incorporation of 6% (dv) ethanol in the antibiotic perfusion solution (see below); the chronic treatment provided 15% (v/v) ethanol as the only drinking source for 60 days after 10 days adaptation with 5 and 10% ethanol. The control animals were offered tap water and were pair-fed to control solid intake16 during the chronic ethanol treatment. Perfusion Techniqu-The recirculating perfusion technique was adapted from techniques used by other authors.5.11,13,21.22The method involves the use of a intestinal segment (-15-cm long), located just below the biliary duct, through which the antibiotic solution is continuously circulated. Briefly, the operating technique is as follows. The animal is anesthetized using sodium thiopenthal (Pentothal, Abbott Laboratories, S.A., Madrid, Spain) at an ip dose of 50 mg/mL (0.1 mL per 100 g of body weight); at least 30 min is required to complete the anesthesia, during which time the animal lies on a platform heated to body temperature. Laparatomy is then performed, and the biliary duct is cannulated. Input and output cannulae are placed a t both ends of the selected intestinal segment. The perfusion solution comes via a peristaltic pump delivering 1 mL * min-’ from a continuously shaken flask. The circuit is closed, returning the output to the flask. All tubing used was silicone tubing. The perfusion solutions consisted of isotonic phosphate buffer (0.05 M, pH 6.9,ionic strength 0.15 with sodium sulfate) which dissolved the different initial concentrations of antibiotic used, 0.2% phenol red to monitorize volume changes, and (in acute ethanol treatment experiments) 6% ethanol. The pH ofthe perfused fluid fell to 6.2 a t the higher antibiotic concentration (50 mM). The solutions were heated to -35°C using a warm-water mechanism to heat the tubing immediately before entering the intestinal segment. After a first washing of the circuit using 30 mL of antibiotic-free buffer solution and filling with the perfusion solution, aliquots of the intestinal contents and the bile and urine were taken every 15 min. The perfusion lasted for a maximum time of 2 h. Three replicate perfusion experiments were made for the entire range of initial antibiotic concentrations, using a different animal in each case. Antibiotic Determination-Quantification of the remaining antibiotic in the samples (intestinal recirculate, bile, urine) was performed using reversed-phase HPLC. The samples were suitably deproteinized and filtered prior to their analysis at the following conditions: Nova-Pak C18 column (Waters Millipore, S.A., Madrid, Spain);mobile phase of 70% phosphate buffer (0.1 M, pH 6.0)and 30% methanol, UV detection a t 254 nm; a flow rate of 1 mL/min; and a maximum pressure of 130 bar. Measured retention times were 2.26 & 0.12 and 3.42f 0.21 min for CFD and CFX, respectively. The antibiotic was monitored using peak height values with twofold injections of each sample. These values yielded concentration values by applying calibration curves obtained from known antibiotic concentrations. In the case of bile and urine samples, the time interval between injections onto the HPLC column was much longer due to the increased retention times of some of the bile and urine constituents as related to the retention times of the antibiotics. Computational Techniques-The effect of variable antibiotic concentrations on the intestinal absorption rate was used to study the absorption mechanisms of CFX and CFD. To calculate the absorption rate for a given initial antibiotic concentration, the remaining antibiotic concentration in the perfusate at each sampling time was corrected t o ideal conditions without osmotic water movement, using phenol red as a tracer. Linear regression was used to estimate the apparent absorption rate constants (/zap), using the natural logarithm of the remaining corrected antibiotic concentration as the independent variable and the sampling time as the dependent variable. This was also the method used to estimate the absorption rate constants of ethanol administered in the acute form. Natural logarithms of the remaining antibiotic concentrations for each perfusion experiment were plotted against the sampling time; the slope of the straight lines thus obtained represent the apparent rate constant (/zap) for the disappearance of antibiotic from the intestinal lumen at a given initial perfusion concentration. In order to analyze the dependence of these constants on the initial antibiotic 512 I Journal of Pharmaceutical Sciences Vol. SO, No. 6, June 1991
concentration, apparent rate constants were transformed into corresponding values a t initial conditions by multiplying them by the initial antibiotic concentration in each case. These initial rate values ( u ) were finally normalized on the intestinal unstretched, wet segment length and the metabolic weight of the corresponding animal. Metabolic weight was considered to be the 0.75 power of the live body weight, following the procedures of other a~thors.23~24 Absorption Kinetics-Several facts point to a saturable intestinal absorption mechanism for cephalosporins. First, the dose dependency of the absorption and the existence of specific absorption areas suggest some kind of specialized, saturable transport4 following Michaelis-Menten kinetics. Secondly, competitive inhibition has been shown for a series of zwitterionic antibiotics.25 In other respects, in the series of papers by Tsuji e t a1.4.11.26.27and Nakashima et al.5 on the intestinal absorption of penicillins and cephalosporins, these authors refer to a mixed-type hyperbolic first-order equation with two different components: a saturating term, similar t o the MichaelisMenten equation, and a linear term analogous to a first-order kinetic equation. Thus, two kinetic models have been proposed to explain the absorption characteristics of cephalosporins: the well-known MichaelisMenten model and the so-called Mixed model. The kinetic equations for these models are shown in eqs 1 and 2, respectively:
(1)
(2) where v represents the apparent absorption rate, V,, is the theoretical maximum absorption rate, K , is the Michaelis-Menten constant, and K is a first-order kinetic term accounting for the simple diffusional aspect of the Mixed model, as well as the intraluminal nonenzymatic degradation of the antibiotic. To obtain estimates of the kinetic parameters of the intestinal absorption, a nonlinear regression fitting program (PCNONLIN, Statistical Consultants, 1986) incorporating iterative residual weighting was used. The t test was used to assess the statistical significance of the kinetic parameters studied in control and treatment conditions, using the parameter values as population means and assuming equality between the parameters as the HO hypothesis. The t values were obtained as the ratio [parameter value]/ [standard errorl.28.29 In order to ascertain a suitable model for the intestinal absorption of CFX and CFD, attention was given to the values of the error standard deviation (S) and the sum ofthe weighted squares residuals (SWRS) that are shown in Table I. Excretion-Biliary and urinary excretion of CFX and CFD was measured from the maximum antibiotic concentration found in the bile and urine samples (MAC, and MAC,, respectively) for each perfusion. These antibiotic concentrations were obtained by quantifying the corresponding peaks in the HPLC analysis of the samples. Table I-Comparison between Mlxed and Mlchaells-Menten Models using the Statistics “Error Standard Deviation” (S) and “Sum of Weighted Squared Residuals” (SWRS) for Cephalexln (CFX) and Cefadroxll (CFD) in Acute and Chronic Controls and Ethanol Treatments Treatment Acute Control CFD Control CFX Ethanol CFD Ethanol CFX Chronic Control CFD Control CFX Ethanol CFD Ethanol CFX
Michaelis-Menten
Mixed
S
SWRS
S
SWRS
0.2732 0.2695 0.1950 0.2300
1.1946 1.3487 0.3427 0.6876
0.2799 0.2664 0.1851 0.2209
1.1754 1.3077 0.3422 0.5859
0.2475 0.2349 0.2526 0.2661
0.7552 0.5050 0.6383 0.7080
0.2748 0.2247 0.2586 0.2582
0.5513 0.4967 0.6021 0.6002
To quantify the relation among the initial antibiotic concentration (IAC) administered to tbe rats and the maximum concentrations (MAC) excreted in each experiment, linear equations were obtained using MAC data as the independent variable and IAC data as the dependent variable. The slope of the straight lines was taken as a measurement of the excretory efficiency for each excretion route, antibiotic, and treatment. These values are summarized as Eb (excretoryefficiency for bile) and E, (excretoryefficiency for urine) in Table 11.
Results Absorption Kinetics-Tables I11 and IV summarize the results from the fitting of our data (see Figures 1and 2) to the Michaelis-Menten and Mixed models, respectively, as the nonlinear regression parameters and their corresponding standard errors for each experimental treatment. It is clear that SWRS values are consistently smaller for the Mixed model fitting than in the Michaelis-Menten model; but, that is to be expected because of the existence of one more parameter in the Mixed model. However, we can use S values to help discriminate the validity of either Mixed or MichaelisMenten models (Table I). Looking a t the S values for CFX, the Mixed fitting yields smaller values than the MichaelisMenten model, while the opposite can be said when looking at the corresponding values for CFD. There is a single exception, in the acute ethanol treatment for CFD, when the S value is smaller for the Mixed model. At the same time, the results of fitting our CFD data to the Mixed model gave negative values for the parameter K of simple diffusion (data not shown), except in the acute treatment with ethanol. These results confirm the idea that the only relevant parameters in that and K,, and, accordingly, Table IV shows fitting were V,, values for V,, and K , obtained from the Michaelis-Menten fitting for the acute and chronic controls and chronic ethanol treatment for CFD absorption. In the case of the Michaelis-Menten model (Table 1111, V, values for CFD and CFX decrease slightly but significantly both with acute and chronic ethanol treatment. The values of K , increase for the two antibiotics under the ethanol treatments, except in the acute ethanol treatment for CFX. On the contrary, following the Mixed model fitting (Table IV), different effects between the results from acute and chronic treatment do appear. First, K, values decrease as a consequence of the acute and chronic treatments for CFX and after the acute treatment decreases significantly in the for CFD. The parameter V,, case of both antibiotics after acute ethanol treatment, while chronic treatment only reveals statistical differences for CFX. The significance of the K parameter can only be assessed for CFX. This parameter reveals significant changes following only acute treatment with ethanol. Excretion-Table I1 lists excretory parameters E, (excretory efficiency for bile) and E, (excretory efficiency for urine),
as well as the corresponding standard error values and the number of animals (experimental points) used; Figures 3 and 4 show these data in a bar-chart format. Urinary excretion of cephalosporins predominates over biliary excretion. Our data show an increase of >lO-fold for urinary E, compared with E,. Thus, by supposing an intestinal segment 15 cm long and a body weight of 200 g, our data assume a maximum biliary concentration of CFX (after 120 min of perfusion) of 43.6 pg * mL-' and an urinary maximum concentration of the same antibiotic of 522 pg * mL-'. These data agree with those already obtained in human beings. For example, levels ranging from 2 t o 40 pg * mL-l of CFX in bile have been observed after an initial dose of 500 mg, while 99% of the dose can be recovered in urine after 5 h.4 Both biliary and urinary excretion efficiencies are better for CFX than for CFD. Kwan and Rogers2 reported an elimination of nearly 100% of a single dose of orally administered CFX (recovered in urine after 6-12 h), while the amount of CFD found in urine 24 h after its administration ranged between 85 and 93%.However, Nightingale et al.30 report a similar renal elimination for both antibiotics after 12 h, but the excretion rate is less for CFD when the time of study is limited to 3 h. The same behavior was previously found by Hartstein et al.1 in a study of the urinary excretion of CFX and CFD after a single po dose of 500 mg in humans. Although these authors use cumulative antibiotic excretion data and not maximum concentration data, it is possible that our data could fit their schema, given our limit of 2 h of intestinal perfusion.
Discussion Our results suggest a saturable mechanism for the intestinal absorption of CFX and CFD, as already pointed out by other authors.5~9~31 This study was made using the maximum antibiotic concentrations permitted by the solubility at the conditions described above (-50 mM), thus raising the level to several times the K, values already reported. It is to be noted that other authors,lO who used maximum antibiotic concentrations of
2 Y
0
10
20
30
40
50
0’ 0
60
W I (mM)
I
10
20
30
40
50
60
W D I (mM)
Figure 1-Relationship between the absorption rate (v) and the initial concentration of CFX perfused in each experimental treatment after fitting to the Mixed model. The plot for the control of the acute treatment acute; (A) chronic. is shown (solid line). Key: (0)control; (0)
Flgure 2-Relationship between the absorption rate (v) and the initial concentration of CFD perfused in each experimental treatment after fitting to the Mixed model. The plot for the control of the acute treatment is shown (solid line). Key: See Figure 1 .
agreement w i t h our data using the same fitting model (0.062 mM * min-’ for C F X and 0.078 mM * min-’ for CFD, assuming an intestinal segment 15 c m long and a 200-g rat). O u r results appear higher than those o f these authors, but no
centration higher than that theoretically reachable in the absence of osmotic effects. O n the other hand, at hypertonic concentrations, a decrease in the concentration w i l l be seen due to the intestinal secretion o f water. So, in the absence of osmotic corrections, V,,, and K , w i l l be lower than after the referred volume change correction. Castaner-Codes et al.31 report data on the intestinal transport o f C F X and CFD that i s also c o n c o r d a n t w i t h o u r s (171.35 and 316.13 g mL-’ min-’ as V,,, values f o r CFX and CFD, respec-
mention i s made of control of osmotic volume changes in their recirculatingperfusion experiments, which also lasted for 120 min. In fact, it i s clear t h a t for hypotonic antibiotic concentrations the intestinal w a l l w i l l absorb water from the perfusing solution, thus yielding a remnant antibiotic con514 I Journal of Pharmaceutical Sciences Vol. 80, No. 6, June 1991
-
-
ACUTE
CHRONIC T
- ctrl - -EtOH-
- ctrl - -EtOH-
Figure 3-Relationship
between the biliary excretory efficiencies (cm-' .g-') for cephalexin (a)and cefadroxil (N) in acute and chronic ethanol treatments, and their respective controls. The asterisk indicates significant differences at p < 0.01 (data are from Table I). ACUTE
CHRONIC
T
* 0
C Q
0
Ea
2
:
+
eV
x a
l
?
-
0 C
L
= ) " n
rameters, K , and V,,,, decrease when ethanol is administered in the acute form, while chronic ethanol treatment seems to have no clear effects on the absorption kinetics. The decrease in the activity of the carrier for CFX and CFD could be explained as a indirect consequence of the effect of ethanol on plasma membranes. So, Beck and Dinda32 suggest conformational changes induced by ethanol on disaccharidase of the brush border membrane to explain the decrease in activity of this enzyme upon acute treatment with ethanol. Uridine, choline, and chenodesoxicholate carriers could also be inactivated by changes in plasma membrane fluidity.6 Recent data33 show a direct effect of chronic ethanol administration on purified glycosidase activities in rat liver and brain, with changes in kinetic profile and activation energy. There is a need for more in depth studies in relation to the effect of ethanol on intestinal transport. Possible areas should include the evaluation of the direct toxic effect of ethanol on isolated hepatocytes. Ethanol Effects on Excretion-In drawing conclusions on the intestinal absorption for CFD and CFX, it should be noted that intestinal absorption, as measured by the V,,, parameter, is better for CFD than for CFX assuming a mixed kinetic model. This gives support to the hypothesis of different transfer mechanisms for these antibiotics in the gut and the liver or kidney, with both urinary and biliary excretion being possible first-order kinetic processes, which could explain the higher excretion values obtained for CFX than for CFD. Acute treatment with ethanol shows an increase in E, in the case of CFD, with decreases in E, for CFX (Table 11). Chronic treatment with ethanol does not produce significant differences. The effect of the acute treatment with ethanol on urinary excretion of CFX could be accounted for by induced enhanced urinary flow. Pohoreckyle reported an almost twofold increase in urine output by rats following an intragastric load of up to 2.5 g kg-' of ethanol, but chronically treated animals excreted a slightly lower volume of urine than did the controls. The almost normal urine volume for chronically treated rats could explain the lack of effect of chronic treatment on the urinary excretion of CFX and CFD. At present we have no explanation for the differential effects of the acute treatment with ethanol on the biliary excretion efficiencies of CFD and CFX. These could be attributed to the different chemical nature of the antibiotics, which could modify the general elimination path. Possibly, given the fact that both CFX and CFD are excreted in a form that is basically unmodified chemically, the presence of the additional parahydroxyl group in CFD compared with CFX could induce hepatic permeability changes which could explain their different excretory efficiencies.
- ctrl - -EtOH-
- ctrl - -EtOH-
Flgure +Relationship between. the urinary excretory efficiencies (cm-' g-') for cephalexin (a)and cefadroxil (N)in acute and chronic ethanol treatments, and their respective controls. The asterisk indicates significant differences at p < 0.01 (data are from Table I).
tively, equivalent to 0.047 and 0.083 mM min-', respectively) but these authors fit their data to a Michaelian-type pharmacokinetic model. From the above discussion, and assuming the hypothesis of the existence of a common carrier system for aminocephalosporins, as repeatedly reported,5,9.25 our results point to a better intestinal absorption for CFD than for CFX. Because of the higher hydrophobicity of CFX (the isobutilic-water partition coefficient for CFX is more than twice than for C F D 9 , this could be indicative of a hydrophilic microenvironment for the cephalosporin intestinal carrier. In keeping with this idea, we can consider the absence of proper values for the K parameter seen in this work (related to free diffusion transport) in the case of CFD (Table IV) as being due to the inadequacy of the fitting in the case of a negligible contribution of the first-order kinetic term of the Mixed model to the absorption of CFX; that is, this antibiotic would be absorbed solely by the carrier mechanism, avoiding purely passive membrane transport. Ethanol Effects on Absorption-A clear effect of ethanol is to raise the K parameter for CFX after acute treatment. This is also the only situation when the Mixed model can be applied to CFD transport. The increase in the passive diffusion component could be attributed to a higher permeability of the intestinal membranes induced by ethanol. Both kinetic pa-
References and Notes 1. Hartstein, A. I.; Patrick, K. E.; Jones, S. R.; Miller,M. J.;Bryant, R. E. Antimicrob. Agents Chemother. 1977,12,93-97. 2 . Kwan, K. C.; Rogers, J. D. In Antibiotics Containing the P-Lactam Structure ZZ Demain, A. L.; Solomon, N. A., Eds.; SpringerVerlag A G Berlin, 1983;pp 278-301. 3. Doluisio, J.T.; Billups, N. F.; Dittert, L. W.; Sugita, E. T.; Swintoskv. J. V. J . Pharm. Sci. 1969.58. 1196-1200. 4. Tsui,' A.; Miyamoto, E.; Kubo, 0.;Yamana, T. J . Pharm. Sci. 1979,68,812-816. 5. Nakashima, E.; Tsuji, A.; Kagatani, S.; Yamana, T. J.Phurm. Dyn. 1984,7,452-464. 6. Wilson, T. H.; Wiseman, G. J . Physiol. 1954,123, 116125. 7. Schurgers, N.; Bijdendijk, J.; Tukker, J. J.; Crommelin, D. J. A. J.Pharm. Scz. 1986,75,117-119. 8. L6 ez del Pino, V.; Hegazy, E.; Hauber, G.;Rernmer, H.; Sciwenk, M.Arch. Toxicol. 1983,Suppl. 6,322326. 9. Kimura, T.; Yamamoto, T.; Ishizuka, R.; Sezaki, H. Biochem. Pharmucol. 1985,34(1), 81-84. 10. Umeniwa, T.; Ogino, 0.; Miyazaki, K.; Arita, T. Chem. Pharm. Journal of Pharmaceutical SciencesI 515 Vol. 80, No. 6, June 7997
Bull. 1979,27,2177-2182. 11. Tsuji, A.; Yamana, T. In @Lactam Antibiotics; Mitsuhashi, S., Ed.: SDrinper-Verlae: New York. 1981:D 235. 12. Tsuji, A.;Gakashim;?, E.; Deguchi, Y.; Nishide, K.; Shimizu, T.; Horiuchi, S.; Ishikawa, K.; Yamana, T. J.Pharm. Sci. 1981,70, 1120-1128. 13. Prieto, J. G.; Barrio, J. P.; Alvarez, A. I.; Zapico, J. Reu. Farmacol. Clin. Exp. 1986,3,99-104. 14. Magnussen, M. P. Acta Phurmacol. Toxicol. 1968.26,130-144. 15. Hayton, W.L. J. Phurm. Sci. 1975,64,1450-1456. 16. Lane, E. A.; Guthrie, S.; Linnoila, M. Clin.Pharmacokinet. 1985, 10,228-247. 17. Larivikre, L.; CaillB, G.; Elie, R. Biopharm. Drug Dispos. 1986,7 , 201-210. 18. Pohoreckv. L. A. Alcohol 1985.2. 659-666. 19. Racusen,-L. C.; Krawitt, E. L. A m . J. Digest. Dis. 1977, 22, 915-920. 20. Ber e er H U. In Methods ofEnz m t i c Analysis; Bergmeyer, H. id.: Verlag Chemie WeinieidAcademic: New York, 1974;pp 1499-1502. 21. Ponz, F.; Ilundain, A.; Lluch, M. Rev. Esp. Fisiol. 1979, 35, 97-104. 22. Prieto, J. G.; Santos, L.; Barrio, J. P.; Alvarez, M. L. J. Pharm. Sci. 1987,76,596-598. 23. Seers,R.;Michels, H. InEmrgy MetabolismofFarm Animals; Ekern,
516 I Journal of Pharmaceutical Sciences Vol. 80, No. 6, June 1991
A.: Sundst~l.F.. Eds.: EAAP 29:Norwav. 1982 DD 135-137. 24. Saadeh, G. H.;‘Mu& M.; Nassar, C. F: Comp: Eiochem. Physwl. 1986,83A,539-541. 25. Miykaki, K.; Ohtani, K.; Umeniwa, K.; Arita, T. J.Phurm. Dyn. 1982.5. , - .555-563. ~~- - - 26. Tsuji, A.; Miyamoto, E.; Hashimoto, N.; Yamana, T. J.Phurm. Sci. 1978.67,1705-1711. 27. Tsuji, A.; Nakashima, E.; Kagami, I.; Nakashima, K.; Yamana, T. J. Pharm. Phurmucol. 1979,31,718-720. 28. Snedecor, G. W.; Cochran, W. G. Mdtodos estadtsticos, 1st Ed.; Compafifa Editorial Continental: S.A., MBxico D. F., 1975;p 131. 29. Sokal, R. R.; Rohlf, F. J. Biometrta. Princzpiosy mitodos estadlsticos en la inuestigacidn biol6gica; Edic. Edit. Blume: Barcelona, Spain, 1979. 30. Nightingale, C.H.; French, M. A.; Quintiliani, R. In pLactam Antibiotics; Mitsuhasi, S., Ed.; Springer-Verlag: New York, 1981; pp 259-267. 31. Castafier-Codes, A.; Merino, M.; Martin, A.; Obach, R.; PIBDelfina, J. M. Eur. Congress Biophurm. Phurmcokin. 1984,2, 170-179. 32. Beck, I. T.; Dinda, P. K. A m . J. Digest. Dis. 1981,26,817-838. 33. Gil Martin, E.; FernBndez-Briera, A.; Calvo Ferntindez, P. Biochem. Pharmucol. 1990,40,925-982. ~~
~~
Volume 80, Number 6
JOURNAL OF PHARMACEUTICAL SCIENCES A publication of the American Pharmaceutical Association
A RTlCLES
Effects of Ethanol on the Pharmacokinetics of Cephalexin and Cefadroxil in the Rat J. P. BARRIO LERA’, A. I. ALVAREZ,AND J. G.PRlETO Received January 23, 1989, from the Laboratorio de Fisiologia Animal, Universidad de Ledn, 24071 Ledn, Spain.
Accepted for publication
July 26, 1990.
Abstract 0 Data are presented on the effect of ethanol on the intestinal
absorption and excretion in rats of two p-lactam antibiotics, cephalexin (CFX)and cefadroxil (CFD). A recirculating perfusion technique within an antibiotic concentration range of 0.5 to 50 mM was used. Ethanol was administered either in an acute form into the intestine or in a chronic form as a 15% drinking solution for 2 months. The results are normalized in relation to the metabolic body weight, intestinal length, and osmotic conditions. Acute ethanol treatment decreases the antibiotic absorption; biliary excretion of CFD is increased, while urinary excretion of CFX is lowered. Chronic treatment shows slight negative effects on the absorption of CFX and CFD. Results are interpreted on the basis of the effect of ethanol on biological membranes. Enhanced urinary excretion after acute ethanol treatment, as well as differences between transport mechanisms, are invoked to explain these effects.
All experimental work on the absorption of pharmacologically active and exogenous substances should be carried out with some estimation of the fate of such compounds in the organism as well as their routes of elimination. In the case of p-lactam antibiotics, the urine is known to be the major excretory route.’ The elimination rates of the active principle (in most cephalosporins) or the degradation products (as with penicillins) a t a given time in urine are also known.2 Tissue distribution of these substances is less well understood; such studies require great experimental effort and complicated mathematical calculations. A number of techniques using either whole animals in situ (“single-pass perfusion,”3 “recirculating perfusion”4.5) or isolated intestinal segments (“everted sacs”6) have been developed to investigate the intestinal absorption of substances. Advantages and disadvantages of these and related approaches have been reviewed elsewhere.7 There is an increasing number of papers dealing with more cell-directed (or even membrane-directed) techniques, such as the use of isolated enterocytes,s membrane vesicles, or liposomes.9 To allow a full understanding of the absorption of a drug or nutrient into the intestinal wall cells (assessing the importance of the brush border region or the ion cotransport phenomena), it is necessary to carry out these kinds of studies. 0022-3549/91/0600-05 1 1$01 .0010 0 1991,American Pharmaceutical Association
But there are still a number of experimental situations where a whole-animal approach is needed, at least as an a priori, first-time study. The transport of p-lactam antibiotics such as cephalexin (CFX) and cefadroxil (CFD) from the intestinal lumen to the blood stream has been extensively studied.5+13 However, no study has been reported on the effects of ethanol on this process. A few workers have studied the intestinal absorption of drugs under the influence of ethanol.lG17 Similarly, the excretion of antibiotics into both the urinary and biliary fluids in the rat is not well documented. While some work has been published on the extension of diuresis in the presence of ethano1,lBno studies have been found that determine whether or not this agent affects antibiotic excretion through bile or urine. This paper compares the effects of ethanol (administered in acute and chronic forms to Wistar rats) on the intestinal absorption and biliary and urinary excretion characteristics of two structurally similar p-lactam antibiotics, CFX and CFD, using variable initial drug concentrations perfused through cannulated, in situ rat duodenum. Since other authors14,15 have not found effects of low ethanol concentrations (0.5%, 2%) on the absorption of several drugs, a higher concentration (6%)of ethanol (similar to that of mild spirits) was chosen for the acute treatment, while the combination of a larger dose (15%)and 2 months of oral administration were the conditions for the chronic treatment.19
Experimental Section Experimental Animals-Female Wistar rats (Rattus noruegicus) were used. A mating schedule was planned to match the different litters to the presumed starting date of the laboratory experiments. Animals from the same litter were grouped from the time of weaning in order to provide enough replicates for the experiments. The animals were offered commercial chow and tap water ad libitum until they reached a body weight of 250 ( 2 30)g. ReagentsCephalexin [CFX;7-(D-cu-amino-a-phenyl acetamidol3-methyl-3-cephem-4-carboxylic acid] and cefadroxil [CFD; 74D-aamino-~-hydroxyphenylacetamido~-3-methyl-3-cephem-4-carboxylic acid] were kindly supplied by the Research Department of Antibi6ticos, S.A. (Le6n, Spain). The initial antibiotic concentrations that were perfused ranged between 0.5 and 50 mM. The upper limit was Journal of Pharmaceutical Sciences I 51 1 Vol. 80, No. 6, June 1991
determined by the solubility of CFX in the buffer solution. Ethanol concentrations were determined spectrophotometrically using alcohol dehydrogenase (Boehringer Mannheim, S.A., Barcelona, Spain) and nicotinamide-adenine-dinucleotide(NAD’; Sigma Chemical Company, St. Louis, MO) according to the procedure of Bergmeyer.2D Ethanol Administration-Ethanol was administered either in acute or chronic form. Acute ethanol treatment involved the incorporation of 6% (dv) ethanol in the antibiotic perfusion solution (see below); the chronic treatment provided 15% (v/v) ethanol as the only drinking source for 60 days after 10 days adaptation with 5 and 10% ethanol. The control animals were offered tap water and were pair-fed to control solid intake16 during the chronic ethanol treatment. Perfusion Techniqu-The recirculating perfusion technique was adapted from techniques used by other authors.5.11,13,21.22The method involves the use of a intestinal segment (-15-cm long), located just below the biliary duct, through which the antibiotic solution is continuously circulated. Briefly, the operating technique is as follows. The animal is anesthetized using sodium thiopenthal (Pentothal, Abbott Laboratories, S.A., Madrid, Spain) at an ip dose of 50 mg/mL (0.1 mL per 100 g of body weight); at least 30 min is required to complete the anesthesia, during which time the animal lies on a platform heated to body temperature. Laparatomy is then performed, and the biliary duct is cannulated. Input and output cannulae are placed a t both ends of the selected intestinal segment. The perfusion solution comes via a peristaltic pump delivering 1 mL * min-’ from a continuously shaken flask. The circuit is closed, returning the output to the flask. All tubing used was silicone tubing. The perfusion solutions consisted of isotonic phosphate buffer (0.05 M, pH 6.9,ionic strength 0.15 with sodium sulfate) which dissolved the different initial concentrations of antibiotic used, 0.2% phenol red to monitorize volume changes, and (in acute ethanol treatment experiments) 6% ethanol. The pH ofthe perfused fluid fell to 6.2 a t the higher antibiotic concentration (50 mM). The solutions were heated to -35°C using a warm-water mechanism to heat the tubing immediately before entering the intestinal segment. After a first washing of the circuit using 30 mL of antibiotic-free buffer solution and filling with the perfusion solution, aliquots of the intestinal contents and the bile and urine were taken every 15 min. The perfusion lasted for a maximum time of 2 h. Three replicate perfusion experiments were made for the entire range of initial antibiotic concentrations, using a different animal in each case. Antibiotic Determination-Quantification of the remaining antibiotic in the samples (intestinal recirculate, bile, urine) was performed using reversed-phase HPLC. The samples were suitably deproteinized and filtered prior to their analysis at the following conditions: Nova-Pak C18 column (Waters Millipore, S.A., Madrid, Spain);mobile phase of 70% phosphate buffer (0.1 M, pH 6.0)and 30% methanol, UV detection a t 254 nm; a flow rate of 1 mL/min; and a maximum pressure of 130 bar. Measured retention times were 2.26 & 0.12 and 3.42f 0.21 min for CFD and CFX, respectively. The antibiotic was monitored using peak height values with twofold injections of each sample. These values yielded concentration values by applying calibration curves obtained from known antibiotic concentrations. In the case of bile and urine samples, the time interval between injections onto the HPLC column was much longer due to the increased retention times of some of the bile and urine constituents as related to the retention times of the antibiotics. Computational Techniques-The effect of variable antibiotic concentrations on the intestinal absorption rate was used to study the absorption mechanisms of CFX and CFD. To calculate the absorption rate for a given initial antibiotic concentration, the remaining antibiotic concentration in the perfusate at each sampling time was corrected t o ideal conditions without osmotic water movement, using phenol red as a tracer. Linear regression was used to estimate the apparent absorption rate constants (/zap), using the natural logarithm of the remaining corrected antibiotic concentration as the independent variable and the sampling time as the dependent variable. This was also the method used to estimate the absorption rate constants of ethanol administered in the acute form. Natural logarithms of the remaining antibiotic concentrations for each perfusion experiment were plotted against the sampling time; the slope of the straight lines thus obtained represent the apparent rate constant (/zap) for the disappearance of antibiotic from the intestinal lumen at a given initial perfusion concentration. In order to analyze the dependence of these constants on the initial antibiotic 512 I Journal of Pharmaceutical Sciences Vol. SO, No. 6, June 1991
concentration, apparent rate constants were transformed into corresponding values a t initial conditions by multiplying them by the initial antibiotic concentration in each case. These initial rate values ( u ) were finally normalized on the intestinal unstretched, wet segment length and the metabolic weight of the corresponding animal. Metabolic weight was considered to be the 0.75 power of the live body weight, following the procedures of other a~thors.23~24 Absorption Kinetics-Several facts point to a saturable intestinal absorption mechanism for cephalosporins. First, the dose dependency of the absorption and the existence of specific absorption areas suggest some kind of specialized, saturable transport4 following Michaelis-Menten kinetics. Secondly, competitive inhibition has been shown for a series of zwitterionic antibiotics.25 In other respects, in the series of papers by Tsuji e t a1.4.11.26.27and Nakashima et al.5 on the intestinal absorption of penicillins and cephalosporins, these authors refer to a mixed-type hyperbolic first-order equation with two different components: a saturating term, similar t o the MichaelisMenten equation, and a linear term analogous to a first-order kinetic equation. Thus, two kinetic models have been proposed to explain the absorption characteristics of cephalosporins: the well-known MichaelisMenten model and the so-called Mixed model. The kinetic equations for these models are shown in eqs 1 and 2, respectively:
(1)
(2) where v represents the apparent absorption rate, V,, is the theoretical maximum absorption rate, K , is the Michaelis-Menten constant, and K is a first-order kinetic term accounting for the simple diffusional aspect of the Mixed model, as well as the intraluminal nonenzymatic degradation of the antibiotic. To obtain estimates of the kinetic parameters of the intestinal absorption, a nonlinear regression fitting program (PCNONLIN, Statistical Consultants, 1986) incorporating iterative residual weighting was used. The t test was used to assess the statistical significance of the kinetic parameters studied in control and treatment conditions, using the parameter values as population means and assuming equality between the parameters as the HO hypothesis. The t values were obtained as the ratio [parameter value]/ [standard errorl.28.29 In order to ascertain a suitable model for the intestinal absorption of CFX and CFD, attention was given to the values of the error standard deviation (S) and the sum ofthe weighted squares residuals (SWRS) that are shown in Table I. Excretion-Biliary and urinary excretion of CFX and CFD was measured from the maximum antibiotic concentration found in the bile and urine samples (MAC, and MAC,, respectively) for each perfusion. These antibiotic concentrations were obtained by quantifying the corresponding peaks in the HPLC analysis of the samples. Table I-Comparison between Mlxed and Mlchaells-Menten Models using the Statistics “Error Standard Deviation” (S) and “Sum of Weighted Squared Residuals” (SWRS) for Cephalexln (CFX) and Cefadroxll (CFD) in Acute and Chronic Controls and Ethanol Treatments Treatment Acute Control CFD Control CFX Ethanol CFD Ethanol CFX Chronic Control CFD Control CFX Ethanol CFD Ethanol CFX
Michaelis-Menten
Mixed
S
SWRS
S
SWRS
0.2732 0.2695 0.1950 0.2300
1.1946 1.3487 0.3427 0.6876
0.2799 0.2664 0.1851 0.2209
1.1754 1.3077 0.3422 0.5859
0.2475 0.2349 0.2526 0.2661
0.7552 0.5050 0.6383 0.7080
0.2748 0.2247 0.2586 0.2582
0.5513 0.4967 0.6021 0.6002
To quantify the relation among the initial antibiotic concentration (IAC) administered to tbe rats and the maximum concentrations (MAC) excreted in each experiment, linear equations were obtained using MAC data as the independent variable and IAC data as the dependent variable. The slope of the straight lines was taken as a measurement of the excretory efficiency for each excretion route, antibiotic, and treatment. These values are summarized as Eb (excretoryefficiency for bile) and E, (excretoryefficiency for urine) in Table 11.
Results Absorption Kinetics-Tables I11 and IV summarize the results from the fitting of our data (see Figures 1and 2) to the Michaelis-Menten and Mixed models, respectively, as the nonlinear regression parameters and their corresponding standard errors for each experimental treatment. It is clear that SWRS values are consistently smaller for the Mixed model fitting than in the Michaelis-Menten model; but, that is to be expected because of the existence of one more parameter in the Mixed model. However, we can use S values to help discriminate the validity of either Mixed or MichaelisMenten models (Table I). Looking a t the S values for CFX, the Mixed fitting yields smaller values than the MichaelisMenten model, while the opposite can be said when looking at the corresponding values for CFD. There is a single exception, in the acute ethanol treatment for CFD, when the S value is smaller for the Mixed model. At the same time, the results of fitting our CFD data to the Mixed model gave negative values for the parameter K of simple diffusion (data not shown), except in the acute treatment with ethanol. These results confirm the idea that the only relevant parameters in that and K,, and, accordingly, Table IV shows fitting were V,, values for V,, and K , obtained from the Michaelis-Menten fitting for the acute and chronic controls and chronic ethanol treatment for CFD absorption. In the case of the Michaelis-Menten model (Table 1111, V, values for CFD and CFX decrease slightly but significantly both with acute and chronic ethanol treatment. The values of K , increase for the two antibiotics under the ethanol treatments, except in the acute ethanol treatment for CFX. On the contrary, following the Mixed model fitting (Table IV), different effects between the results from acute and chronic treatment do appear. First, K, values decrease as a consequence of the acute and chronic treatments for CFX and after the acute treatment decreases significantly in the for CFD. The parameter V,, case of both antibiotics after acute ethanol treatment, while chronic treatment only reveals statistical differences for CFX. The significance of the K parameter can only be assessed for CFX. This parameter reveals significant changes following only acute treatment with ethanol. Excretion-Table I1 lists excretory parameters E, (excretory efficiency for bile) and E, (excretory efficiency for urine),
as well as the corresponding standard error values and the number of animals (experimental points) used; Figures 3 and 4 show these data in a bar-chart format. Urinary excretion of cephalosporins predominates over biliary excretion. Our data show an increase of >lO-fold for urinary E, compared with E,. Thus, by supposing an intestinal segment 15 cm long and a body weight of 200 g, our data assume a maximum biliary concentration of CFX (after 120 min of perfusion) of 43.6 pg * mL-' and an urinary maximum concentration of the same antibiotic of 522 pg * mL-'. These data agree with those already obtained in human beings. For example, levels ranging from 2 t o 40 pg * mL-l of CFX in bile have been observed after an initial dose of 500 mg, while 99% of the dose can be recovered in urine after 5 h.4 Both biliary and urinary excretion efficiencies are better for CFX than for CFD. Kwan and Rogers2 reported an elimination of nearly 100% of a single dose of orally administered CFX (recovered in urine after 6-12 h), while the amount of CFD found in urine 24 h after its administration ranged between 85 and 93%.However, Nightingale et al.30 report a similar renal elimination for both antibiotics after 12 h, but the excretion rate is less for CFD when the time of study is limited to 3 h. The same behavior was previously found by Hartstein et al.1 in a study of the urinary excretion of CFX and CFD after a single po dose of 500 mg in humans. Although these authors use cumulative antibiotic excretion data and not maximum concentration data, it is possible that our data could fit their schema, given our limit of 2 h of intestinal perfusion.
Discussion Our results suggest a saturable mechanism for the intestinal absorption of CFX and CFD, as already pointed out by other authors.5~9~31 This study was made using the maximum antibiotic concentrations permitted by the solubility at the conditions described above (-50 mM), thus raising the level to several times the K, values already reported. It is to be noted that other authors,lO who used maximum antibiotic concentrations of
2 Y
0
10
20
30
40
50
0’ 0
60
W I (mM)
I
10
20
30
40
50
60
W D I (mM)
Figure 1-Relationship between the absorption rate (v) and the initial concentration of CFX perfused in each experimental treatment after fitting to the Mixed model. The plot for the control of the acute treatment acute; (A) chronic. is shown (solid line). Key: (0)control; (0)
Flgure 2-Relationship between the absorption rate (v) and the initial concentration of CFD perfused in each experimental treatment after fitting to the Mixed model. The plot for the control of the acute treatment is shown (solid line). Key: See Figure 1 .
agreement w i t h our data using the same fitting model (0.062 mM * min-’ for C F X and 0.078 mM * min-’ for CFD, assuming an intestinal segment 15 c m long and a 200-g rat). O u r results appear higher than those o f these authors, but no
centration higher than that theoretically reachable in the absence of osmotic effects. O n the other hand, at hypertonic concentrations, a decrease in the concentration w i l l be seen due to the intestinal secretion o f water. So, in the absence of osmotic corrections, V,,, and K , w i l l be lower than after the referred volume change correction. Castaner-Codes et al.31 report data on the intestinal transport o f C F X and CFD that i s also c o n c o r d a n t w i t h o u r s (171.35 and 316.13 g mL-’ min-’ as V,,, values f o r CFX and CFD, respec-
mention i s made of control of osmotic volume changes in their recirculatingperfusion experiments, which also lasted for 120 min. In fact, it i s clear t h a t for hypotonic antibiotic concentrations the intestinal w a l l w i l l absorb water from the perfusing solution, thus yielding a remnant antibiotic con514 I Journal of Pharmaceutical Sciences Vol. 80, No. 6, June 1991
-
-
ACUTE
CHRONIC T
- ctrl - -EtOH-
- ctrl - -EtOH-
Figure 3-Relationship
between the biliary excretory efficiencies (cm-' .g-') for cephalexin (a)and cefadroxil (N) in acute and chronic ethanol treatments, and their respective controls. The asterisk indicates significant differences at p < 0.01 (data are from Table I). ACUTE
CHRONIC
T
* 0
C Q
0
Ea
2
:
+
eV
x a
l
?
-
0 C
L
= ) " n
rameters, K , and V,,,, decrease when ethanol is administered in the acute form, while chronic ethanol treatment seems to have no clear effects on the absorption kinetics. The decrease in the activity of the carrier for CFX and CFD could be explained as a indirect consequence of the effect of ethanol on plasma membranes. So, Beck and Dinda32 suggest conformational changes induced by ethanol on disaccharidase of the brush border membrane to explain the decrease in activity of this enzyme upon acute treatment with ethanol. Uridine, choline, and chenodesoxicholate carriers could also be inactivated by changes in plasma membrane fluidity.6 Recent data33 show a direct effect of chronic ethanol administration on purified glycosidase activities in rat liver and brain, with changes in kinetic profile and activation energy. There is a need for more in depth studies in relation to the effect of ethanol on intestinal transport. Possible areas should include the evaluation of the direct toxic effect of ethanol on isolated hepatocytes. Ethanol Effects on Excretion-In drawing conclusions on the intestinal absorption for CFD and CFX, it should be noted that intestinal absorption, as measured by the V,,, parameter, is better for CFD than for CFX assuming a mixed kinetic model. This gives support to the hypothesis of different transfer mechanisms for these antibiotics in the gut and the liver or kidney, with both urinary and biliary excretion being possible first-order kinetic processes, which could explain the higher excretion values obtained for CFX than for CFD. Acute treatment with ethanol shows an increase in E, in the case of CFD, with decreases in E, for CFX (Table 11). Chronic treatment with ethanol does not produce significant differences. The effect of the acute treatment with ethanol on urinary excretion of CFX could be accounted for by induced enhanced urinary flow. Pohoreckyle reported an almost twofold increase in urine output by rats following an intragastric load of up to 2.5 g kg-' of ethanol, but chronically treated animals excreted a slightly lower volume of urine than did the controls. The almost normal urine volume for chronically treated rats could explain the lack of effect of chronic treatment on the urinary excretion of CFX and CFD. At present we have no explanation for the differential effects of the acute treatment with ethanol on the biliary excretion efficiencies of CFD and CFX. These could be attributed to the different chemical nature of the antibiotics, which could modify the general elimination path. Possibly, given the fact that both CFX and CFD are excreted in a form that is basically unmodified chemically, the presence of the additional parahydroxyl group in CFD compared with CFX could induce hepatic permeability changes which could explain their different excretory efficiencies.
- ctrl - -EtOH-
- ctrl - -EtOH-
Flgure +Relationship between. the urinary excretory efficiencies (cm-' g-') for cephalexin (a)and cefadroxil (N)in acute and chronic ethanol treatments, and their respective controls. The asterisk indicates significant differences at p < 0.01 (data are from Table I).
tively, equivalent to 0.047 and 0.083 mM min-', respectively) but these authors fit their data to a Michaelian-type pharmacokinetic model. From the above discussion, and assuming the hypothesis of the existence of a common carrier system for aminocephalosporins, as repeatedly reported,5,9.25 our results point to a better intestinal absorption for CFD than for CFX. Because of the higher hydrophobicity of CFX (the isobutilic-water partition coefficient for CFX is more than twice than for C F D 9 , this could be indicative of a hydrophilic microenvironment for the cephalosporin intestinal carrier. In keeping with this idea, we can consider the absence of proper values for the K parameter seen in this work (related to free diffusion transport) in the case of CFD (Table IV) as being due to the inadequacy of the fitting in the case of a negligible contribution of the first-order kinetic term of the Mixed model to the absorption of CFX; that is, this antibiotic would be absorbed solely by the carrier mechanism, avoiding purely passive membrane transport. Ethanol Effects on Absorption-A clear effect of ethanol is to raise the K parameter for CFX after acute treatment. This is also the only situation when the Mixed model can be applied to CFD transport. The increase in the passive diffusion component could be attributed to a higher permeability of the intestinal membranes induced by ethanol. Both kinetic pa-
References and Notes 1. Hartstein, A. I.; Patrick, K. E.; Jones, S. R.; Miller,M. J.;Bryant, R. E. Antimicrob. Agents Chemother. 1977,12,93-97. 2 . Kwan, K. C.; Rogers, J. D. In Antibiotics Containing the P-Lactam Structure ZZ Demain, A. L.; Solomon, N. A., Eds.; SpringerVerlag A G Berlin, 1983;pp 278-301. 3. Doluisio, J.T.; Billups, N. F.; Dittert, L. W.; Sugita, E. T.; Swintoskv. J. V. J . Pharm. Sci. 1969.58. 1196-1200. 4. Tsui,' A.; Miyamoto, E.; Kubo, 0.;Yamana, T. J . Pharm. Sci. 1979,68,812-816. 5. Nakashima, E.; Tsuji, A.; Kagatani, S.; Yamana, T. J.Phurm. Dyn. 1984,7,452-464. 6. Wilson, T. H.; Wiseman, G. J . Physiol. 1954,123, 116125. 7. Schurgers, N.; Bijdendijk, J.; Tukker, J. J.; Crommelin, D. J. A. J.Pharm. Scz. 1986,75,117-119. 8. L6 ez del Pino, V.; Hegazy, E.; Hauber, G.;Rernmer, H.; Sciwenk, M.Arch. Toxicol. 1983,Suppl. 6,322326. 9. Kimura, T.; Yamamoto, T.; Ishizuka, R.; Sezaki, H. Biochem. Pharmucol. 1985,34(1), 81-84. 10. Umeniwa, T.; Ogino, 0.; Miyazaki, K.; Arita, T. Chem. Pharm. Journal of Pharmaceutical SciencesI 515 Vol. 80, No. 6, June 7997
Bull. 1979,27,2177-2182. 11. Tsuji, A.; Yamana, T. In @Lactam Antibiotics; Mitsuhashi, S., Ed.: SDrinper-Verlae: New York. 1981:D 235. 12. Tsuji, A.;Gakashim;?, E.; Deguchi, Y.; Nishide, K.; Shimizu, T.; Horiuchi, S.; Ishikawa, K.; Yamana, T. J.Pharm. Sci. 1981,70, 1120-1128. 13. Prieto, J. G.; Barrio, J. P.; Alvarez, A. I.; Zapico, J. Reu. Farmacol. Clin. Exp. 1986,3,99-104. 14. Magnussen, M. P. Acta Phurmacol. Toxicol. 1968.26,130-144. 15. Hayton, W.L. J. Phurm. Sci. 1975,64,1450-1456. 16. Lane, E. A.; Guthrie, S.; Linnoila, M. Clin.Pharmacokinet. 1985, 10,228-247. 17. Larivikre, L.; CaillB, G.; Elie, R. Biopharm. Drug Dispos. 1986,7 , 201-210. 18. Pohoreckv. L. A. Alcohol 1985.2. 659-666. 19. Racusen,-L. C.; Krawitt, E. L. A m . J. Digest. Dis. 1977, 22, 915-920. 20. Ber e er H U. In Methods ofEnz m t i c Analysis; Bergmeyer, H. id.: Verlag Chemie WeinieidAcademic: New York, 1974;pp 1499-1502. 21. Ponz, F.; Ilundain, A.; Lluch, M. Rev. Esp. Fisiol. 1979, 35, 97-104. 22. Prieto, J. G.; Santos, L.; Barrio, J. P.; Alvarez, M. L. J. Pharm. Sci. 1987,76,596-598. 23. Seers,R.;Michels, H. InEmrgy MetabolismofFarm Animals; Ekern,
516 I Journal of Pharmaceutical Sciences Vol. 80, No. 6, June 1991
A.: Sundst~l.F.. Eds.: EAAP 29:Norwav. 1982 DD 135-137. 24. Saadeh, G. H.;‘Mu& M.; Nassar, C. F: Comp: Eiochem. Physwl. 1986,83A,539-541. 25. Miykaki, K.; Ohtani, K.; Umeniwa, K.; Arita, T. J.Phurm. Dyn. 1982.5. , - .555-563. ~~- - - 26. Tsuji, A.; Miyamoto, E.; Hashimoto, N.; Yamana, T. J.Phurm. Sci. 1978.67,1705-1711. 27. Tsuji, A.; Nakashima, E.; Kagami, I.; Nakashima, K.; Yamana, T. J. Pharm. Phurmucol. 1979,31,718-720. 28. Snedecor, G. W.; Cochran, W. G. Mdtodos estadtsticos, 1st Ed.; Compafifa Editorial Continental: S.A., MBxico D. F., 1975;p 131. 29. Sokal, R. R.; Rohlf, F. J. Biometrta. Princzpiosy mitodos estadlsticos en la inuestigacidn biol6gica; Edic. Edit. Blume: Barcelona, Spain, 1979. 30. Nightingale, C.H.; French, M. A.; Quintiliani, R. In pLactam Antibiotics; Mitsuhasi, S., Ed.; Springer-Verlag: New York, 1981; pp 259-267. 31. Castafier-Codes, A.; Merino, M.; Martin, A.; Obach, R.; PIBDelfina, J. M. Eur. Congress Biophurm. Phurmcokin. 1984,2, 170-179. 32. Beck, I. T.; Dinda, P. K. A m . J. Digest. Dis. 1981,26,817-838. 33. Gil Martin, E.; FernBndez-Briera, A.; Calvo Ferntindez, P. Biochem. Pharmucol. 1990,40,925-982. ~~
~~
