PROD RUG STRATEGIES FOR BYPASSING THE

FIRST·PASS METABOLISM OF PROPRANOLOL

by

Wei Wei Chu

A thesis submitted to the faculty of

The University of Utah

in partial fulfillment of the requirements for the degree of

Master of Science

Department of Pharmaceutics

The University of Utah

December 1987

Copyright © Wei Wei Chu 1987

All Rights Reserved

THE UNIVERSITY OF UTAH GRADUATE SCHOOL

SUPERVISORY COMMITTEE APPROVAL

of a thesis submitted by

Wei Wei Chu

This thesis has been read by each member of the following supervisory committee and by majority vote has been found to be satisfactory.

/1/,3/'67

THE UNIVERSITY OF UTAH GRADUATE SCHOOL

FINAL READING APPROVAL

To the Graduate Council of The University of Utah:

I have read the thesis of Wei Wei Chu

in its

final form and have found that (I) its format, citations, and bibliographic style are consistent and acceptable: (2) its illustrative materials including figures, tables, and charts are in place; and (3) the final manuscript is satisfactory to the Supervisory

Committee and is ready for submission to the Graduate School.

Dale

Member. Supervisory Commillee

Approved for the Major Department

Chairman Dean

Approved for the Graduate Council

B.G. Dick

Deiln of The Graduale School

ABSTRACT

Propranolol is a nonspecific beta-adrenergic antagonist used for the treatment of

cardiac arrhythmias, angina pectoris and hypertension. A significant problem in

propranolol therapy is that it undergoes extensive presystemic metabolism after oral ad­

ministration leading to reduced bioavailability and significantly greater intersubject

variability in blood levels after oral than after intravenous administration. In previous

studies, Garceau et al. demonstrated that the hemisuccinate ester of propranolol, when

administered orally to beagle dogs, yields propranolol levels eight times higher than an

equivalent dose of propranolol hydrochloride, suggesting that the prodrug approach may

~e an effective means of avoiding first-pass metabolism of drugs which undergo extensive

first-pass elimination.

The purpose of this study is to obtain preliminary information to be used in sub­

sequent mechanistic studies of the avoidance of first-pass metabolism by prodrugs. The

results conclude that O-acyl ester prodrugs of propranolol are suitable as model com­

pounds for mechanistic studies focussing on the use of the prodrug approach to bypass

first-pass metabolism by the liver. The Spraque Dawley rat has also been identified as a

suitable animal model for such studies. However, since both the acetate and succinate

esters were similar in bioavailability in the current study, both being higher than

propranolol, lipophilicity alone may not be the determinig factor in these results.

CONTENTS

ABS~CT ••••••••••••••••••••••••••••••••••••••••••••••••••••••••• iv

LIST OF FIGURES ••••••••••••••••••••••••••••••••••••••••••••••••••• vii

LIST OF TABLES •••••••••••••••••••••••••••••••••••••••••••••••••••• viii

ACKNOWLEDGMENTS ............................................... ix

Chapter 1. INTRODUCTION .................................................. 1

2. LITERA.TU'R,E REVIEW ••• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 3

2.1 Morphology Of The Liver ••••••••••••••••••••••••••••••••••••• 3 2.2 Models Of Hepatic Transport And Metabolism Kinetics....... •. • 10 2.3 Prodrugs For Bypassing First-Pass Metabolism •••••••••••••••.• 14 2.4 Pharmacology Of Propranolol •••••••••••••••••••••••••••••••. 16 2.5 First-Pass Metabolism Of Propranolol •••••••••••••••••.••••••• 16 2.6 Possible Strategies For Bypassing The First-Pass Metabolism Of

Propranolol ................................................. 22

3. STATEMENT OF THE PROBLEM ••••.•.•••••••.••••..••••..•••••.•• 23

4. EXPERIMENTAL METHODS •••• . • • • • • • • • • • • • . • . . • • • . • • . • . • • . • • • • • • 26

4.1 Chemical Synthesis Of Propranolol Prodrugs •••••.••..•••••.•• 26 4.1.1 Synthesis of propranolol hemisuccinate hydrochloride salt 26 4.1.2 Synthesis of propranolol acetate hydrochloride salt •••••. 27

4.2 Preparation Of Stock Solutions ••••••••••••••••••••••••••• 0... 27 4.3 Analysis Of Propranolol And Prodrugs Of Propranolol In Plasma 27 4.4 High Performance Liquid Chromatographic Analysis • 0 • 0 •••• 0 • • 28 4.5 Stabilities Of Propranolol And Prodrugs In Buffer Solutions 0" 0 • 30 4.6 In Vitro Blood Hydrolysis Kinetics ••• 0 o •• 0 0 • '0 • 0 • o ••• 0 0 •• 0 0 00' 30 4.7 In Vivo Pharmacokinetic Studies ••• 0 0 0 •• 0 •••••• 0 •• 0 •• 0 • 0 • 0 • • • 30 4.8 Calculation of Pharmacokinetic Parameters •••••• 0 • 0 ••••••• 0 0 • 31

5. RESULTS AND DISCUSSION • 0 • 0 • 0 •••••••• 0 •••••• 0 • • • • • • • • • • • • • • • • • 32

5.1 Assay Validation ••••••••••••••••••••••••••••••••••••••••••••• 32 5.2 Prodrug Stability In Stock Solutions And Quenched Samples 32 5.3 Prodrug Bioconversion At pH 7.40 In Buffer And In Plasma •••••• 38 5.4 In Vivo Pharmacokinetic Studies • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 38

6. CONCLUSIONS ................................................... 54

REFERENCES 55

vi

LIST OF FIGURES

1. Principal Surfaces of Hepatocytes in Relation to the Perihepatocellular Spaces ............................................................. 4

2. The Microvascular Unit of Liver Parenchyma, the Hepatic Acinus. courtesy J.L.Campra and T.B.Reyndds, reprinted with permission from The Liver: Biology & Pathology, Chapter 37, 1982, Copyright 1982, Raven Press N.Y. 6

3. The Outflow Curves of Multiple-indicator Dilution Method. courtesy C.A.Goresky, reprinted with permission from The Liver: Biology & Pathol-ogy, chapter 34, 1982, Copyright 1982, Raven Press, N.Y. .................... 8

4. Proposed Metabolic Pathways of Propranolol in the Liver. . . . . . . . . . . . . . . . . . .. 20

5. Structures of Propranolol and its Prodrugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 24

6. HPLC Chromatography of Propranolol (e), Propranolol Hemisuccinate (.) and Propranolol Acetate (+) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 29

7. Recovery of Propranolol (e), Propranolol Hemisuccinate (.) and Propranolol Acetate (+) in 200111 Spiked Plasma after C-18 RP SEP-PAK Purification ......................................................... 35

8. Validation of the Stability of Propranolol (e), Propranolol Hemisuccinate (.) and Propranolol Acetate (+) in Quenching Solution at 4°C. . . . . . . . . . . . . . .. 37

9. Semilogarithmic Plots of the Concentrations of Propranolol (e), Propranolol Hemisuccinate (+) and Propranolol Acetate (*) Versus Time in PH 7.40 Phosphate Buffer at 37°C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 39

10. Semilogarithmic Plots of In-vitro Hydrolysis of Propranolol (e), Propranolol Hemisuccinate (.) and Propranolol Acetate (+) in Rat Plasma at 37°C . . . . . . . .. 40

11. Plasma Concentrations of Propranolol in Sprague Dawley Rats after Oral Administration of Propranolol (e), Propranolol Hemisuccinate (.) and Propranolol Acetate (+) at Dose 10 mg/kg (n=3) . . . . . . . . . . . . . . . . . . . . . . . . . . .. 42

12. Plasma Concentrations of Propranolol Hemisuccinate after Oral Ad-ministration of Propranolol Hemisuccinate HCl at Dose 10 mglkg(n=3) ........ 48

13. Plasma Concentrations of Propranolol Acetate after Oral Administration of Propranolol Acetate at Dose 10 mglkg(n=3) .............................. 49

LIST OF TABLES

1. Validation of the SEP-PAKMethod for Propranolol and its Prodrugs in Aqueous Solution .................................................... 33

2. Recovery of Propranolol in Spiked Plasma after SEP-P AK Purification . . . . . . . .. 34

3. The Hydrolysis of PRO·HCI, HS-HCI and AC-HCI in PH 4.00 Phosphate Buffer at 37°C and 4°C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 36

4. Plasma Concentrations of Propranolol (Mean±SEM) in Sprague Dawley Rats after Oral Administration of Propranolol HCI (n=3) at a Dose of 10 mg/kg .............................................................. 43

5. Plasma Concentrations of Propranolol (Mean±SEM) in Sprague Dawley Rats after Oral Administration of Propranolol Hemisuccinate HCI (n=3) (Dose = 10 mg/kg Propranolol HCI Equivalents) ........................... 44

6. Plasma Concentrations of Propranolol (Mean±SEM) in Sprague Dawley Rats after Oral Administration of Propranolol Acetate HCI (n=3) (Dose = 10 mg/kg Propranolol HCI Equivalents) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 45

7. Pharmacokinetic Parameters Obtained from Least-Squares Regression of the Propranolol Blood Concentrations Resulting from Oral Administration of Propranolol and its Acetate and Succinate Esters According to Equation 1 .................................................................. 46

8. Plasma Concentrations of Propranolol Hemisuccinate (In Propranolol Equivalent Unit) after Oral Administration of Propranolol Hemisuccinate HCI (Dose = 10 mg/kg Propranolol HCI Equivalents) . . . . . . . . . . . . . . . . . . . . . . .. 50

9. Plasma Concentrations of Propranolol Acetate (in Propranolol Equivalent Unit) after Oral Administration of Propranolol Acetate HCI (Dose = 10 mg/kg Propranolol HCI Equivalents) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 51

10. Apparent Absorption and Elimination Rate Constants Obtained from Least-Squares Regression of the Propranolol Hemisuccinate and Acetate Blood Concentrations after Oral Administration According to Equation 1 . . . . . .. 52

ACKNOWLEDGMENTS

I would like to thank my advisor, Bradley D. Anderson, for his valuable advice. I

am also thankful to William I. Higuchi and Jeffrey L. Fox, who point me to many inter­

esting issues about this work. I am indebted to my husband, Hwa Chung, whose under­

standing and love makes possible my going through and finishing this thesis. Finally, to

my parents I extend my greatest gratitute, since without them I would not have stood

here and pursued my goal.

CHAPTER!

INTRODUCTION

The ultimate goal of research in drug delivery is to develop strategies for delivering

pharmacologically active agents to specific target sites in a highly efficient and controlled

manner. In contrast to this ideal, present-day drug therapy is a highly inefficient process

due to a multitude of competing processes and biological barriers which combine to reduce

the quantity of an administered drug reaching its intended target. An understanding of

these barriers to efficient drug delivery and the development of methods to circumvent

them is essential.

For orally administered drugs, first-pass liver uptake and metabolism is a formid­

able barrier to efficient delivery. Extensive first-pass metabolism, in addition to decreas­

ing the percentage of dose reaching its intended site of action, often leads to serious

variability in bioavailability, necessitating careful monitoring of patient blood levels.

While first-pass metabolism can be avoided by selecting alternative routes of administra­

tion (Lv., transdermal, rectal, etc.), oral administration is generally the preferred route.

Rational approaches for bypassing first-pass metabolism which could be applied to drug

candidates which exhibit promising pharmacological activity but undergo extensive first­

pass metabolism when administered orally would be quite valuable.

There are a number of examples of the use of the prodrug approach to reduce first­

pass metabolism and thereby increase oral bioavailability. For example, Dobrinska et a1.

[12,47,48] demonstrated that the highly lipophilic pivaloyloxyethyl ester of methyldopa

exhibited a 2.3-fold higher systemic availability of methyldopa following an ora1 dose com­

pared to an equivalent oral dose of methyldopa due at least partially to decreased first­

pass metabolism. Formation of the mono-O-sulfate conjugate of methyldopa was shown to

2

be lower for the prodrug than for methyldopa [47, 48]. Bodor et al. [7], studying various

classes of transient derivatives of L-dopa, which is used in the treatment of Parkinsonism,

showed that derivatives such as the diacetyl-L-dopa ester, the benzyl ester and various

dipeptides effectively provided protection against metabolism which occurred in the

gastrointestinal tract and/or during the first passage through the liver, resulting in a

significantly better bioavailability of the drug. Phenolethanolamines, which are used as

adrenergic agents, are of limited utility due to extensive gut wall first-pass metabolism.

For example, 80% of orally administered isoproterenol is sulfated in the intestinal wall in

the first-pass [58]. The oral bioavailability of Etilefrine® (o.-[(ethylamino)- methyl] -3'­

hydroxybenzyl alcohol) is 0.50, even though it is completely absorbed, due to gut wall

metabolism [58]. To avoid the attack of conjugating enzymes at the 3'- hydroxy group,

Etilefrine® is masked by formation of the 3'-acetate prodrug [58].

The antihypertensive drug, propranolol, also exhibits low and variable

bioavailability after oral administration due to extensive first-pass metabolism in the

liver. Garceau et al. [16J have shown that following oral administration of propranolol

hemisuccinate in dogs, plasma propranolol levels were eight times higher than after an

equivalent dose of propranolol. Thus, the prodrug effectively protects the drug from first­

pass elimination by the liver [16J.

The sequence of events which leads to drug elimination in the first-pass through the

liver and the mechanism by which prodrugs avoid this unfavorable outcome are not wen

understood. The determination of these mechanisms and the development of chemical

modification strategies which can be applied to highly extracted compounds is the long­

range goal of research in this area. The purpose of the present work is to obtain prelimi­

nary data in rats on the first-pass metabolism of two chemically disparate prodrugs of the

model compound propranolol, a compound which is known to undergo extensive first-pass

metabolism [9, 18,42,44]. Information on these propranolol esters will serve as a start­

ing point for more in-depth studies to ascertain the mechanism by which these compounds

avoid first-pass metabolism.

CHAPTER 2

LITERATURE REVIEW

2.1 Morphology Of The Liver

The liver is the most important and effective organ in the body for elimination of

drugs. The structural organization of the parenchymal and vascular elements of the liver

uniquely adapts it to serve as a guardian interposed between the digestive tract and the

rest of the body [8]. Thus, a major function of the liver involves the uptake of substrates

from the gastrointestinal tract (GI tract) and their subsequent storage, metabolism, and

distribution to blood and bile.

Physiologically, there are two flow systems interspersed in the liver [8]. Blood flows

into the liver from the portal vein and hepatic artery and leaves via the hepatic vein.

These vessels are interspersed at 90° angles to each other in the liver. The portal vein

provides the liver with 70% to 75% of the blood supply of the human liver, and the hepatic

artery provides the remaining 25% to 30% [8].

The space between the terminal portal venule and the terminal hepatic vein is filled

with single cell·thick sheets of hepatocytes lining specialized capillaries, the liver

sinusoids [8](Figure 1). These channels, which may be regarded as the end organs of the

circulation, have a unique structure. They are lined by flattened endothelial cells con­

taining fenestrae of various sizes-- from 0.1 Jlm to 2 Jlm, so even high molecular weight

(up to 250,000 gm per mole) substances can be sieved through these fenestrae [8].

Beneath lies the space of Disse, comprising much of the the functional extracellular space

of the liver. The fenestrae provide continuity between the vascular space and the space of

Disse. No continuous anatomic barrier exists between the plasma and the Disse space,

and hence dissolved substances have free access to the underlying surface of the liver cells

(Figure 1).

UPTAKE

HEPATOCYTE

Figure 1: Principal Surfaces of Hepatocytes in Relation to the Perihepatocellular Spaces

The parenchymal liver cells, or hepatocytes, number about 250 billion in the normal

adult organ and occupy about 80% of the parenchymal volume of the liver. They are

polyhedral multifaceted cells with eight or more surfaces. Their diameters vary from 19.0

J.1m to 20.5 J.1m, with the average of 20 J.1m [22]. Biochemically, the hepatocyte plasma

membrane has a relatively high amount of glycosylated protein and lipids and a complex

spectrum of polypeptides [34]. The plasma membrane of hepatocytes serves to maximize

their surface area for solute uptake [38]. This design is optimal for the functioning of the

liver as a barrier to the oral delivery of foreign substances. The cell membrane is not only

considered as a static structure protecting the cell against the loss of essential con­

stituents by diffusion and allowing molecules and ions to exchange according to their lipid

solubility, molecular size and electric charge, but also the cell membrane is actively in­

volved in the transfer of some substances into and out of the cell due to the presence of

various carrier systems in the membrane [20]. Therefore, the effect of the cell membrane

on the entry of materials into liver cells varies with the nature of the substance.

Once substances penetrate through the fenestrae into the extracellular space, they

undergo dilution in the extracellular space and, depending on their structure, may be

bound to hepatocytes or transported into the cell and subsequently metabolized, excreted

into the bile, or released back into the Disse space. First-pass metabolism of an orally

administered drug depends on the relative rates of these processes occurring within

hepatocytes.

The functional unit of the liver is the simple acinus (Figure 2) [8, 22], consisting of a

small parenchymal mass. Blood flows from the terminal portal venule into acinar

sinusoids (zone 1) and flows sequentially through zone 2 and into zone 3, where it exits

near the terminal hepatic venules. Acinar zones 1,2 and 3, respectively, represent areas

supplied with blood varying in quality with regard to oxygen and nutrient contents [22].

The function, structure, and shape of a given cell depends on its specific position in

the liver cell plate and the various zones of the acinus. Several studies using histochemi­

cal techniques or microdissection have found differences in the distribution of enzymes

among hepatocytes of the liver acinus [22]. Zonal heterogeneity may play an important

p. v.·~· ... _-r~~+..l

- -." " / , I \

\ \

i' ,

Figure 2: The Microvascular Unit of Liver Parenchyma, the Hepatic Acinus. courtesy J.L.Campra and T.B.Reyndds, reprinted with

permission from The Liver: Biology & Pathology, Chapter 37, 1982, Copyright 1982,

Raven Press N.Y.

\ J \

6

7

role in metabolism. By microdissection of the acinar zones followed by enzymatic

analysis, Novikoff et al. [36, 37] have found that periportal hepatocytes (zone 1) contain

numerous mitochondria with higher levels of glucose-6-phosphatase. Other researchers

[21,41] also have shown that phosphoenol-pyruvate carboxykinase, presumably a rate­

limiting enzyme in gluconeogenesis, and fructose-1,6-diphosphatase are deposited

preferentially in periportal hepatocytes. In contrast, pyruvic kinase, an enzyme involved

in glycolysis, is higher in the hepatocytes of acinar zone 3. These data suggest that acinar

zone 1 hepatocytes are predominantly engaged in gluconeogenesis, while cells of acinar

zone 3 participate predominantly in glycolysis [22]. There is no sharp border existing

between gluconeogenic and glycolytic hepatocytes, however, since enzymes corresponding

to each pathway can be detected in both zones. Although the amount of lipid present in

the hepatocytes is subject to metabolic variation, lipid is more concentrated in the

centrilobular zone (zone 3) of recently fed animals [11]. Also the activities of the enzymes

beta-hydroxybutyric dehydrogenase [36, 51] and esterase [36, 37] are apparently higher in

cells of acinar zone 3. Thus, the metabolic function of the centrilobular hepatocytes is

more likely related to lipid metabolism [36,37]. Several lines of evidence suggest

[22] that cells closer to the terminal hepatic venule (acinar zone 3) contribute

predominantly to the metabolism of drugs via the cytochrome P-450 system.

Since most of the metabolic enzymes lie in the hepatocyte interior, uptake into and

transport across the hepatocyte plasma membrane must be the first steps in the sequence

of the liver metabolism of drugs. The time course for the passage of a given orally ad­

ministered drug through the liver and its access to the hepatocytes is therefore of interest.

The investigation of hepatic transport kinetics was pioneered by Goresky [20] in studies of

single-pass multiple-indicator dilution outflow curves from the liver (Figure 3). A single

injection, multiple indicator dilution technique consists of rapid injection of a reference

substance and a study substance into the blood flowing into the organ. An injection

catheter was placed in the portal vein and a venous collection catheter was placed in the

left main hepatic vein. Some typical outflow curves were obtained by Goresky when using

the multiple indicator dilution [20]. Labeled red blood cells which can not pass through

• o ---::::I o I')

D

'"

• red blood eell

t I a beled sue rose

• I a beled glucose

I ~ t, \ t) ~ l,.

• • •••• I ................ _.

l ...... •. .. .... . . . . D--------~~ ________ .. ~ ________ ~_··_-~

1D aD 3D Time, seeond

Figure 3: The Outflow Curves of Multiple-indicator Dilution Method. courtesy C.A Goresky, reprinted with permission from

The Liver: Biology & Pathology, chapter 34, 1982, Copyright 1982, Raven Press, N. Y.

8

9

the sinusoidal fenestrae into the space of Disse are used as the vascular reference and

labeled sucrose (14C-sucrose) as the second or extracellular reference. After a lag time of

approximately 6 seconds the concentration of red blood cells in the exiting fluid rises

rapidly to a peak in less than 10 seconds. Virtually 100% of the red blood cells introduced

are cleared from the liver within 20 seconds. As compared with the curve of the red cell,

the 14C-sucrose curve demonstrates the effect of distribution in the interstitial or Disse

space on the outflow curve. The outflow appearance is delayed , the curve rises to a later

and lower peak, and the curve downslope decays more slowly [20]. Lateral diffusion oc,­

curs virtually instantaneously at each point along the length of the sinusoid while axial

gradients are preserved. The mean residence time of the 14C-sucrose in the liver is on the

order of seconds (10-20 seconds). This relatively short residence time suggests that

hepatocyte uptake must be fairly rapid for first-pass metabolism to be significant. The

outflow curve of 3H-D-Glucose which rapidly penetrates into hepatocytes begins to rise at

the same time as that of the labeled sucrose but increases slowly to a peak which is later

and substantially lower than that for sucrose [20]; also the curve is biphasic with a much

slower elimination phase. This curve reflects a partitioning of glucose between the in­

tracellular and extracellular spaces. Since glucose is not metabolized in this experiment,

it is eventually completely cleared from the liver. All of these outflow curves are

broadened, however, due to the distribution of sinusoidal and exchanging vessel transit

times [20]. In conclusion, the mean residence time of a substance which can not penetrate

into the hepatocyte is on the order of seconds, suggesting that uptake into hepatocytes

must be relatively rapid for first-pass metabolism to be significant. Not every substance

is taken up by the cells. The rate of the uptake process therefore depends on the nature of

the substance, lipophilicity, molecular weight etc ..

Even though the anatomical structure of the liver and the plasma membrane of

hepatocytes serves to facilitate solute uptake, it is well recognized that a lack of

metabolism for some compounds may be in part related to their low permeability [17, 61].

Goresky [20], using a multiple- indicator dilution technique, showed that inhibiting the

uptake rate of galactose across the liver cell membrane with glucose significantly in-

10

creased the throughput component in outflow curves thus indicating that inhibition of cell

uptake rates may be an effective means of minimizing first-pass metabolism.

2.2 Models Of Hepatic Transport And Metabolism Kinetics

Elimination of substrates from the blood by the intact liver depends on the

processes of metabolism, hepatic flow and anatomical arrangements of the sinusoids. Two

well defined models have been widely used to predict the effect of changes in blood flow,

protein binding and drug metabolizing enzyme activity on the hepatic clearance of drugs.

The venous equilibrium model (well-stirred model) assumes that the liver is a single well

mixed compartment, such that the concentration of unbound drug throughout the liver is

uniform and equal to that in the hepatic venous effluent. As pointed out by Forker [31J,

this model ignores the solute concentration gradient that develops along each sinusoid as

the inevitable consequence of solute removal and it fails to account for imperfect mixing

in the extrahepatic circulation. The sinusoidal model (parallel tube model) regards the

liver as a series of parallel identical tubes with enzymes distributed evenly around the

tubes and the concentration of drug declining exponentially along the length of the tube

[1, 27,29, 60J. Further modifications of the sinusoidal model take into account the

heterogeneity of distribution of metabolizing enzymes, shunt pathways [3J, and in corpora-

tion of cooperative enzyme elimination kinetics [26].

The mathematical equations for these two models in the perfused liver studies are

for the venous equilibrium model:

and for the sinusoidal model:

R . e-f·CLinl I Q C =------

ss Q. (l-e-fCLwIQ)

where Cas = steady-state concentration; R = dose rate administered by constant infusion

into portal vein; f = fraction of unbound drug in perfusate; CLint = intrinsic hepatic

clearance of unbound drug; and Q = circuit flow rate. From the above equations can be

derived [60] the following two new equations which relate the unbound steady-state con-

11

centration (Cuss) after a constant rate infusion into the portal vein to the same deter­

minants of hepatic drug elimination. For the venous equilibrium model:

For the sinusoidal model:

f-R -e-fCLin.,IQ

Cuss = Css-f= Q.(l-e-fCLw /Q)

It can be seen by inspection of these equations that, unlike the sinusoidal model, the

venous equilibrium model predicts that the unbound perfusate drug concentration (Cuss)

will be independent of both the fraction of unbound drug (f) in the perfusate and the

circuit flow rate (Q).

Which of these two commonly used models is the more appropriate to describe the

hepatic elimination of drugs and other compounds is still in dispute [1, 27, 29, 39, 40, 60].

Studies with lidocaine [40], lignocaine [1], pethidine [1] and propranolol [27) have been

presented as evidence to support the venous equilibrium model. According to the above

equations relating the steady-state perfusate drug concentration in the reservoir of a per­

fused liver circuit after constant infusion into the portal vein, the steady-state concentra­

tion of drug in the reservoir of a perfused liver circuit for the venous equilibrium model is

independent of hepatic blood flow, while for the sinusoidal model, the steady-state con-

centration of drug decreases with lower blood flow. The behavior of lidocaine, lignocaine

and pethidine under linear conditions to changes in hepatic blood flow rate were ex-

amined in the perfused rat liver [1, 27, 40]. The steady-state output concentrations of

these substances (lidocaine, lignocaine and pethidine) leaving the liver were not sig-

nificantly changed with varying hepatic blood flow. Thus, the data appear to be predicted

better by a "well-stirred" model than by a "parallel tube" model [1,27,40).

The venous equilibrium model also predicts that elimination of unbound drug will

be independent of the degree of protein binding, whereas the sinusoidal model predicts a

marked increase in elimination with an increasing unbound fraction. In experiments

with propranolol, the unbound steady-state propranolol concentration in the hepatic

venous effluent remained unchanged, despite an almost 7 -fold increase in the free fraction

12

of propranolol perfusing the rat liver [27]. The data conform precisely to the predictions

of the venous equilibrium model and appear to be incompatible with the sinusoidal model,

which predicts a 1 ~O-fold decrease in unbound concentration [27].

On the other hand, Weisiger et al. [60] have illustrated that concentration gradients

are formed during the uptake of 125I-thyroxine by the perfused rat liver. Autoradiog­

raphy of tissue slices after perfusion of the portal vein at physiologic flow rates with

protein-free buffer containing 125I-thyroxine demonstrated a rapid exponential fall in den-

sity with distance from the portal venule, declining by half for each 8% of the mean length

of the sinusoid. Analysis of the data using models in which each sinusoid was

represented by different numbers of sequentially perfused compartments (1-20) indicated

that at least eight compartments were necessary to account for the magnitude of the

gradients seen. These results [60] are incompatible with the venous equilibrium model,

which predicts a uniform concentration at all points along the sinusoid, and suggest that

it may be necessary to consider the effects of sinusoidal concentration gradients when

studying the hepatic removal of efficiently extracted substances.

Keiding and Chiarantini [29] have suggested that elimination should take place in

hepatocytes lining sinusoidal tubes, perfused unidirectionally from the inlet to the outlet,

following Michaelis-Menten kinetics with a maximal elimination rate (V max) and the half­

saturation concentration (~). The mathematical description of these phenomena gives:

V max' C' C' = Ci-Co

V = Km +C" In(Ci /CJ

This is the Michaelis-Menten relation where the concentration (C') is the logarithmic

average of Ci and Co, and v is the elimination rate. This relation is independent of flow

rate (Q), but Co depends on Q at a given elimination rate:

C = v o Q. (e(V m.ru; -u)/Q.Km-l)

Changes of flow at larger values of V mJ~ will cause larger changes of Co than at

smaller values of V mall\n. When V malQ·l\n is less than 0.1, Co is practically inde­

pendent of flow.

In the rat liver perfusion studies of Keiding [29], reduction of the hepatic blood flow

13

at a constant galactose elimination rate decreased the outflow concentration and in­

creased the inflow concentration, whereas there was no significant change in the logarith­

mic average concentration. The outflow concentration (Co) adapted previously is not an

appropriate approximation to the sinusoidal concentration for hepatic elimination of

galactose. In that case, therefore, Co should be replaced by C'. Only in the limiting case

of the sinusoidal model, when Vmu!Q'~ ~ 0, may C' be approximated by Co. These

observations are consistent with the predictions of the sinusoidal perfusion model and

demonstrate that the decreasing concentration along the sinusoids has to be taken into

account in the evaluation of elimination kinetics in the intact liver as first suggested by

Goresky et a1. [20]. In addition, there are numerous examples of the existence of a

"lobular gradienttl in hepatic elimination (6, 28, 32], consistent with the predictions of the

sinusoidal model.

Forker and Luxon [15] have demonstrated some examples of anatomic and

metabolic heterogeneity within the liver lobule. They think that the kinetics observed in

the liver are caused by the summation of the individual sinusoidal transport steps. The

maximal rate of removal of galactose and ethanol occurs in the periportal acinar zone 1,

whereas metabolism of lidocaine and propranolol would be expected to occur maximally in

the centrilobular acinar zone 3, where enzymes responsible for biotransformation of these

drugs are principally located (27]. However, Anderson et a1. (2J demonstrated using

autoradiography that higher concentrations of propranolol deposit in the periportal zone

(zone 1) of the liver lobule after the injection of propranolol into the portal vein (2]. They

pointed out that propranolol was efficiently cleared from the perfusate in the periportal

zone (zone 1).

Since there is no simple anatomical explanation for the applicability of either

model, a complete description of the hepatic elimination of a particular substance is

generally based on the experimental verification of the appropriate model (27].

14

2.3 Prodrugs For Bypassing First-Pass Metabolism

There are a number of examples of the use of the prodrug approach to reduce first­

pass metabolism and thereby increase oral bioavailability. Most of the published ex­

amples, however appear to have focussed on bypassing gastrointestinal metabolism. For

example, the systemic availability of methyldopa (L-a-methyl-3,4-

dihydroxyphenylalanine) was increased 2.3-fold after an oral dose of the pivaloyoxyethyl

(POE) ester compared to an equivalent dose of methyldopa itself [12, 47,48].

Methyldopa's oral bioavailability is reduced due to first-pass O-sulfation in the gastroin­

testinal tract [12, 47]. The highly lipophilic O-acyl prodrug enhanced the absorption,

resulting in reduced exposure of the parent drug to the conjugating enzyme. The intact

prodrug is subsequently hydrolyzed to methyldopa in the liver. The amount of sulfate

conjugate recovered in urine is equivalent to about one third of the systemically available

dose of oral methyldopa compared to about one eleventh for the POE ester [12]. Ad­

ministration of methyldopa as the POE ester prodrug thereby not only increases the

bioavailability of free methyldopa to the general circulation [12], but also serves to protect

the drug from sulfation during the first-pass.

L-dopa (L-3,4-dihdroxyphenylalanine) is still generally accepted as the first drug of

choice in the management of Parkinsonism disease [7]. Long-term therapy with L-dopa

is, however, associated with a number of therapeutic problems. L-dopa is usually ad­

ministered orally and, in man as in dog, the material in solution appears to be well ab­

sorbed, primarily in the small bowel by a special carrier transport mechanism [7, 34]. But

the drug is extensively metabolized in the gastrointestinal tract and during its first pas­

sage through the liver, so that relatively little arrives in the blood as intact L-dopa [7].

Bodor and his coworkers studied various classes of transient derivatives of L-dopa as

prodrugs to attempt to solve these problems [7]. A number of them such as diacetyl,

benzyl esters and the peptide (the N-formyldiacetyl L-dopa), result in significantly higher

L-dopa bioavailability and a much more favorable plasma L-dopa/dopamine (the

metabolite of Dopa) ratio [7]. The most interesting observation is that after a dipeptide

prodrug was compared to L-dopa in an enteric coated formulation, the prodrug was found

15

to be superior even to the enteric L-dopa formulation, which suggests that besides protec­

tion in the gastrointestinal tract, the prodrug provides protection against metabolism

during absorption and enhances the absorption of L-dopa possibly by increasing the pas­

sive transport contribution [7].

Etilefrine® (a-[(ethylamino)-methyl]-3' -hydroxybenzyl alcohol) contains a free

hydroxyl group in the meta-position of a phenyl ring and is therefore subject to gut wall

first-pass sulfation [58]. To avoid the attack of conjugating enzymes at the 3'- hydroxy

group, it was masked by acylation to form the 3'-{O-pivaloyl) derivative. The prod~g

showed favorable solubility and improved lipophilicity and exhibited reduced sulfation in

the first-pass [58].

'While the above examples serve to illustrate the utility of the prodrug approach in

bypassing first-pass metabolism in the gastrointestinal tract, the primary interest of this

work is to develop strategies useful in avoiding first-pass metabolism by the liver. There

are limited data dealing with the use of prodrugs to bypass liver first-pass elimination.

One particularly interesting study was a comparison of the oral bioavailability in dogs of

the succinate ester of propranolol with propranolol by Garceau et al. [16]. Following oral

administration of propranolol hemisuccinate, plasma propranolol levels were eight times

higher than after an equivalent oral dose of propranolol. The plasma propranolol

hemisuccinate levels peaked early, and levels were negligible at six hours after dosing.

These data suggested that the prodrug was rapidly absorbed and converted to

propranolol. As an index of the amount of the prodrug absorbed orally, the AUe of the

prodrug after oral administration was compared to that after intravenous dosing. The

authors claimed that 70% of the oral dose of the prodrug was absorbed [16]. This study

demonstrated the potential of the prod rug approach for protecting highly metabolized

drug such as propranolol from liver first-pass elimination. Because of its low

bioavailability due to first-pass metabolism by the liver [9, 18, 23, 42, 44] and because the

prodrug approach seemed to be effective in improving bioavailability [16], propranolol and

prodrugs of propranolol were selected as model compounds for the present study.

16

2.4 Pharmacology Of Propranolol

As mentioned, propranolol is a nonselective ~-adrenergic blocking agent that is used

widely for the treatment of hypertension, the prophylaxis of angina pectoris, and the con­

trol of certain types of cardiac arrhythmias. It blocks ~1 and ~2 receptors competitively

and does not exhibit any intrinsic agonistic properties.

The most important effects of ~-adrenergic blocking drugs are on the cardiovascular

system, predominantly due to actions on the heart. Propranolol decreases heart rate and

cardiac output, prolongs mechanical systole, and slightly decreases blood presure in rest­

ing subjects [19]. Peripheral resistance is increased as a result of compensatory sym­

pathetic reflexes, and blood flow to all tissues except the brain is reduced [19]. Total

myocardial oxygen consumption and coronary blood flow are also decreased as a result of

reductions of heart rate, ventricular systolic pressure and contractility [19]. ~-Adrenergic

agonists are known to increase modestly the release of norepinephrine from adrenergic

nerve terminals. Propranolol blocks this effect and such impairment of the release of

norepinephrine following sympathetic nerve stimulation might contribute to the an­

tihypertensive effects of the drug. Reduction in cardiac output occurs rather promptly

after administration of propranolol. However, the hypotensive effects of propranolol do

not usually appear as rapidly. Propranolol is particularly useful clinically in combination

with vasodilators that act directly on the vasculature [19].

2.5 First-Pass Metabolism Of Propranolol

The oral bioavailability of propranolol is quite low, varying with animal species and

the oral dose. Walle et al. [52] illustrated that the oral bioavailability of propranolol in

man (62-86 kg) was 18.9-23.5% at doses ranging from 20 mg to 80 mg. Other phar­

macokinetic studies in healthy adults and in patients with various disease states have

demonstrated as much as 10- to 20-fold variation in the plasma concentrations of

propranolol between individuals after oral, but not intravenous, doses of the drugs

[42,52,53]. The problem is more severe in controlled release systems. AUe estimates

suggest that the bioavailability of a slow-release oral formulation following a single dose

was about one-third that of the conventional preparation in man [13].

17

In rats, the bioavailability was 7.5-25.3% with the oral dose ranging from 1.0 mg/kg

to 10.0 mg/kg [25], and in dogs (average 24.9 kg), the systemic bioavailability of a 20 mg

oral dose was 8.0 ± 8% [30]. Iwamoto and Lo [25, 30], respectively, have found that there

is no gastrointestinal first-pass metabolism of propranolol in rats and dogs. Rather, the

cause of the low bioavailability of propranolol is extensive liver first-pass elimination.

Some researchers [45] have postulated that hepatic metabolism is not the only

process which might be involved in a concentration-dependent change in hepatic extrac­

tion - that the ability of the liver to accumulate unchanged drug is also important in the

overall removal process. Because the liver receives a portal blood supply draining from

the gut, much higher portal venous concentrations of propranolol are obtained after oral

administration than intravenous administration. In general, the extent of uptake

depends on the affinity of solute for the hepatocyte plasma membrane or limited intracel­

lular binding sites. The extent of metabolism may depend on the extent of uptake. Higher

portal vein concentrations of propranolol result in lower hepatic extraction and reduced

metabolism of propranolol in the liver. Both the pattern of metabolism and the relation­

ship between circulating drug concentrations and dose change depending on the route of

administration [45].

Early in-vivo studies of the liver uptake process of propranolol in dogs by Shand et

al. [42, 43, 44, 45] established that the liver extraction ratio is extremely high during the

intravenous administration of propranolol. Propranolol concentrations in the inferior

vena cava increased linearly with the dosage. Drug concentrations in the hepatic vein

were always considerably lower and also increased linearly with the dosage. When the

inferior vena caval concentrations were below 0.2 J,Lgiml, hepatic extraction was greater

than 90% and if the concentration increased to 2.0 J,Lgiml, the extraction was reduced to

80% [43]. These data agree well with later findings [25] in rats that hepatic extraction

ratios range from 92.5% to 74.7% with oral dosages ranging from 1.0 mg/kg to 10.0 mg/kg.

Due to the fact that hepatic extraction exhibits a nonlinear relationship with dosage in

man and rats, McEwen, Shand, Evans [14, 33,44,45] have suggested that there may be

two processes which result in the concentration-dependent change in hepatic extraction:

18

one is a saturable, high-affinity low-capacity binding process, and the other one is a low­

affinity high-capacity process. Later, Anderson et al. [2] confirmed that statement in per­

fused isolated rat liver studies, and supplemented that the low affinity process can be

saturated. Thus the low affinity uptake probably involved a binding site and not merely a

"partitioning" of propranolol between liver and perfusate as suggested previously [2].

Octanol-water partition coefficients, determined for propranolol by Davis and Elson

[10], suggested that propranolol should bind readily to the hepatocyte plasma membrane.

The partition coefficient of propranolol at pH 7.12 is 101.58 at 22.5°C, even though

propranolol is mostly ionized at this pH. The nonspecific interaction of propranolol with

phospholipid membranes has been investigated by Herbette and his colleagues [24].

Neutron diffraction utilizing propranolol deuterated in the naphthalene moiety shows

that the naphthalene moiety of propranolol partitions into the hydrocarbon core of the

dimyristoyl lecithin bilayer, and that the charged amine side chain is most likely posi­

tioned in the aqueous phospholipid head group region [24].

Later, many in-vitro and in-vivo animal studies [2,4] have investigated the deposi­

tion of propranolol in the liver. Anderson [2], in studying isolated rat livers which were

perfused with propranolol in a simplified recirculation system without plasma proteins or

erythrocytes, demonstrated that the propranolol levels in the reservoir declined biphasi­

cally. The first phase (alpha-) is extremely fast and almost flow-limited despite the use of

high perfusion flow rates [2J. Also, the first phase is not significantly diminished by

lowering the temperature or removing oxygen from the system. If transport systems are

involved in this phase, they would have to be extremely rapid processes [2, 38]. On the

other hand, binding processes would be expected to show rapid kinetics [2J. The authors

suggested that the second (beta-) phase in the uptake process is associated with the

metabolism of propranolol and is temperature and oxygen dependent [2]. After 20

minutes of anoxia (95% N2-5% CO2), propranolol was added to the perfusate; there was a

rapid uptake of propranolol by the liver. Repeated additions of propranolol gave similar

results: equilibrium was achieved in about 15 minutes and it appeared that no beta-phase

existed. Introduction of oxygen at 100 minutes led to large oxygen consumption by the

19

liver, accompanied by a progressive removal of propranolol from the perfusate. Intro­

duction of a second drug (imipramine, desipramine, nortriptyline etc.) after the liver had

been loaded with propranolol did cause some release of propranolol from the liver [2],

resulting in reduction of the rate constant of the beta-phase.

The uptake of propranolol by the liver appears to be altered by phenobarbital

pretreatment, indicative of a change in the capacity of the high affinity process. No dis­

cernible change was observed in the low affinity component [2]. Shand's group has ex­

amined the binding of propranolol to rat liver microsomes and suggested that microsomal

binding might account for the high affinity site in rat liver [33]. In Anderson's paper [2],

the significant influence of temperature, oxygen and several drugs on the beta-phase

slope supports this hypothesis. Other investigators have reached the supporting conclu­

sion concerning a closely related compound, alprenolol: microsomal binding may be re­

lated to high affinity uptake of alprenolol by the perfused liver [50]. Unfortunately, an

extremely low capacity of high affinity binding sites «10 J..lg/g wet weight liver) was found

in Shand's study, indicating that high affinity binding sites are not of quantitative impor­

tance [33]. Von Bahr et al. [49] instead attributed the high affinity binding of propranolol

to rat liver cytochrome P-450. However, the identification of binding sites and the

detailed mechanism by which propranolol is metabolized in the hepatocyte still remain to

be evaluated.

The primary metabolic pathways of propranolol include glucuronidation, side-chain

oxidation and ring oxidation (Figure 4) [54, 55, 56, 57]. Walle and his colleagues

[52] suggested that glucuronidation could be an important pathway for the presystemic

removal of single oral doses of propranolol. The area under the plasma concentration­

time curves (AUCs) of propranolol glucuronide (PG) and propranolol were of the same

order after an intravenous dose, 0.05 mg/kg, while after oral doses of 20 mg and 80 mg,

the AUCs of PG were seven times the AUCs of propranolol. The plasma propranolol

levels were not significantly different between 0.05 mg/kg after intravenous administra­

tion and 20 mg after oral administration, but the AUC of PG after the oral dose was seven

times higher than after the intravenous dose [52]. In conflict with these data, Walle

20

GLUe

GLUCURONIDATION

CHAIN OXIDATION

ALDEHYCE

GLUC= Glucurcnic Acid Conjugate

Figure 4: Proposed Metabolic Path ways of Propranolol in the Liver

21

recently found in studies of the metabolites in urine of normal man that the ring oxida­

tion is the "first choice" pathway in the metabolism of propranolol (Figure

4) [54, 55, 56, 57] and is the main determinant of the low and highly variable oral

bioavailability of propranolol. A possible explanation may be that the appearance of

propranolol in the systemic circulation after oral administration is associated with a

change in the pattern of metabolism which may result from the saturation of the oxida­

tive metabolism, or of an alternate pathway at high portal venous drug concentrations

[45].

As more information accumulates, it clearly suggests that hepatic uptake of

propranolol reflects primarily physical binding, probably to several sites with differing

affinities, capacities and locations within the liver [21 .. McEwen et a1. [14, 33, 44, 45] have

suggested that there are at least two binding sites-high-affinity low-capacity binding

sites and low-affinity high-capacity binding sites involved in the liver uptake process as

discussed previously. The fact that phenobarbital increased the capacity of the high­

affinity uptake suggests that an inducible protein or lipoprotein is involved [2). Since

uptake kinetic curves showed that the slower (beta-) phase is a temperature-dependent

process [2), metabolic enzymes are believed to be involved in the hepatic uptake process.

Dose-dependent drug elimination is generally attributed to saturation of the drug­

metabolizing enzymes [2).

Uptake may be followed by a multitude of distributive processes inside the cell in­

cluding binding to cytosolic proteins and enzymes and transport into various subcellular

organelles. Some of these binding processes may be followed by enzyme catalyzed

degradation of the substrate to form metabolites or conjugates which then may them­

selves distribute throughout the hepatocyte and transport across the plasma or can­

nalicular membranes, ultimately appearing in the bile or bloodstream.

2.6 Possible Strategies For Bypassing The First-Pass Metabolism Of Propranolol

22

The evidence available for the hemisuccinate ester of propranolol, which exhibited

higher systemic availability in dogs than propranolol after oral administration, suggests

that prodrugs may be an effective means of bypassing the first-pass metabolism of drugs

which undergo extensive hepatic elimination. However, the mechanism by which the

hemisuccinate ester avoids first-pass mechanism is unknown. Two possible mechanisms

are (1) reduced uptake due to the hydrophilic nature of the propranolol hemisuccinate;

and (2) avoidance of first-pass glucuronidation of the beta-hydroxy group which is blocked

in the succinate ester. The former possibility follows from the fact that propranolol

hemisuccinate is zwitterionic at physiological pH, and therefore likely to be poorly trans­

ported through the hepatocyte membrane. Reduced uptake into the hepatocyte would

lead to reduced hepatic extraction and minimal metabolism, since the enzymes involved

are intracellular. Avoidance of first-pass glucuronidation is also considered as a pos-

sibility in light of the observation that the AUC of propranolol glucuronide after an oral

dose of propranolol was seven times higher than after an intravenous dose [16]. Since the

site of glucuronide formation is blocked in esters of propranolol, ester hydrolysis would

have to occur prior to glucuronidation. This additional requirement may lead to reduced

first-pass elimination.

Of the two mechanisms suggested above, the latter seems less likely in view of the

fact that most of the evidence available suggests that ring oxidation, not glucuronidation,

is the primary determinant of the low and highly variable oral bioavailability of

propranolol [54, 55, 56, 57]. Therefore, the hypothesis that the reduced hepatic elimina­

tion of the hemisuccinate ester is due to its reduced lipophilicity is favored. This

hypothesis can be tested by designing prodrugs varying in lipophilicity.

CHAPTER 3

STATEMENT OF THE PROBLEM

Propranolol, l-isopropylamino-3-(l-naphthoxy)-2-propanol (Figure 5), is a non­

specific beta-adrenergic antagonist used for the treatment of cardiac arrhythmias, angina

pectoris and hypertension. A significant problem in propranolol therapy is that it un­

dergoes extensive presystemic metabolism after oral administration leading to reduced

bioavailability and significantly greater intersubject variability in blood levels after oral

than after intravenous administration. Recently, however, the hemisuccinate ester has

been shown to yield propranolol levels eight times higher after oral administration to dogs

than an equivalent oral dose of propranolol hydrochloride. Although the mechanism by

which this is achieved is not understood, these results suggest that bioreversible chemical

modification of a parent drug to form a prodrug may be an effective strategy to employ to

minimize first-pass metabolism of drugs which are extensively eliminated in the first­

pass. A rational design strategy, however, requires an understanding of the mechanism

by which prodrugs minimize first-pass metabolism.

The purpose of this study is to obtain preliminary information to be used in sub­

sequent mechanistic studies of the avoidance of first-pass metabolism by prodrugs. To

undertake such mechanistic studies, appropriate model compounds and a suitable animal

model are necessary. Propranolol esters were selected as potential model compounds and

Spraque Dawley rats were chosen as a possible animal model. Experiments were con­

ducted to determine whether or not, in the rat, the systemic availability of orally adminis­

tered propranolol is increased by prodrug administration. The esters selected were the

acetate and hemisuccinate. These compounds are expected to differ significantly in

lipophilicity compared to propranolol. The hemisuccinate, being zwitterionic at

physiological pH, is expected to be significantly less lipophilic than propranolol, while the

24

~lecular We1Jht Compound

I. R· -H 29~.81 Propranolol

39~.88 Propranolol Bemisuccinate

III. R •• COCH.1 337.85 Propranolol Acetate

Figure 5: Structures of Propranolol and its Prodrugs

25

acetate should be more lipophilic than propranolol. Therefore, the relative

bioavailabilities of these compounds were examined to obtain a preliminary indication of

the role which prodrug lipophilicity may play in first-pass metabolism.

CHAPTER 4

EXPERIMENTAL METHODS

In this study, two bioreversible esters were compared with their parent drug

(propranolol) to evaluate the influence of the pro-moiety on the relative rates of prodrug

bioconversion and overall oral bioavailability. In vitro hydrolysis studies were carried out

in rat plasma and in buffer solution at various pH's to determine free propranolol and

prodrug concentration-time profiles. In vivo pharmacokinetic studies in which free

propranolol and the remaining prodrug concentration were monitored versus time were

conducted in rats. In order to monitor low concentrations of propranolol and un­

hydrolyzed prodrugs in blood, a high performance liquid chromatographic (HPLC) assay

capable of measurement of submicrograms of propranolol and pro drugs in 200

microliter(JlD volumes of blood was developed.

4.1 Chemical Synthesis Of Propranolol Prodrugs

4.1.1 Synthesis of propranolol hemisuccinate hydrochloride salt

Propranolol hemisuccinate hydrochloride salt (HS-HCl) was obtained from Ayerst

(purity 99.5%) or synthesized by a modification of the method of Nelson and Walker [35].

Propranolol hydrochloride (PRO-HCl), 500 mg (1.69 mmoles) was stirred at 85-90°C with

500 mg of succinic anhydride (5.00 mmoles) in 1 ml dimethylformamide (DMF) for four

hours. The mixture was then cooled down to room temperature (25°C), water was added

to dissolve the O-hemisuccinate product and the solution was washed with ether to

remove DMF and the N-succinylated product. The aqueous solution was then evaporated

in a rotary evaporator to yield propranolol hemisuccinate HCl (HS-HCl) as a white

precipitate. The product was analyzed qualitatively by IR and quantitatively by HPLC,

versus the Ayerst reference compound to confirm its structure and purity.

27

4.1.2 Synthesis of propranolol acetate hydrochloride salt

O-acetylpropranolol was prepared by a method adapted from that of Nelson and

Walker [35]. PRO-HCI, 1.50 grn (5.07 mmoles) was stirred at 100°C with 1 ml acetic

anhydride (10.58 mmoles) until the reaction was completed (the mixture became clear).

Water was added to dissolve the O-acetylated salt. Then ethyl acetate was added to wash

the aqueous solution several times. The N-acetylated and diacetylated products are in-

capable of forming salts, hence ethyl acetate can remove them from aqueous solution.

The combined aqueous solution was evaporated in a rotary evaporator to yielq.

propranolol acetate HCI (AC·HCl) as a white precipitate. The purity of the product was

established by UV and HPLC.

4.2 Preparation Of Stock Solutions

In order to conveniently prepare various concentrations of propranolol and its

prodrugs for kinetic studies and for standard solutions when using HPLC, propranolol

HCl (PRO-HCn, propranolol hemisuccinate HCI (HS-HCn and propranolol acetate HCI

(AC·Hcn were prepared in saline as stock solutions at concentrations of 6.42 mM,

6.59mM and 6.30 mM, respectively. All solutions were adjusted to pH 4.00 with phos-

phoric acid after dissolution in saline and stored at 4°C until use.

4.3 Analysis Of Propranolol And Prodrugs Of Propranolol In Plasma

Plasma samples containing various concentrations of PRO-HCI, HS-HCI and AC­

HCI were combined with 2 ml methanol to quench the enzyme activity. Quenched

samples were centrifuged at 3,000 rpm for 2 minutes. To validate the stability of

prodrugs in quenched solutions, 2 ml of methanol were spiked with 200 JlI of prodrug in

pH 7.40 phosphate buffer. The resulting solutions were stored at 4°C for 10 to 20 days,

then analyzed by HPLC. After centrifugation, the supernatant was collected. The

precipitate was rinsed twice with methanol and recentrifuged. The rinse solution was

collected and added into the supernatant, and was evaporated by PIERCE Reacti-Therm

Heating Module under nitrogen, then was reconstituted to 1 ml with 1 ml of 20%

11eOHlWater adjusted to pH 3.00 with phosphoric acid and passed through a C-18 Sep-

28

Pak. cartridge followed by an additional 1 ml of the same solvent to remove the serum

constituents from the Sep-Pak.. Approximately 5 ml of a solution containing 50%

Acetonitrile,20% Methanol, and 0.6% Triethylamine adjusted to pH 3.50 was then used

to quantitatively extract propranolol and prodrugs. Then, the 5-ml samples were

evaporated under nitrogen, and were reconstituted with HPLC mobile phase to ap­

propriate concentrations prior to analysis.

Validation of the C-18 Sep-Pak procedure was also conducted using the above

method at various concentrations of propranolol and its prodrugs in 200 J..1l buffer and

plasma.

4.4 High Performance Liquid Chromatographic Analysis

A modular high performance liquid chromatographic system consisting of an

automated sample injector (Wisp model 710A; Waters Associates, Milford, MA) operated

at 200-2000 J..1l injection volume, a constant-flow pump operated at 1.2 mllmin, a reversed

phase column packed with 5-J..1m Spheri-5 C-18 RP, a variable-wavelength UV detector

operated at 219 nm, and a digital integrator was used for all analyses.

The mobile phase was altered depending on the compound analyzed, but the mobile

phase composition was generally 45% acetonitrile in water buffered with phosphoric acid

at pH 3.50-4.00. N,N-Dimethyloctylamine (0.05%) was added as a polar modifier for the

separation of propranolol , prodrugs and the plasma components, in order to reduce tail­

ing on the chromatograph. Propranolol and its prodrugs (HS,AC) are completely

separated on the chromatograph, the retention times are 7.35 minutes for propranolol,

10.71 minutes for HS and 12.36 minutes for AC (Figure 6).

Standard solutions of propranolol and the prodrugs were prepared in mobile phase

at concentrations ranging from 7.00x10·7• 8.00xlO..s molar and stored at 4°C prior to use.

Product concentrations were determined using peak height ratios. The peak height

response versus concentration of standards was linear throughout the above concentra­

tion range.

Estimated detection limits of these compounds in plasma using the above method

and assuming a HPLC sample injection volume of 200 J..11 were: PRO-HCI 1.01x10..s M;

c E c 11 e M • al

" M . l- N a: r • I-

a: ... ,

c

E ... " • 0 f'"

t-a::

•

I,.

Figure 6: HPLC Chromatography of Propranolol (e), Propranolol Hemisuccinate (.) and Propranolol Acetate (+)

29

30

HS-HCIl.04xl0-8 M; AC-HCI1.52xl0-8 M.

4.5 Stabilities Of Propranolol And Prodrugs In Buffer Solutions

Two hundred I.Ll of each drug stock solution were added into 2 ml of pH 4.00 and

7.40 sodium phosphate buffers to investigate the stability of each drug at 37°C and/or

4°C. After dilution with buffers, final concentrations of propranolol, propranolol hemisuc­

cinate and propranolol acetate were 5.8x10-4 M, 5.99xl0-4 M and 5.73xl0-4 M, respec­

tively. Aliquots (200 ).11) were taken out at 0, 60, 120 and 360 minutes after dosing and

quenched with 2 ml methanol. The samples were diluted under the concentration ranges

of the standards and stored at 4°C prior to analysis for the remaining contents of

propranolol and its prodrugs. The degradation rates of each drug were thus determined

from semilogarithmic plots of concentration remaining versus time.

4.6 In Vitro Blood Hydrolysis Kinetics

Hydrolysis rates of prodrugs were determined in Spraque Dawley rat plasma.

Twenty-five ml of the fresh plasma was collected and stored at 4°C until use. Plasma

samples were incubated at 37°C for 15 minutes. Three samples were spiked with 4xl0·5

M ofPRO-HCI, HS-HCI and AC-HCI, respectively, in a 37°C water bath. Aliquots (200).11)

were withdrawn at timed intervals and quenched in 2 ml of pure methanol to stop the

enzyme activity. The quenched plasma samples were purified using the Sep-Pak proce­

dure described previously and diluted to the appropriate concentrations for analysis of the

remaining content of propranolol and its prodrugs in the solution. Hydrolysis rates were

determined from semilogarithmic plots of the molar concentrations of free propranolol

and prodrugs versus time.

4.7 In Vivo Pharmacokinetic Studies

To determine in vivo pharmacokinetics of propranolol and the two prodrugs, each

was administered orally to male Sprague Dawley rats weighing 250-400 gm. On day 1 of

the study, a two-piece silastic-polyethylene catheter was implanted into the right jugular

vein of each animal under ether anesthesia as described by Weeks and Davis [59]. Only the

silastic catheter portion of the catheter was inserted into the vein. For sampling, the

31

polyethylene extension of the catheter was passed subcutaneously to emerge at the dorsal

base of the neck. To avoid the use of heparin, the catheter was flushed daily with a small

volume of saline (0.5 ml) to maintain patency. Food was withheld overnight before experi­

ment. The first pharmacokinetic study was performed on day 2. Nine experimental rats

were randomly divided into three groups. Each group was treated as follows:

Group 1· PRO-HCl stock solution Group 2· HS-HCl stock solution Group 3· AC-HCl stock solution

Two hundred J..lI of blood was withdrawn from the jugular vein through the indwelling

catheter at timed intervals and immediately added into 2 ml of methanol to precipitate

blood components and thereby stop the plasma esterase activity_ Then, following the

analytical procedure discussed above, the pharmacokinetic curves were determined from

semilogarithmic plots of the molar concentrations of free propranolol and prodrugs versus

time.

4.8 Calculation of Pharmacokinetic Parameters

Plasma elimination curves of drug and prodrugs obtained after oral administration

were analyzed by least squares regression analysis. The areas under the plasma

concentration-time curves were estimated arithmetically (AUC 0-4) by the trapezoidal

rule for the observed value (t=4).

CHAPTER 5

RESULTS AND DISCUSSION

5.1 Assay Validation

The estimation of the relative rates and extent of bioconversion of propranolol

prodrugs in biological fluids required a highly specific technique to differentiate among

the free propranolol, prodrugs and the biological impurities. A pre analytical purification

step was necessarily developed for the current study, utilizing a C-18 RP Sep-Pak®

cartridge to separate propranolol and its esters from the plasma proteins and lipids.

Table 1 shows the percent recovery versus solute concentration for propranolol,

propranolol hemisuccinate and propranolol acetate from aqueous solutions. Table 2 and

Figure 7 show the percent recovery as a function of solute concentration in spiked plasma

for propranolol (PRO-HC}), propranolol hemisuccinate (HS-HCl) and propranolol acetate

(AC-HC}). As demonstrated, the percent recovery is quantitative over concentrations

ranging approximately from 9.65xlO-6 M to 7.04xlO-s M for one ml of aqueous solution or

plasma passing through the Sep-Pak, a range which includes the amounts measured in

this study. In general, less than 6% of the compounds are lost during sample preparation.

5.2 Prodrug Stability In Stock Solutions And Quenched Samples

The stock solutions of each drug were prepared in pH 4.00 phosphate buffer solu­

tion. As evident in Table 3, propranolol and its prodrugs are sufficiently stable in pH 4.00

buffer at 4°C but degrade slowly at 37°C.

Methanol was used as a quenching solution to terminate the enzyme activity for

blood samples which could not be analyzed immediately.

As represented in Figure 8, propranolol and its ester prodrugs are very stable at

4°C in the quenching solution. After a hundred hours, the remaining amount of these

compounds still exceeds 95% of the initial amount. In order to ensure that analytical data

Table 1: Validation of the SEP-PAK. Method for Propranolol and its Prodrugs in Aqueous Solution

Concentration Sample -6 -6 Recovery

x10 M x10 glml

Propranolol 16.20 4.79 98.30 B.48 2.51 91.94 B.11 2.40 95.40 7.69 2.27 101.50 7.16 2.11 102.00 1.84 0.54 99.80 1.62 0.48 95.90 0.76 0.23 106.10

Mean :!: SEH 98.45 :!: 0.59

Pr opr anolol 13.40 5.30 98.70 Hemisuccinate 7.11 2.81 98.70

6.70 2.65 93.40 3.24 1.28 99.20 2.74 1.08 96.50 0.67 0.27 102.30

Mean :!: SEJ-1 98.15 :!: 0.49

Propranolol 3.8~ 1.30 98.10 Acetate 1.93 0.65 98.40

0.48 0.16 95.80 1-1ean :!: SE}1 97.44 :!: 0.49

33

%

Table 2: Recovery of Propranolol in Spiked Plasma after SEP-PAK Purification

Concentration Sample -6 -6 Recovery

xl0 M xl0 g/ml

Propranolol 7.69 2.27 101.50 1.84 0.54 99.80 0.76 0.22 106.10 0.36 0.11 100.00 0.18 0.05 92.90 0.09 0.03 94.30

Mean! SE11 99.10 ! 0.81

Propranolol 7.11 2.81 101.00 Hemisuccinate 3.55 1.41 96.20

1.78 0.70 96.90 0.71 0.28 88.90 0.36 0.14 92.90 0.07 0.03 101.80

Mean ! SEM 97.10 ! 0.98

Propr anolol 9.65 3.26 99.10 Acetate 4.83 1.63 93.90

2.41 0.81 93.90 0.96 0.33 90.10 0.48 0.16 94.70

Mean :!: SEM 94.30 :!: 0.74

34

"

35

100 - • • • ·-t • t - • • .t t t .0

• • 0

H

.. 70 >. .... <U > .0 0 U Q)

c:x: 11.1 ISO en ~

.. 0

:10

.0

10

CI D.01 D.10 "U:IO

-6 Solute Concentration, 10 ,/ml

Figure 7: Recovery of Propranolol (.), Propranolol Hemisuccinate (_) and Propranolol Acetate (.) in 200~1 Spiked Plasma

after C·18 RP SEP·PAK Purification

Time

0

60

120

360

Time

0

11

20

Table 3: The Hydrolysis of PRO-HC], HS-HCI and AC-HCI in PH 4.00 Phosphate Buffer at 37°C and 4°C

0 Remainin2 % (min.) At 37 C PRO HS AC

100.00 100.00 100.00

98.57 98.97 96.97

96.20 97.94 90.91

95.72 92.78 84.85

0

(day) At 4 C

100.00 100.00 100.00

96.40 97.99 98.42

98.09 97.04 97.10

36

100

80

80

70

N eo .. >. ... QJ SO > 0 (,) Q)

40 ~

30

20

10

0

I. • • -t-. e! • + • + • •

10 20 30 40 150 eo 70 so so 100

Time, hour

Figure 8: Validation of the Stability of Propranolol (e), Propranolol Hemisuccinate (.) and Propranolol Acetate (. ) in

Quenching Solution at 4°C

37

38

would not be biased significantly as a result of in vitro hydrolysis, plasma samples were

processed and analyzed promptly in all experiments.

5.8 Prodrug Bioconversion At pH 7.40 In Buffer And In Plasma

Kinetic studies in dilute aqueous solutions (7x10""M) were conducted at pH 7.40 in

a 37°C water bath to obtain estimates of the stability of propranolol and its ester prodrugs

under these conditions. Figure 9 clearly shows that propranolol is stable under these

conditions but both the propranolol hemisuccinate and acetate undergo first-order

hydrolysis with half-lives of 281 minutes, 84 minutes, respectively.

Due to the fact that propranolol prodrugs are more stable in pH 4.00 buffer than in

pH 7.40 buffer, the ester prodrugs appear to be catalyzed by hydroxide ion (OH-) in water.

AC-HCl appears to be the most sensitive compound to this base-catalyzed hydrolysis reac­

tion. The data are not sufficient to provide information on the pH-hydrolysis rate profile

and the mechanism of the hydrolysis reaction is still unknown.

To gain more information on the influence of pro-moiety structure on the relative

rates of prodrug bioconversion, the hydrolysis of the ester prodrugs were also determined

in rat plasma at a concentration of 4.00x1 O-sM. Shown in Figure 10 are semilogarithmic

plots of the concentration of propranolol and its prodrugs remaining at a given time ver­

sus time in rat plasma. Both HS and AC appear to undergo first-order hydrolysis in

plasma samples. First-order hydrolysis rate constants for each compound were obtained

from the slopes of such plots (Figure 10). The half-lives of HS and AC are calculated to be

35 minutes and 9 minutes, respectively. Propranolol is not degraded to any significant

extent in rat plasma.

The ester prodrugs of propranolol are hydrolyzed much more rapidly in plasma

samples than in aqueous buffer at the same pH - possibly by interaction with nonspecific

carboxylesterases.

5.4 In Vivo Pharmacokinetic Studies

The oral bioavailabilities of the prodrugs were compared with propranolol in

Sprague Dawley rats. The relationship between time and plasma levels of propranolol

e ,a~~~-.----__ ------------------____ __

150

lIa

1a ~--------------------------------------~---.a co Ba .0 10a ,ac 140 1Sa 180 lIao Time, minute

Figure 9: Semilogarithmic Plots of the Concentrations of Propranolol (e), Propranolol Hemisuccinate (+) and Propranolol Acetate <*)

Versus Time in PH 7.40 Phosphate Buffer at 37°C

39

1Di~~ __ ~ ____________________________________________ _

10 aD 30 40 150

Time, minute eo 70

Figure 10: Semilogarithmic Plots of In-vitro Hydrolysis of Propranolol (e), Propranolol Hemisuccinate (.) and

Propranolol Acetate (.) in Rat Plasma at 37°C

110

40

41

was determined in each of three rats after 10 mglkg oral doses of propranolol

hydrochloride and equivalent doses of the hemisuccinate and acetate esters. The

propranolol blood concentrations versus time are shown in Figure 11 for propranolol, the

hemisuccinate and the acetate esters. Plasma propranolol concentrations are con­

siderably higher following administration of the prodrugs (Tables 4, 5, 6).

Blood concentration (C)-time curves for each animal were analyzed by least-squares

regression (RSTRIP,Micromath Inc., SLC, UT) using the fonowing one-compartment

equation to reflect simple first-order absorption and elimination processes.

C = A(e-Kt -e-KGt)[lJ

The pharmacokinetic parameters obtained are shown in Table 7. Efforts were made

to employ a two-compartment equation, but the improvement in fit did not warrant the

inclusion of additional parameters.

Absorption of the prodrugs and their conversion to propranolol are rapid, with peak

plasma levels of propranolol having been attained at 0.67 hours after administration of

both prodrugs and propranolol (Tables 4, 5, 6 & Figure 11). Since blood sampling was

terminated at four hours after dosing, only a few points are available for calculating the

half-lives. Based on these values, the disappearance half-life of propranolol is 69 minutes

after oral PRO-HCI, which is very similar to Iwamoto's results (45-55 minutes) and

Bianchetti's data (63 minutes) [5]. The disappearance half-life of plasma propranolol is 42

minutes after oral HS-HCI and 38 minutes after oral AC-HCI administration according to

the one-compartment elimination equation (Table 7). The elimination half-life of

propranolol was therefore significantly altered after prodrug administration. The reason

is probably that drug concentrations in the liver are sufficiently high after oral

propranolol to saturate metabolic enzyme systems, leading to reduced hepatic extraction

from the systemic circulation. On the other hand, the liver is not saturated by the plasma

propranolol after oral propranolol prodrugs, resulting in a faster elimination rate of the

plasma propranolol from the systemic circulation. Consistent with this interpretation,

the mean hepatic extraction ratio of 14C-propranolol given intravenously, has been

reported to be remarkably decreased after pretreatment with unlabelled propranolol or

...... e ....... CO

"" I 0 -.,. "'"" 0 ...... 0 C CO $of Q. 0 $of

&:I..

CO e ct)

CO

"'"" ~

R,4

.,2

_,C

1,B

1.8

1.4

1,.

1.C

C,B

C.-

C.4

C,.

C.C 1 • 3

Time, hour

Figure 11: Plasma Concentrations of Propranolol in Sprague Dawley Rats after Oral Administration of Propranolol ce), Propranolol

Hemisuccinate C.) and Propranolol Acetate C+) at Dose 10 mg/kg Cn=3)

42

Time

0.00

0.12

0.33

0.67

1.50

2.50

4.00

Table 4: Plasma Concentrations of Propranolol (Mean±SEM) in Sprague Dawley Rats after Oral Administration of

Propranolol HCl (n=3) at a Dose of 10 mg/kg

-6 Plasma Propranolol Concentration (x10 g/mI)

(hr.) Mean:tSEM

0.00 0.00 0.00 0.00:t0.00

0.30 0.38 0.33 0.34:t0.02

0.59 0.78 0.64 0.67:t0.06

1.10 1.33 1.06 1.16:t0.08

0.57 0.47 0.47 0.50:t0.03

0.29 0.32 0.24 0.28:t0.02

0.15 0.19 0.09 0.14:t0.03

43

44

Table 5: Plasma Concentrations of Propranolol (Mean±SEM) in Sprague Dawley Rats after Oral Administration of Propranolol

Hemisuccinate HCI (n=3) (Dose = 10 mWlcg Propranolol HCI Equivalents)

-6 Plasma Propranolol Concentration (xl0 g/ml)

Time (hr.) Mean:!:SEM

0.00 0.00 0.00 0.00 O.OO:!:O.OO

0.12 0.77 0.56 0.77 0.70:!:0.07

0.33 1.83 1.48 1.80 1.70:!:0.11

0.67 2.74 2.06 2.65 2.48:!:0.21

1.50 1.77 1.30 1.73 1.60:!:0.15

2.50 0.76 0.71 0.74:!:0.03

4.00 0.61 0.19 0.40:!:0.21

Table 6: Plasma Concentrations of Propranolol (Mean±SEM) in Sprague Dawley Rats after Oral Administration of Propranolol Acetate

HCl (n=3) (Dose = 10 mg/kg Propranolol HCl Equivalents)

-6 Plasma Propranolol Concentration (xlO g/ml)

Time (hr.) MeantSEM

0.00 0.00 0.00 0.00 O.OOtO.OO

0.12 1.15 0.58 1.41 1.05tO.25

0.33 1.38 1.43 1.57 1.46tO.06

0.67 2.94 2.21 2.24 2.46±0.24

1.50 2.06 1.28 2.05 1.80tO.26

2.50 0.81 0.69 0.87 0.79tO.05

4.00 0.08 0.36 0.22tO.11

45

Table 7: Pharmacokinetic Parameters Obtained from Least-Squares Regression of the Propranolol Blood Concentrations Resulting from Oral

Administration of Propranolol and its Acetate and . Succinate Esters According to Equation 1

A UC (hr ~g/mI)

46

Rat Compound K (hr- 1) ka (hr- 1) o - 4 hrs.a 0 _CX:!b

1 PRO-HCl 0.585 2.86 1.85 1.94 2 PRO-HCl 0.477 4.52 2.03 2.14 3 PRO-HC1 0.726 3.05 1.65 1.64

Average±SD 0.60t.12 3.48t.91 1.84t.19 1.91t.25

4 HS-HCI 0.496 2.41 5.24 6.98 5 HS-HCl 1.124 1.30 3.99 4.08 6 HS-HCI 1.318 1.35 4.94 4.67

AveragetSD 0.98t.43 1.69±.63 4.72t.65 5.24t1.53

7 AC-HCI 1.25 1.30 5.35 5.10 8 AC-HC1 1.41 1.44 3.88 3.39 9 AC-HC1 0.64 2.80 5.21 5.57

AveragetSD 1.10±.41 1.85t.83 4.81t.81 4.69±1.1S

aFrom trapezoidal integration.

bFrom best fit according to Equation II].

47

during portal venous administration of unlabelled propranolol [46].

The apparent absorption rate constant, Ka (Table 7) based on propranolol ap­

pearance in the blood, was significantly reduced after prodrug administration. Either

slower absorption of the prodrugs compared to propranolol, or relatively slow (rate­

limiting) bioconversion after absorption, could account for this.

Estimates of the mean area under the plasma level-time curves (AUC) obtained

from trapezoidal integration of the data between 0 and four hours are listed in Table 7.

The overall bioavailability of the prodrugs, based on the AUC, is significantly higher than

that of propranolol. Following oral administration of HS-HCI and AC-HCI, the AUC of

plasma propranolol curves are 2.63 and 2.75 times higher, respectively, than after an oral

dose of PRO-HCl. The AUC (0-4 hours) is 4.84 ± 0.25 J.Lg/ml x hr. after HS-HCI, 5.07 ±

0.37 Jlg/ml x hr. after AC-HCI and 1.84 ± 0.06 Jlg/ml after PRO-HCl. In Iwamoto's paper

[25], the oral bioavailability after a 10 mg/kg oral dose of propranolol in male Wistar rats

is 0.25. If the same value is assumed for the oral bioavailability of propranolol in the

present study, the relative oral bioavailabilities of HS-HCI and AC-HCI are estimated to

be 0.66 and 0.69, respectively.

The appearance and disappearance of the intact prodrugs from plasma were also

monitored. Again the data were analyzed by least-squares regression using a simple

biexponential equation(Eq 1). The levels of intact prodrugs in plasma at various times

are shown in Figures 12 and 13. Following oral administration, unchanged prodrug levels

peak at 0.67 hours after dosing, and are negligible at four hours after dosing (Tables 8

and 9), while the plasma propranolol levels also peak at 0.67 hours but are still substan­

tial four hours after dosing. The half-lives of plasma hemisuccinate and acetate dis­

appearance are 25 and 22 minutes (Table 10), respectively. These results suggest rapid

absorption and conversion of the prodrugs. The relatively high concentrations of the

prodrugs in the blood and higher values of their absorption rate constants compared to

their elimination rate constants coupled with the comparable magnitudes of the prodrug

elimination rate constants and the propranolol formation constants suggest that prodrug

bioconversion, rather than absorption, governs the rate of propranolol appearance in the

\C) I

~ 3,D & ...... DO

C .... _.D .. cu 4J CU c::

.-4 (J (J

:I fI)

.-4 e

1.D C1.I :::: ~ 0 ~ 0 c:: C'C ,. Q.. 0 ,.

r::I.o

C'C e fI)

cu ....., r::I.o

D .• ~------~--------~--____ -L ______ ~ __ _ , 3

Time, hour

Figure 12: Plasma Concentrations of Propranolol Hemisuccinate after Oral Administration of Propranolol Hemisuccinate

HCl at Dose 10 mglkg(n=3)

48

\0 I

i 8.0 ........

DO

0 -.. CIJ

""" C\1

""" CIJ

~ ~ 0 ~ 0 c C\1 a-Q. 0 a-

Q.;

C\1 e CfJ C\1 ~ Q.;

'.0

, • 3 Time. hour

Figure 13: Plasma Concentrations of Propranolol Acetate after Oral Administration of Propranolol Acetatee

at Dose 10 mg!kg(n=3)

49

Table 8: Plasma Concentrations of Propranolol Hemisuccinate ( in Propranolol Equivalent Unit) after Oral Administration or Propranolol Hemisuccinate

Hel (Dose = 10 mglkg Propranolol He} Equivalents)

-6 Time(hr.) HS-HCI Plasma Concentration (xlO g/ml) Mean:!:SEM

(Propranolol Equivalent Concentration)

0.00 0.00 0.00 0.00 0.00:t0.00

0.12 1.35 1.13 1.63 1.37:t0.14

0.33 2.39 1.40 1.94 1.91±0.29

0.67 2.86 1.70 2.43 2.33±0.34

1.50 0.88 0.52 0.72 0.71±0.10

2.50 0.23 0.25 0.24±0.01

4.00 0.05 0.05:t0.00

50

Table 9: Plasma Concentrations of Propranolol Acetate (in Propranolol Equivalent Unit) after Oral Administration of Propranolol Acetate

HCI (Dose = 10 mgJkg Propranolol HCI Equivalents)

-6

51

Time (hr.) AC-HCl Plasma Concentration(xlO g/ml) Mean:SEM (Propranolol Equivalent Concentration)

0.00 0.00 0.00 0.00 0.00:0.00

0.12 2.30 2.34 2.1S 2.27:0.05

0.33 3.17 3.2S 3.37 3.27:0.06

0.67 3.S9 3.S4 3.96 3.90:t0.03

1.50 '1.27 1.21 1.lS 1.22:t0.03

2.50 0.53 0.37 0.45:t0.08

4.00 0.10 0.10 0.10:t0.00

Table 10: Apparent Absorption and Elimination Rate Constants Obtained from Least-Squares Regression of the Propranolol Hemisuceinate and

Acetate Blood Concentrations after Oral Administration According to Equation 1

Rat Compound K (hr- 1) ka (hr- 1)

4 HS-HCl 1.92 2.38 5 HS-HCI 0.99 6.34 6 HS-HCl 1.15 5.37

Average±SD 1.35±.50 4.70±2.06

7 AC-HCl 1.06 4.84 8 AC-HCl 1.07 5.67 9 AC-HCl 2.21 2.27

Average±SD 1.45±.41 4.26±1.77

52

53

blood.

The mechanism by which prodrugs of propranolol produce significantly higher

bioavailabilities of propranolol is unknown. The simplest rationalization is that the

prodrugs themselves are not metabolized in the first-pass through the liver and sub­

sequently undergo hydrolysis in the systemic circulation to propranolol. In vitro studies in

Lltese laboratories of the hydrolysis of HS-HCI and AC-HCI in buffer and in rat plasma

(pH 7.4, 37°C) establish that propranolol ester hydrolysis is catalyzed in plasma. The

half-life values were 281 min and 35 min for HS-HCI and 84 min and 9 min for AC-HCI in

buffer and in plasma, respectively. Thus, hydrolysis rates in plasma were sufficiently

rapid for hydrolysis in the systemic circulation to contribute to the overall rates of ester

hydrolysis in vivo.

Further experiments to determine the mechanism by which prodrugs successfully

avoid metabolism in their first-pass through the liver and the extent to which this

favorable property can be optimized via changes in chemical structure are planned.

CHAPTER 6

CONCLUSIONS

In previous studies propranolol has been shown to undergo extensive presystemic

metabolism after oral administration, leading to reduced bioavailability and significantly

greater intersubject variability in blood levels after oral than intravenous administration.

Garceau et al. demonstrated that the hemisuccinate ester of propranolol, when adminis­

tered orally to beagle dogs, yields propranolol levels eight times higher than after an

equivalent dose of propranolol hydrochloride [16], suggesting that the prodrug approach

may be an effective means of avoiding first-pass metabolism of drugs which undergo ex­

tensive first-pass elimination. The mechanism by which prodrugs might lead to

decreased first-pass metabolism is unknown, however.

In the current study, the plasma propranolol AUC's of HS-HCl and AC-HCI were

found to be 2.6 times and 2.7 times higher, respectively, than that of propranolol after

oral administration in rats. The data confirm that O-acyl ester prodrugs of propranolol

are suitable as model compounds for mechanistic studies focussing on the use of the

prodrug approach to bypass first-pass metabolism by the liver. The Spraque Dawley rat

has also been identified as a suitable animal model for such studies.

Given the probable difference in lipophilicity between the succinate, which is zwit­

terionic at physiological pH, and the acetate ester, which, though largely protonated at

pH 7.4 would be expected to be more lipophilic than propranolol, one hypothesis tested

was that the zwitterionic succinate exhibits higher bioavailability due to reduced uptake

into hepatocytes. However, since both the acetate and succinate esters were similar in

bioavailability, both being higher than propranolol, lipophilicity alone may not be the

determining factor in these results.

REFERENCES

1. AB.Ahmad, P.N.Bennett, M.Rowland, J. Pharm. Pharmacol., Vol. 35, 1983, pp. 219-224.

2. J.H.Anderson, R.C.Anderson and L.S.lben, J.Pharmacol.Exp.Ther., Vol. 206, 1978, pp. 172-180.

3. L.Bass, R.Robinson, A.J.Bracken, J.Theor.Biol., Vol. 72,1978, pp. 161-163.

4. C.Von Bahr, J.Hermansson and M.Line, J.Pharmacol.Exp.Ther., Vol. 222, No. 2, 1982, pp. 458-462.

5. G.Bianchetti, J.L.Elghozi, R.Momeni, P.Meyer,P.L.Morselli, J. Pharm. Exp. Ther., Vol. 214, No. ,1980, pp. 682-687.

6. B.L.Blitzer, J.L.Boyer, Gastroeuterology, Vol. 82, 1982, pp. 346-357.

7. N.Bodor, KB.Sloan, T.Higuchi, J.Med.Chem., Vol. 20(11), 1977, pp. 1435-1445.

8. J.L.Campra and T.B.Reynolds, in The Liver:Biol.& Path., Raven Press, N.Y., 1982, pp. 627-645, Ch. 37.

9. G.M.Cohem and D.S.Davies, Br.J.Pharmacol., Vol. 50, No.3, 1974, pp. 472p.

10. S.S.Davis and G.Elson, J.Pharm.Pharmacol., Vol. 26, 1974, pp. 90p.

11. H.W.Deane, Anat.Rec., Vol. 88,1944, pp. 39-65.

12. M.R.Dobrinska, W.Kukovetz, E.Bublerand and H.L.Leidy et.al., Biopharm., Vol. 1 0, No.6, 1982, pp. 587-600.

J.Pharm.&

13. O.H.Drummer, J.McNeil, E.Pritchard, W.J.Louis, J.Of Pharm.Sci., Vol. 70, No. 9, 1981, pp. 1030-1032.

14. G.H.Evans, G.R.Wilkinson, D.G.Shand, J.Pharmacol & Exp.Ther., Vol. 186(3), 1973, pp. 447-454.

15. E.L.Forcker, B.ALuxon, Am.J.Physiol., Vol. 234, 1982, pp. G532-540.

16. Y.Garceau,I.Davis and J.Hasegawa, J.Pharm.Sci., Vol. 67, No. 10, 1978, pp. 1360-1363.

17. J.R.Gillette, Ann.N.YAcad.Sci., Vol. 179, 1979, pp. 43.

18. M.Gibaldi, R.N.Boyes and S.Feldman, 1338-1340.

J.Pharm.Sci., Vol. 60, 1970, pp.

19. A.G.Gilman, L.S.Goodman, T.W.Rall,F.Murad, in Goodman and Gilman's The Pharmocological Basis Of Therapeutics, Macmillan Publishing Co., 1985, pp. 190-195, Ch. 9.

20. C.AGoresky, in The Liver:Biol.& Path., Raven Press, N.Y., 1982, pp. 581-599, Ch.34.

21. W.G.Guder and U.Schmidt, Hoppe Seylers Z.Physiol.Chem., Vol. 357, 1976, pp. 1793-1800.

22. J.J.Gumucio and D.L.Miller, in The Liver:Biol.& Path., Raven Press, N.Y.,

56

1982, pp. 647-661, Ch. 38.

23. A.Hayes and R.G.Cooper, 302-311.

J.Pharmacol.Exp.Ther., Vol. 176, No.2, 1971, pp.

24. L.Herbette, AM.Katz and J.M.Sturtevant, 1983, pp. 259-269.

Molecular Pharmacol., Vol. 24,

25. Klwamoto and J.Watanabe, J.Pharm.Pharmacol., Vol. 37,1985, pp. 826-828.

26. S.Johnson, S.Keiding, J.Theor.Biol., Vol. 89,1981, pp. 549-556.

27. D.B.Jones, D.J.Morgan, G.W.Mihaly, L.K Webster, R.ASmallwood, J.Pharmacol & Exp.Ther., Vol. 229(2),1984, pp. 522-526.

28. S.Keiding, S.Johansen, LMidtboll, .4m.J.Physiol., Vol. 237, 1979, pp. 316-324.

29. S.Keiding, E.Chiarantini, J.Pharmacol & Exp.Ther., Vol. 205, 1978, pp. 465-470.

30. M.W.Lo, D.J.Effeney, S.M.Pono, B.M.Silber and S.Riegelman, J.Pharmacol.Exp.Ther., Vol. 221, No.2, 1982, pp. 512-515.

31.

32.

33.

34.

B.A.Luxon, E.L.Forker, Federation Proc., Vol. 43, 1984, pp.161.

T.Matsumura, R.G.Thurman, Am.J.Physiol., Vol. 244, 1983, pp. 656-659.

J.McEwen, D.G.Shand and G.R.Wilkinson, 1974,pp.14p-141p.

D.KF.Meijer, C.Neef and G.M.M.Groothuis, pp. 167-189.

Br.J.Pharmacl., Vol. 52, No.1,

Topics In Pharm.Sci., Vol. ,1983,

35. W.L.Nelson and R.B.Walker, Chem.Path.Pharmacol., Vol. 22, No.3, 1978, pp. 435-445.

36. AB.Novikoff, J.Histochem.Cytochem., Vol. 7,1959, pp. 240-244.

37. AB.Novikoffand E.Essner, Am.J.Med., Vol. 29,1960, pp.102-131.

38. H.Ong,P.Suich and C.Marchand, Drug Metabolism & Disposition, Vol. 9, No.6, 1981, pp. 529-534.

39. KS.Pang, M.Rowland, J.Pharmacol. & Biopharm., Vol. 5(6),1977, pp. 625-650.

40. KS.Pang, M.Rowland, J.Pharmacol. & Biopharm., Vol. 5(6),1977, pp. 65-680.

41.

42.

V.Schmidt, H.Schmidt, W.G.Guder, 1978, pp. 193-198.

Hoppe Seylers Z.Physiol.Chem., Vol. 359,

D.G.Shand, E.M.Nuckolls and J.AOates, pp.112-120.

Clin.Pharmacol.Ther., Vol. 11, 1970,

43. D.G.Shand, G.H.Evens and A.S.Nies Life Sci., Vol. 10,1971, pp.1417-1421.

44. D.G.Shand and R.E.Rangno, Pharmacol, Vol. 7,1972, pp. 159-168.

45. D.G.Shand and R.E.Rangno and G.H.Evans, Pharmacol, Vol. 8(4-6), 1972, pp. 344-352.

46. T. Suzuki, S. Isozaki, T. Ohkuma, T. Rikihisa, J.Pharm. Dyn., Vol. 3,1980, pp. 603-611.

47. S.Vickers, C.A.Duncan, S.D.White and G.O.Breault et aI., J.Pharmacol. & Exp. Ther., Vol. 6(6),1978, pp. 640-645.

48. S.Vickers, C.A.Duncan, H.G.Ramjit and M.R.Dobrinska et aI., J.Pharmacol. & Exp. Ther., Vol. 12(2),1983, pp. 241-246.

49. C.Von Bahr, E.Fellenius, T.Fried, Suppl.I.92.

57

Acta.Pharmacol. Toxicol., Vol. 31, 1972, pp.

50. C.Von Bahr, H,Vadi, ANorlin, P.S.Moldeus, Acta. Pharmacol.Toxicol., Vol. 38, 1976, pp. 513-524.

5l.

52.

AWachstein, E.Meisel, Am.J.Clin.Phthol., Vol. 27,1957, pp.13-23.

T.Walle, T.C.Fagan, E.C.Coradi, U.K. Walle and T.E.Gaffuey, Ther., Vol. 26,1979, pp. 167-172.

Clin. Pharmacol.

53. T.Walle, E.C.Conradi, U.K. Walle, T.C.Fagan, T.E.Gaffuey, Clin.Pharmacol.Ther., Vol. 27, 1980, pp. 23-31.

54. T.Walle, U.K. Walle, n.R.Knapp, E.C.Conradi and E.M.Bargai-, Drug Metaol. & Dispos., Vol. 11(4), 1983, pp. 344-349.

55. T.Walle, U.K.Walle, L.S.Olanoff, Drug Metabol. & Dispos., Vol. 13(1), 1984, pp. 204-209.

56. T.Walle, Drug Metaol. & Dispos., Vol. 13(3), 1984, pp. 279-282.

57. T.Walle, U.K. Walle, M.J.Wilson, T.C.Fagan, T.E.Gaffuey, Clin. Pharmacol., Vol. 18,1984, pp. 741-747.

58. J.Wanger, H.Grill, D.Henschler, 1423-1427.

J.Pharma.Sci., Vol. 69(12), 1980, pp.

59. J.R.Weeks, J.D.Davis, J. Appl. Physiol., Vol. 19,1964, pp. 540-54l.

60. R.AWeisger, C.M.Mendel, R.R.Cavalier, J.Pharm.Sci., Vol. 75(3), 1986, pp. 233-237.

61. G.R.Wilkinson, D.G.Shand, Clin.Pharmacol.Ther., Vol. 18, 1975, pp. 377.