Downloaded from
`
`jpet.aspetjournals.org
`
` at ASPET Journals on May 16, 2016
`
`0022-~/86/2332-0277102.00/0
`THE JoURNAL or PHARMACOLOGY AND ExPBJUN&NTAL TH&RAPEtmcs
`~t C 1985 by The Amerieon Society for Pharmoc:oloo and Esperimental Tberopeutia
`
`Vol. 233, No.2
`PrinUd in U.S.A.
`
`Reduced Extraction of /-Propranolol by Perfused Rat Liver in the
`Presence of Uremic Blood1
`
`NORIKAZU TERAQ2 and DANNY D. SHEN3
`Department of Pharmaceutics, Schoof of Pharmacy, State University of New York at Buffalo, Amherst, New York
`Accepted for publication January 28, 1985
`
`ABSTRACT
`Previous in vivo studies have shown that the presystemic clear(cid:173)
`ance of p.o.-administered /eva-isomer of propranolol is inhibited
`in rats with uranyl nitrate-induced acute renal failure. A series of
`steady-state single-pass rat liver perfusion studies were per(cid:173)
`formed to explore the probable mechanism of the observed
`metabolic inhibition. When livers from normal rats were perfused
`with blood perfusate prepared from normal donor animals, a high
`extraction ratio (Eh) of 0.974 ± 0.005 (mean ± S.D.) was ob(cid:173)
`served at an influent /-propranolol concentration of 400 ngfml,
`i.e., only 2.6% of drug entering the liver escaped single-pass
`extraction. The extraction of /-propranolol was significantly lower
`(i.e., Eh= 0.906 ± 0.017) when livers isolated from uranyl nitrate(cid:173)
`induced renal failure rats were perfused with uremic blood; such
`that there was an approximate 3-fold increase in the amount of
`drug escaping single-pass extraction (i.e., from 2.6 to 9.4%).
`
`This difference in hepatic extraction is quantitatively consistent
`with the increase in p.o. systemic availability of /-propranolol
`observed in our previous in vivo study with the uranyl nitrate(cid:173)
`induced renal failure rat model. When livers from normal rats
`were cross-perfused with uremic blood, extraction of /-propran(cid:173)
`olol was depressed to almost the same level (i.e. , Eh = 0.927 ±
`0.009) as when livers from renal failure animals were perfused
`with uremic blood. In contrast. livers from renal failure rats cross(cid:173)
`perfused with normal blood exhibited comparable extraction for
`/-propranolol (Eh = 0.970 ± 0.010) as normal livers perfused with
`normal blood. These results indicate that the diminution in pre(cid:173)
`systemic hepatic extraction of /-propranolol in the uranyl nitrate
`renal failure rat model is due to the presence of an inhibitory
`factor in the uremic blood. No apparent alterations in the intrinsic
`activities of the hepatic transport and/or drug-metabolizing en(cid:173)
`zyme systems were observed.
`
`There is mounting evidence that alterations in the biotrans(cid:173)
`formation of drugs can occur in renal failure (Reidenberg,
`1977). Studies in our laboratory with the uranyl nitrate-induced
`acute renal failure rat model demonstrated that the presystemic
`hepatic clearance of the levorotatory or S-(-)-isomer of pro(cid:173)
`pranolol is inhibited during uremia after both single and repet(cid:173)
`itive p.o. administration of the beta blocker (Terao and Shen,
`1983, 1984). Similar results with racemic propranolol were
`reported recently by Katayama et al. (1984) using the same
`nephrotoxic acute renal failure rat model. These animal data
`are consistent with the results of two early clinical studies
`(Lowenthal et al., 1974; Bianchetti et al., 1976), which showed
`a 3- to 7-fold higher blood or plasma concentration of racemic
`propranolol after a single p.o. administration of the drug in
`end-stage renal failure patients as compared to patients or
`volunteers with normal renal function. However, two more
`recent studies (Stone and Walle, 1980; Wood et al., 1980)
`showed either a small or negligible increase in circulating
`propranolol concentrations in chronic renal failure patients
`
`Received for publication July 6, 1984.
`1 This work was oupported by U.S. Public Health Service Grants HL-25797
`and GM-20852.
`'Preoent address: Chupi Pharmaceutical Co., Ltd., 1-9 Kyobashi 2-Chome,
`Cbuo-ku, Tokyo, 104, Japan.
`• Preoent addreos: Department of Pharmaceutics, BG-20, School of Pharmacy,
`University of Washington, Seattle, WA 98195.
`
`receiving regular propranolol therapy. As was pointed out in
`our earlier reports (Terao and Shen, 1983, 1984), the apparent
`conflicts may be attributed to differences in experimental de(cid:173)
`sign and patient selection between studies. Furthermore, these
`fmdings suggest that the clinical factors affecting the metabo(cid:173)
`lism of propranolol in uremia are complex and multifactorial
`in nature. Hence, there is a need for experimental studies in
`animal models to elucidate the probable mechanism(s) respon(cid:173)
`sible for the decreased first-pass clearance of p.o. administered
`propranolol during uremia, which in turn may provide some
`insights as to when and how the metabolism and disposition of
`propranolol could be affected in chronic renal failure patients.
`In the present study, an in situ perfused rat liver preparation
`was used to assess the extraction of 1-propranolol by livers
`isolated from normal and renal failure animals. These experi(cid:173)
`ments were designed specifically to explore two plausible mech(cid:173)
`anisms for the apparent inhibitory effect of renal failure on the
`hepatic extraction of 1-propranolol: 1) the presence of endoge(cid:173)
`nous inhibitors in the uremic blood or 2) an intrinsic alteration
`in the hepatic uptake and/or metabolic processes during renal
`failure. The two causal mechanisms can readily be distin(cid:173)
`guished by performing a series of "cross" perfusion experiments,
`which entail the perfusion of livers isolated from normal ani(cid:173)
`mals with blood perfusate prepared from uremic donor animals
`and the counter-perfusion of livers isolated from renal failure
`animals with blood perfusate prepared from normal donor.
`277
`
`Boehringer Ex. 2021
`Mylan v. Boehringer Ingelheim
`IPR2016-01565
`Page 1
`
`

`
`Downloaded from
`
`jpet.aspetjournals.org
`
` at ASPET Journals on May 16, 2016
`
`Vol.233
`
`Saqical proceclure. The liver donor animal was anesthetized with
`ether and placed on a movable operating platform. An abdominal
`midline incision was made to expoae the region of the portlJ hepatiB
`without diaturbing the liver. Tbe bile dt~et was located and cannulated
`with a PE-10 tubing (Clay Adams, Parsippany, NJ) filled with heparin(cid:173)
`ized saline (100 U/ml).
`A l0011e ligeture was placed around the inferior vena cava proximal
`to the merger of the right renal vein. A bolus doae of heparin (300 U)
`in isotonic aaline was injected into the inferior vena cava at a Bite 5 to
`10 mm distal to the l0011eligeture. Tbe l0011eligeture was left in place
`until the conclusion of the portal and hepatic vein cannulation proce(cid:173)
`dure.
`The hepato-portal vein was expoaed and the pyloric branch of the
`vein was tied oft'. A polyethylene cannula (fabricated from a 1.5" length
`ofPE-160 tubing), filled with heperinizedsalins (100 U/ml) andcloaed
`in the outflow end with a metal plug (to prevent air from intruding
`into the cannula), was inserted into the portal vein. Tbe tip of the
`cannula was placed about 5 mm away from tbe first bifurcation of the
`portal vein before ita entry into the various lobes of the liver.
`The thorax was cut open e:r:poeing the right atrium. The thoracic
`inferior vena cava was isolated and a heparinized polyethylene cannula
`(PE-160) was inserted into tbe cava quickly. The tip of the caval
`cannula was advanced to tbe junction of the hepatic vein. Tbe preplaced
`ligeture around the inferior vena cava was then tightened. As a final
`step, the hepatic artery was ligeted to prevent fluid leakage during
`perfusion. The entire 81Ugical procedure was usually completed in 5 to
`7min.
`PerfWiate preparation. The perfuaate medium for both preperfu(cid:173)
`aion and blood perfusion were prepared fresh daily. Blood-donor ani(cid:173)
`mals (between 13 to 15 normal or renal failure rata per perfusion
`experiment) were anesthetized with ether and blood was withdrawn
`from the abdominal aorta into a heparinized glass syringe. Tbe poolad
`heparinized blood (at a final heparin concentration of 10 U/ml) was
`poured through a 3" x 3" 81Ugical gauze sponge, into a 250-ml volu(cid:173)
`matric glass cylinder to remove any large blood clots that may be
`present.
`Krebe-bicarbonate buffer solution (millimolar) (NaCI, 118.5; Na(cid:173)
`HCO., 25.0; KCl, 4.74; CaCla, 2.54; Mg804, 1.04; and KH,.PO., 1.19;
`m
`
`Fig. 1. Schematic ~ of the
`rat lver perfusion system. The portal and
`caval COI118Cti0118 of the lver to the lnc:l(cid:173)
`vidual perfusion circuits -
`shown for
`each phase of the experiment: warnHIP
`(1), praperfusion (II) lnl blood perfusion
`(Ill). Key: P, JUI'IP; F, liter; T, bubble trap.
`
`278
`
`Methods
`
`Animals. Male SpriiiiiJI!·Dawley rata weighing between 300 to 350 g
`were uaed in all of the perfuaion experiments. The animals were allowed
`food and water ad libitum until the time of sacrifice. Following the
`procedure establiahed by Giacomini et al. (1981), renal failure was
`induced by a single i.v. injection of uranyl nitrate (5 mgjkg) 5 days
`before the perfusion experiments. Control animals received an equal
`volume injection of saline. To aaseas the degree of uremia, plasma uree
`nitrogen concentration was determined by a colorimetric procedure
`using a commercial assay kit (Sigma Chemical Co., St. Louis, MO).
`Perfualon apparatu. A schematic representation of the liver
`perfuaion apparatus is shown in figure 1. The whole assembly of the
`perfuaion apparatus, except for two variable speed roller pumps (Maa(cid:173)
`terflex, models K-7565 and K-7553, Cole-Parmer Instrument Co., Chi(cid:173)
`cago, IL), was houaed in a temperature-controlled Plexiglas cabinet
`(MRA Corp., Clearwater, FL). The temperature inside the cabinet was
`maintained at 37"C.
`The apparatus was made up of two independent circuits of perfusion.
`One circuit, referred to as the preperfuaion circuit, was uaed in the
`initial part of the experiment to clear the liver of ali reeidual blood by
`perfuaing with a blank buffer medium. The preperfusion step also
`allowed tbe stabilization of liver temperature at 37"C. The other circuit
`was uaed for the actual perfuaion with drug containing rat blood
`perfusate.
`The preperfuaion circuit consisted of a perfusate reservoir, a Lucite
`filter equipped with a disc of rme-meshed (120-140 threads per inch)
`100% Japanese silk and an air bubble trap. The various parts of the
`assembly were connected with silicone tubing. The buffer medium was
`aerated continuously with a prewarmed humidified gas mixture of95%
`oxygen and 5% carbon dioxide (Carbogen, Airco, Inc., Murray Hill,
`NJ).
`In the blood perfuaion circuit, the perfusate was oxygenated by
`passage through a 20-ft coil of Silastic tubing (0.058" inside diameter,
`0.077" outside diameter; Dow Coming Co., Midland, MI), which was
`enclosed in a glass chamber flushed continuously with prewarmed
`humidified carbogen gas. The blood perfusate was stirred continuously
`and gently in the reservoir by a magnetic stirrer.
`
`I :1 2
`
`I
`
`5 •
`
`t
`
`Blood Reurvair
`
`n
`
`Liver
`
`~
`I Waate
`
`Boehringer Ex. 2021
`Mylan v. Boehringer Ingelheim
`IPR2016-01565
`Page 2
`
`

`
`Downloaded from
`
`jpet.aspetjournals.org
`
` at ASPET Journals on May 16, 2016
`
`1985
`
`pH= 7.4) was equilibrated with Carbogen gas for 10 min. Bovine serum
`albumin (Fraction V, Sigma Chemical Co.) and glucose were dissolved
`in the buffer to yield concentrations of 5 and 0.1 %, respectively. Two
`hundred milliliter of the buffer solution (i.e., preperfusion medium)
`was placed in the reservoir of the preperfusion circuit where it was
`warmed to 37'C and kept oxygenated.
`Another 100 mi of the Krebs-bicarbonate buffer solution was added
`to 200 ml of the pooled heparinized rat blood to make a one-third
`diluted whole blood perfusate. 1-Propranolol hydrochloride (Ayerst
`Laboratories, New York, NY) was added to the final blood perfusate
`to a concentration of 500 ng/ml. Sodium taurocholate (Sigma Chemical
`Co.) dissolved in saline was also added to the blood perfusate to obtain
`a concentration of 30 ,..M. The perfusion medium (300 ml) was mixed
`thoroughly and placed in the perfusate reservoir. The blood perfusate
`was then recirculated through all the components of the perfusion
`circuit (without the liver) at a constant flow rate of 10 ml/min for 30
`to 40 min (see schema I in f~g. 1). The equilibration period was necessary
`to allow the blood perfusate to warm up to 37'C and be fully equilibrated
`with carbogen gas. Preliminary experiments also indicated that there
`was an unavoidable loss of drug from the perfusate ( -20%) due to
`adsorption onto glassware& and silicone tubing during the initial period
`of recirculation.
`Liver perfusion. I) Preperfusion. After the cannulation procedure
`the animal was transferred into the perfusion cabinet. Preperfusion
`with drug-free buffer medium was initiated immediately by connecting
`the portal cannula to the outflow port (No. 3, schema II in fig. 1) in
`the preperfusion circuit. The effluent from the liver was discarded by
`connecting the caval cannula to the drainage port (No. 5). The liver
`was preperfused for 10 min at a flow rate of 15 ml/min.
`During the preperfusion the position of the liver and the inlet and
`outlet tubes were adjusted so that there was no obstruction to the
`perfusate flow. The surface of the liver was irrigated with 5 to 10 ml of
`warm saline. A saline-moistened gauze sponge was gently placed over
`the liver to avoid dehydration of the organ. The temperature of the
`liver was monitored by placing a thermistor probe (Model 43-TD,
`Yellow Springs Instrument Co., Yellow Springs, OH) between two
`lobes of the liver. The position of the bile duct was edjusted to obtain
`a constant flow of bile.
`2) Single-po8B extraction study. The steady-state extraction of /(cid:173)
`propranolol was measured during single-pass perfusion of livers from
`control and renal failure animals with both normal and uremic blood.
`The four following perfusion arrangements were performed: i) normal
`liver perfused with normal blood; (ii) uremic liver perfused with uremic
`blood; (iii) normal liver perfused with uremic blood; and (iv) uremic
`liver perfused with normal blood.
`At the end of the preperfusion period, the portal cannula was
`disconnected from the preperfusion circuit and connected quickly to
`the outflow port (No. 1) of the blood perfusion circuit. Care was taken
`not to disturb the position of the liver and the caval cannula connection.
`The perfusate flow rate through the liver was set at 10 ml/min.
`The effluent (venous) blood perfusate was withdrawn from an in(cid:173)
`line septum-type sampling port situated at the caval cannula outlet, at
`2, 5, 8, 12, 15, 18 and 21 min after the commencement of blood
`perfusion. The influent (arterial) blood perfusate was collected from a
`sampling port placed at tbe inlet to the portal cannula, at 0, 7, 13 and
`22 min after the start of blood perfusion. Bile was collected during the
`steady-state interval, i.e., 14 to 24 min, in a tared polypropylene tube
`(Walter Sarstedt, Inc., Princeton, NJ). Bile volume was estimated
`gravimetrically, assuming a density of unity. At the completion of the
`perfusion experiment, the liver was excised, rinsed with warm saline,
`blotted dry and weighed.
`The perfusate blood samples were split in halves. One portion was
`centrifuged to obtain the "plasma" fraction. The perfusate blood and
`plasma along with the bile and liver specimens were stored frozen at
`-20'C pending drug metabolite assay. Only the results on the perfusate
`blood and plasma will be presented in this report.
`The concentration of /-propranolol in perfusate blood and plasma
`was measured by a high-performance liquid chromatographic method
`
`Metabolic Inhibitors in Uremic Rat Blood
`
`279
`
`described earlier (Terao and Shen, 1982). The assay had a coefficient
`of variation of less than 4% (at the concentration of 10 ng/ml) and a
`sensitivity limit of0.5 to 1 ngfml (with a sample volume of0.1 ml).
`Protein binding of !-propranolol in arterial perfusate plasma col(cid:173)
`lected at 13 min of perfusion and in venous perfusate plasma at 18 min
`were determined by equilibrium dialysis according to a procedure
`described previously (Terao and Shen, 1983).
`3) Viability tJBBe88ment. Livers prepared from control and renal
`failure animals were compared with respect to their viability during
`perfusion with normal rat blood perfusate. In these experiments per(cid:173)
`fusion was performed in the recirculating mode to minimize the volume
`of rat blood required. The preperfusion was the same as described
`earlier. Sixty milliliters of the blood perfusate was recirculated (by
`connecting the caval cannula to port No. 2 shown in fig. 1) at a flow
`rate of 10 ml/min for a period of 3 hr. Bile salt was replenished by
`infusing 18 ,..mol of sodium taurocholate per hr into the reservoir.
`Bile flow rate was monitored throughout the experiment. Arterial
`(portal) and venous (caval) blood P., and P.., and arterial blood pH
`were determined before and after 1, 2 and 3 hr of perfusion (IL Micro
`13/123 pH/Blood Gas Analyzer, Instrumentation Laboratory, Lexing(cid:173)
`ton, MA). Perfusate samples were also collected at hourly intervala for
`the following biochemical measurement& in perfusate plasma: glucose,
`albumin, total protein, urea nitrogen, creatinine and glutamic pyruvic
`transaminase. All of the biochemical measurements were performed on
`a centrifugal chemistry analyzer (Centrifichem Analyzer 400, Union
`Carbide Corporation, Clinical Diagnostics, Rye, NY).
`Protein binding and blood to pluma coneentration ratio. In
`vitro binding of /-propranolol to proteins in the rat perfusate plasma
`and drug partitioning between blood cells and perfusate plasma were
`determined. Blank blood perfusate was prepared by diluting freshly
`collected rat blood (at a hematocrit of 0.45) from normal animals with
`one-third volume of Krebs-bicarbonate buffer. Earlier work (Terao and
`Shen, 1983) has shown that uremia does not affect serum protein
`binding and blood to serum partitioning of /-propranolol. Therefore,
`all subsequent experiments were conducted with perfusate prepared
`with blood from normal donor rats. To measure the partitioning of
`drug between red blood cells and perfusate plasma, a tracer amount (3
`x 10' dpm) of 1-["H)propranolol (specific activity, 22.4 Ci/mmol; New
`England Nuclear, Boston, MAl and varying amounts of cold carrier
`were added to 3 ml of blank blood perfusate to achieve a wide range of
`pre-equilibrium perfusate blood concentrations from 8 ngfml to 300
`,..gfml. The blood samples were gently mixed at 37'C for 60 min.
`Preliminary experiments showed that partition equilibrium is reached
`within an hour of incubation at 37'C. At the end of incubation, an
`aliquot of the blood was removed and the remaining portion was
`centrifuged to obtain the plasma fraction. Concentration of radioactive
`/-propranolol in the perfusate blood and plasma (10-,..J aliquot&) were
`determined by direct liquid scintillation counting in premixed commer(cid:173)
`cial scintitlor (Aquosol, New England Nuclear). Quenching was cor(cid:173)
`rected by external standard channel ratio method. For perfusate plasma
`protein binding determinations, blank perfusate plasma was dialyzed
`against 0.134 M phosphate buffer (pH 7.4), spiked with varying con(cid:173)
`centrations of cold /-propranolol (ranging from 10 ng/ml to 500 ,..gfml)
`and a tracer quantity of 1-["H)propranolol (10' dpmfml), for a period
`of 6 hr at 37'C. After dialysis, aliquot& (50 to 100 ,..!) of the perfusate
`plasma and buffer were assayed for total radioactivity.
`Pharmacokinetic calcula&iona. The single-pass hepatic extrac(cid:173)
`tion ratio (E•l of /-propranolol was calculated from the steady-state
`arterial (C.) and venous (C.) perfusate blood concentration as follows:
`E.= c.-c.
`c.
`All of the measured arterial concentrations were pooled to provide an
`average steady-state estimate of C .. The steady-state venous perfusate
`blood concentration was calculated by averaging the measurements at
`15, 18 and 21 min of perfusion.
`Assuming that the liver behaves like a well-stirred compartment
`(i.e., with drug in the venous blood being in instantaneous equilibrium
`
`(1)
`
`Boehringer Ex. 2021
`Mylan v. Boehringer Ingelheim
`IPR2016-01565
`Page 3
`
`

`
`Downloaded from
`
`jpet.aspetjournals.org
`
` at ASPET Journals on May 16, 2016
`
`Vol.233
`
`ever, a trend toward a higher venous P .. with time, indicating
`a very small (<4%) but gradual decline in oxygen consumption
`over the 3 hr of perfusion.
`The plasma biochemical measurements from the viability
`experiment are presented in table 1. There was no remarkable
`difference in any of the biochemical indices between livers
`prepared from normal control and renal failure animals. Plasma
`glucose showed a gradual decline during recirculating perfusion
`reflecting a steady metabolic consumption. On the other hand,
`albumin and total protein concentrations increased (-20%)
`over the 3-hr period, most likely due to continuous net produc(cid:173)
`tion of plasma proteins by the liver. As expected, serum creat·
`inine and urea nitrogen concentrations (of blood samples taken
`before sacrifice or surgery) were elevated markedly in the renal
`failure donor animals as compared to the normal donor animals.
`However, normal concentrations of creatinine and urea nitro(cid:173)
`gen were observed during perfusion of uremic liver with blood
`perfusate prepared from normal rats, indicating that preperfu(cid:173)
`sion with the buffer medium was effective in removing these
`uremic substances from the hepatic vasculature and tissue. The
`perfusate plasma transaminase activities remained normal
`throughout the experiment, suggesting that hepatocellular in(cid:173)
`tegrity was maintained at all times.
`For both control and uremic livers, a constant bile flow was
`maintained for the first 2 hr (control, 282 ± 48 ,d/30 min;
`uremic, 286 ± 62 ,.1/30 min). Thereafter, bile flow decreased by
`approximately 15% over the last hour of perfusion.
`Overall, the results of the viability experiment indicate no
`apparent difference in the physiologic integrity of the perfused
`livers prepared from normal control and renal failure animals
`under the present experimental conditions.
`Protein binding and partitioning in blood perfwlate.
`The relationship of percentage of bound and blood to plasma
`concentration ratio of !·propranolol to the total drug concen(cid:173)
`tration in perfusate plasma is graphically displayed in figure 3.
`Both the percentage of bound and perfusate blood to plasma
`concentration ratios were relatively constant at total (i.e.,
`bound and unbound) perfusate plasma concentrations below 1
`,.g/ml. At higher concentrations, a gradual decrease in the
`perfusate plasma protein binding of !-propranolol was observed.
`There was also a concurrent increase in the blood to plasma
`ratio, which is likely due to enhanced partitioning of !-propran(cid:173)
`olol into the cellular fraction as a result of the increase in the
`unbound fraction of drug in perfusate plasma. In order to avoid
`the complexities of concentration-dependent changes in the
`distribution of !-propranolol within the blood perfusate during
`passage acroas the hepatic sinusoid, the arterial perfusate drug
`concentration was kept below 1 ,.g/ml. Considerations based
`on these observations and the detection limitation of drug assay
`in the perfusate effiuent led to the decision of setting the
`influent !-propranolol concentration at 500 ng/ml for the sub·
`sequent single-pass extraction study.
`Single-p888 perfusion experiments. The mean results of
`the four sets of perfusion experiments (i.e., normal liver with
`normal blood, uremic liver with uremic blood, normal liver with
`uremic blood and uremic liver with normal blood) are presented
`in table 2. Four perfusion runs were performed per experiment.
`The time course of mean arterial and venous blood perfusate
`concentrations for each set of perfusion experiment is presented
`in figure 4.
`There were no significant differences in the mean body
`weight (of the donor animals), liver weight, hematocrit and
`
`280
`
`Ten10 8nd Shen
`
`with that in the liver), it has been shown that the hepatic extraction
`ratio can be mathematically related to hepatic blood flow rate ( Q.J and
`intrinsic clearance (CI;.,) (Rowland et al., 1973; Wilkinson and Shand,
`1975). The latter parameter expresses the capability of the liver to
`remove drug from the perfusate in the absence of any blood flow rate
`limitation and is governed by the rate of uptake and/or metabolic
`transformation of !-propranolol by the hepatocyte&. Assuming that the
`kinetics of intrahepatic distribution and metabolic removal are linear
`processes, this relationship is expressed by the following equation:
`
`(2)
`
`Upon rearrangement of the equation, an expression for Cl;., can be
`obtained:
`
`Cl;., = E.-Q.
`1- E.
`
`(3)
`
`Because the perfusion flow rate is controlled in the perfusion prepara(cid:173)
`tion, the intrinsic metabolic clearance of !-propranolol in the perfused
`rat liver can be estimated from the extraction ratio.
`
`Results
`
`Viability of perfuaed uremic liver. The resulta of the
`perfusate blood pH, P .. and P.., measurements from the via(cid:173)
`bility experiment are presented in figure 2. No significant
`differences in these physiological parameters were observed
`between the perfused livers from the normal control and renal
`failure animals. Both the perfusate blood pH and the arterial
`and venous P.., remained constant during the entire period of
`perfusion. Adequate oxygenation of the blood perfusate was
`achieved as evident by a steady and high arterial oxygen tension
`(@ 280 mm Hg). One indicator of the viability of the perfused
`liver preparation was an effective oxygen utilization as reflected
`in the large arterial-venous difference in P ... There was, how-
`
`... 50~0
`
`%
`e
`
`0
`
`•
`
`e. t i
`
`N
`0
`u
`IL
`
`30
`
`0
`
`60
`
`0
`•
`g
`
`i v
`g
`
`A
`
`120
`
`180
`
`...
`280}0
`:J: 260
`E
`E
`
`0 ~r
`
`IL
`
`40
`
`l!l
`0
`
`% 't 8
`
`Q
`
`Q
`
`Q A
`
`0
`
`Iii
`
`Iii v
`
`60
`
`120
`
`180
`
`•
`
`0
`
`8
`
`120
`
`180
`
`...
`a 7.3
`"C
`~ 7.2
`c
`
`0
`
`60
`
`Time of Perfusion, min.
`Fig. 2. Mean blood gas/pH measurements from the viability experiment
`for normal control liver (0) and uremic liver <•>· A and V represent arterial
`and venous data, respectively (n = 3).
`
`Boehringer Ex. 2021
`Mylan v. Boehringer Ingelheim
`IPR2016-01565
`Page 4
`
`

`
`1985
`
`Metabolic Inhibitors In Uremic R8181ood
`
`281
`
`TABLE 1
`Blocllemlclll mMIIUrementa on perfuute plauna abtlllned from the viHIIIty experiment
`AI Ivers wall perfused with bklod perfusate prepared from normal animals.
`
`Downloaded from
`
`jpet.aspetjournals.org
`
` at ASPET Journals on May 16, 2016
`
`Plrlusion Tme
`
`om
`275±66
`286±25
`3.8±0.4
`3.7 ± 0.2
`4.5±0.2
`4.4 ±0.2
`10±3
`10 ± 1
`0.8± 0.2
`0.7 ± 0.1
`18 ± 1
`17 ±6
`
`&Om
`231 ±40
`233± 56
`3.8±0.2
`3.7 ± 0.1
`4.7±0.2
`4.7± 0.1
`11 ± 2
`15±2
`0.8 ± 0.1
`0.9±0.1
`19±3
`19 ± 10
`
`120m
`195±33
`193±36
`4.0± 0.2
`3.8 ± 0.1
`4.9±0.2
`4.9±0.2
`14±3
`19±2
`0.9±0.1
`0.9± 0.2
`23± 1
`21 ± 7
`
`1aom
`196 ± 29
`185±30
`4.0 ± 0.2
`3.9 ± 0.1
`5.3± 0.2
`5.0 ± 0.1
`17±3
`23±2
`0.9 ± 0.2
`0.9± 0.2
`27±6
`25±3
`
`0.043). In addition, the unbound fraction of 1-propranolol was
`consistently higher in the arterial perfusate plasma than the
`venous plasma perfusate by about 5%. Considering the rather
`large arterial-venous drug concentration gradient, these differ(cid:173)
`ences can probably be attributed to a small degree of concen(cid:173)
`tration dependency in the protein binding and blood cell to
`plasma partitioning of l-propranolol in the blood perfusate.
`More critical is the observation that there was no change in
`blood to plasma concentration ratio with time. Therefore, all
`subsequent di8CU88ion will refer only to the perfusate blood
`data.
`In all cases, steady-state concentration of 1-propranolol in
`the venous perfusate blood, i.e. steady-state extraction, was
`reached within 15 min after the start of blood perfusion (see
`fig. 4). A small peak in venous concentration was observed
`consistently during the flrBt 5 min of blood perfusion. The
`reason for this fluctuation in effluent perfusate concentration
`during the equilibration period is not clear.
`Steady-state hepatic extraction ratio and estimates of intrin(cid:173)
`sic hepatic clearance of 1-propranolol for the four sets of per(cid:173)
`fusion experiments are summarized in table 3. 1-Propranolol
`was removed effectively from normal blood perfusate during a
`single passage through the liver isolated from normal control
`animals. The high hepatic extraction ratio (0.974 ± 0.005) is
`consistent with data from our earlier in vivo studies (Terao and
`Shen, 1983) which showed that the average systemic availabil(cid:173)
`ity of 1-propranolol after single p.o. administration (6 mg/kg)
`is only 7% in the normal rat. A much lower extraction ratio
`(0.906 ± 0.017) was observed in the uremic liver perfused with
`uremic blood, such that the mean venous to arterial concentra(cid:173)
`tion ratio (i.e., fraction of drug escaping hepatic extraction) is
`3 to 4 times higher in the uremic liver perfused with uremic
`blood as compared to normal liver perfused with normal blood
`(0.094 vs. 0.026). The magnitude of increase in the effluent
`drug concentration is similar to the relative difference in the
`in vivo p.o. systemic availability of l-propranolol between nor(cid:173)
`mal and renal failure animals (7 vs. 18%) reported earlier
`(Terao and Shen, 1983). The data from these homologous
`perfusion experiments confmn our hypothesis that the change
`in systemic availability of I-propranolol induced by renal failure
`is attributed to an inhibition of first-pass hepatic clearance.
`
`Measlnments
`
`Donor
`
`S..h"
`
`Glucose, mg/dl
`
`Albumin, gfdl
`
`Total protein, g/dl
`
`lkea nitrogen, mg/dl
`
`Creatinine, mg/dl
`
`GPT."IU/ml
`
`110 ± 40°
`NC"
`RF"
`95±23
`3.2 ± 0.6
`NC
`RF
`3.1 ±0.5
`5.8± 1.2
`NC
`RF
`5.5± 1.0
`16±4
`NC
`RF
`155 ± 10
`1.1±0.3
`NC
`RF
`4.9 ± 0.6
`27±7
`NC
`RF
`27±5
`• Measlnments on serum obt8ned from livw-donor animals before surgery.
`• NC, normal control rats. n • 3.
`"MeM±S.D.
`• RF, AJIIIII faikn rats, n • 3.
`• GPT. glu1amic pyruvic transarT*Iase.
`
`~ 100
`
`a:
`
`.. E • .. 80
`I! .. • i 60
`• Q.
`. !: ... 40
`c: " 0 m
`
`1/! 20
`
`100
`10
`.1
`.01
`Totol Perfuaote Ploama Conc.,l'-g/ml
`
`2.0
`
`0
`
`.!
`.!:!
`;;
`a:
`1.5 u c:
`(.) .. E
`1.0 • .. a: ... 0
`
`0
`0.5 iii
`1000
`
`Fig. 3. Variations in plasma protein binding and blood-plasma partitioning
`of /-propranolol In rat blood perfusate madlum as a function of plasma
`perfusate drug concentration. Cone. concentration.
`
`steady-state bile flow between the four groups of experiments.
`A relatively uniform degree of uremia was induced in all renal
`failure donor animals as indicated by the consistency in plasma
`urea concentrations (Ca. 140 to 160 mg/dl) measured before
`the time of sacrifice or surgery.
`The arterial blood perfusate concentration measured during
`the perfusion was 10 to 25% less than the anticipated concen(cid:173)
`tration of 500 ng/ml. This apparent loss of drug is probably
`due to adsorption of 1-propranolol onto surfaces of glasswares
`and silicone tubing. Although the actual arterial drug concen(cid:173)
`tration varied from run to run (range 301 to 442 ng/ml), the
`input concentration remained constant (within ±5%) during
`each course of perfusion. The small between-run variation in
`the influent perfusate concentrations is not expected to affect
`the extraction ratio of 1-propranolol.
`As expected, binding to perfusate plasma proteins and the
`blood to plasma concentration ratio of 1-propranolol did not
`differ between the four perfusion arrangements. This confirms
`our previous in vivo finding that uremia has no effect on the
`distribution of I-propranolol within the various blood elements.
`I-Propranolol concentration in the perfusate blood was always
`slightly lower than in the perfusate plasma. The blood to plasma
`concentration ratio was slightly higher for the arterial as com(cid:173)
`pared to the venous perfusate blood (0.912 ± 0.048 vs. 0.903 ±
`
`Boehringer Ex. 2021
`Mylan v. Boehringer Ingelheim
`IPR2016-01565
`Page 5
`
`

`
`Downloaded from
`
`jpet.aspetjournals.org
`
` at ASPET Journals on May 16, 2016
`
`Ttlnlo and 8hen
`
`282
`
`TABLE2
`
`Vol. 233
`
`327 ± 10°
`10.5 ± 1.1
`
`13.0±3.6
`7.6± 1.6
`113 ± 29
`
`372 ± 17
`397 ±34
`
`9.8± 1.9
`10.6 ±