DRUG-DRUG
`
`INTERACTIONS:
`
`SCIENTIFIC AND
`
`REGULATORY
`
`PERSPECTIVES
`
`Albert P. Li
`In Vitro Technologies Inc.
`University of Ivlarylan
`echnology Center
`"more, Maryland
`
`’
`
`ACADEMIC PRESS
`
`San Diego London Boston New York Sydney Tokyo Toronto
`
`Vanda Exhibit 2043 - Page 1
`
`Vanda Exhibit 2043 - Page 1
`
`

`
`This book is printed on acid—free paper.
`
`Copyright © 1997 by ACADEMIC PRESS
`
`All Rights Reserved.
`No part of this publication may be reproduced or transmitted in any form or by any
`means, electronic or mechanical, including photocopy, recording. or any information
`storage and retrieval system. without permission in writing from the Publisher.
`The appearance of the code at the bottom of the first page of a chapter in this book
`indicates the Publisher‘s consent that copies of the chapter may be made for
`personal or internal use of specific clients. This consent is given on the condition.
`however, that the copier pay the stated per copy fee through the Copyright Clearance
`Center, Inc. (222 Rosewood Drive. Danvers. Massachusetts 01923), for copying
`beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent
`does not extend to other kinds of copying, such as copying for general distribution. for
`advertising or promotional purposes. for creating new collective works, or for resale.
`Copy fees for pre-1997 chapters are as shown on the title pages. if no fee code
`appears on the title page. the copy fee is the same as for current chapters.
`1054-3589/97 $25.00
`
`Academic Press
`a division 0fHarcom'I Bl'(1(‘(’ & C(}H1[J(lI1V\'
`525 B Street. Suite 1900. San Diego. California 92101—4495. USA
`http://\v\v\v.apnet.com
`
`Academic Press Limited
`24-28 Oval Road. London NW1 7DX. UK
`http://\v\v\v.hbuk.co.uk/ap/
`
`International Standard Book Number: 0-12-03L’94'~1—1
`
`Vanda Exhibit 2043 - Page 2
`
`

`
`Vanda Exhibit 2043 - Page 3
`
`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`

`
`F. Peter Guengerich
`
`such as lung and small intestine, where an appreciable contribution to overall
`metabolism of a drug can occur depending on the route of administration.
`Fortunately, it appears that the metabolism of most drugs can be ac-
`counted for by a relatively small subset of the P4505. One estimate is that
`290% of human drug oxidation can be attributed to six enzymes: P4505
`1A2, 2C9/10, ZC19, ZD6, 2E1, and 3A4 (Guengerich, 1995; Wrigliton and
`Stevens, 1992). Further, most could probably be attributed to P4505 1A2,
`2C9/10, ZD6, and 3A4 and, by some estimates, half can be attributed to
`P450 3A4 (Guengerich, 1995; Guengerich et £11., 1994a). This view is based
`primarily on (in uitro) microsomal studies done with drugs studied to date
`and may change somewhat with time. For instance, the fraction of drugs
`oxidized by P450 ZD6 may be too high in current estimates because of the
`ease of identifying these and the attention that has been given to this particu-
`lar enzyme. Nevertheless, the concept that most drug oxidations are cata-
`lyzed primarily by a small number of P450 enzymes is important in that
`the approaches to identifying drug—drug interactions are feasible, both in
`z/itro and in viva.
`I
`This chapter operates from the premise that many significant drug—drug
`interactions can be understood in terms of P4505. However, drug—drug
`interactions are more complex for at least two reasons. First, some drug-
`drug interactions can be attributed to pharmacokinetic differences due to
`other enzymes such as monoamine oxidases, flavin-containing monooxygen—
`ases, UDP—glucuronosyl transferases, and sulfotransferases. These and other
`so—called “drug-metabolizing” (or “xenobiotic—metabolizing”) enzymes also
`show the characteristics of induction and inhibition by drugs that are associ-
`ated with P4505, although most have not yet been studied as extensively.
`The other aspect of drug—drug interactions is that some of these are probably
`pharmacodynamic instead of pharmacokinetic. For instance, drugs can com-
`pete for binding to a receptor directly related to the pharmacological re-
`sponse.
`
`ll. Potential Consequences of Drug—Drug Interactions
`
`the major effects of drug—drug
`From a pharmacokinetic standpoint,
`interactions can be understood in terms of causing the disposition of a durg
`to be unusually slow or fast. The major consequence is a high or low plasma
`and tissue level of the drug.
`If the metabolism of a drug is impeded due to enzyme inhibition, then
`a high plasma level may follow (Fig. 1). One of the major effects will be
`increased pharmacological activity, and this may or may not be a problem,
`depending on the therapeutic window. Of course, not only the desired effect
`
`Vanda Exhibit 2043 - Page 4
`
`

`
`Plasma
`level of
`
`drug
`
`FIGURE I
`
`Time (arrows show repeated doses)
`Effect of enzyme inhibition on drug metabolism and plasma drug levels.
`
`be anticipated. Another possibility is that when the major pathway of metab-
`olism of a drug is blocked, secondary pathways may become more favorable.
`This can be a problem if the secondary pathway leads to a toxic product.
`An example of this is seen with the analgesic phenacetin (no longer on the
`U.S. market). If O-deethylation (P450 1A2) is slow, then other pathways
`are favored that lead to quinoneimine formation and methemoglobinemia
`(Fischbach and Lenk, 1985; Klehr et ((1., 1987). Another possibility is that
`the increased level of a drug due to inhibition of the P450 involved in its
`oxidation may lead to inhibition of another P450. Although direct evidence
`for such a situation has not been presented, one could postulate that accumu-
`lation of quinidine due to P450 3A4 inhibition might lead to inhibition of
`P450 ZD6, an enzyme for which quinidine is an inhibitor but not a substrate
`(Guengerich er £11., 1986b; Otton at al., 1984).
`When levels of P450 (or, for that matter, another enzyme) are induced,
`the major consequence is a lack of therapeutic effectiveness. Although this
`might seem to be a common event, the number of real clinical situations in
`which this has been a problem are rather limited. Two of the best documented
`examples are cyclosporin and 17oz-ethynylestradiol (uide infra). Another
`possibility with a pro-drug is that activation may be too rapid and a seriously
`high level of active drug could result. This could be a problem, as one of
`the primary reasons for developing pro—drugs is to avoid a transiently high
`level of the active drug. However, no good examples of clinical problems
`resulting from a phenomenon of this type are known yet.
`There are two other possibilities that can be considered in regard to
`issues of drug—drug interactions. One involves metabolism of chemical car-
`cinogens. Most of the P4505 that transform drugs can also oxidize chemical
`carcinogens (Guengerich and Shimada, 1991; Guengerich, 1995). The possi-
`bility exists that a P450 induced by a drug could lead to enhanced levels
`
`Vanda Exhibit 2043 - Page 5
`
`Vanda Exhibit 2043 - Page 5
`
`

`
`F. Peter Guengerich
`
`of DNA—carcinogen adducts due to increased carcinogen activation. The
`induction of human P450s 1A1 and 1A2 by omeprazole was postulated to
`present a risk due to such considerations (Diaz er a/., 1990). Whether or
`not this is a serious issue is unknown, as levels of P450 1A2 are only one
`of many factors linked to cancer risk from known carcinogenic substrates
`for the enzyme (Lang et a1., 1994). Nevertheless, most pharmaceutical com-
`panies and the Food and Drug Administration (FDA) would rather avoid
`drugs that induce P450 1A subfamily enzymes, which have been suggested
`to be related to cancer development (Ioannides and Parke, 1990, 1993).
`Again, it should be emphasized that increased cancer risk due to P450 (1A
`or other) induction is still a hypothesis. Another matter to consider is that
`some P4505 are involved in the detoxication of potential carcinogens and
`that induction or inhibition might have an impact on this process, as well
`as the activation (Guengerich and Shimada, 1991; Richardson et a1., 1952;
`Nebert, 1989).
`The other possibility involves the influence of drugs on P450s involved
`in the transformation of endogenous compounds, i.e., those normally found
`in the body. This matter has not been extensively investigated, but some
`possibilities exist. Depending on the tissue, 17B—estradiol
`is oxidized by
`P450 3A4 (Brian et 511., 1990; Guengerich et ((1., 1986a), P450 1A2 (Guo
`et (11,, 1994), or P450 1B1 (Liehr et a1., 1995). It is not known what the
`impact of changes in these enzymes is on total body levels. Testosterone is
`a substrate for P450 3A4 (Guengerich et £11., 1986a; Waxman et (11,, 1988).
`Another case to consider is animals devoid of bilirubin UDP-glucuronosyl
`transferase activity, who can be administered P450 1A inducers to lower
`their levels of bilirubin to nontoxic levels (Kapitulnik and Gonzalez, 1993;
`Kapitulnik and Ostrow, 1978).
`A final point to consider is that some drugs may affect the disposition
`of chemicals in foods and beverages through P450 interactions. For instance,
`the drug disulfiram (Antabuse) inhibits P450 2E1. This would affect ethanol
`oxidation by P450 2E1, although the more serious effect is on aldehyde
`dehydrogenase (Guengerich et ((1., 1991). The P450 1A2 inhibitor furafylline
`blocks caffeine N3-demethylation to the point where severe insomnia is
`associated with drinking coffee (Sesardic er ((1., 1990; Kunze and Trager,
`1993).
`
`III. Use of Information about Human P450s
`
`Much of the information abot drug metabolism by human P450s has
`been acquired in the past decade. Induction was first seen in clinical settings
`in the 1950s (Remmer, 1959) and there were many in I/itro studies with
`
`Vanda Exhibit 2043 - Page 6
`
`

`
`and CDNA cloning methods were used to obtain DNA sequences for the
`human P4505 (Gonzalez, 1989).
`In recent years the access to human tissue samples in the United States
`and Europe has facilitated characterization of P450 reactions catalyzed by
`human P4505. The availability of the recombinant human P4505 expressed
`in various systems has also facilitated studies on their catalytic selectivity
`(Gonzalez et al., 1991a,b; Guengerich et ((1., 1996). Thus, it is now relatively
`straightforward to use in uitro studies to determine which P4503 oxidize a
`particular drug and which drugs can inhibit oxidations catalyzed by this
`P450. The in uitro determination of inducibility is not as easily done, but
`a number ofpossibilities exist with cultures ofhuman hepatocytes (Guillouzo
`et £11., 1993; Loretz et (11,, 1989) (also see chapter by A. P. Li).
`It is also possible to do logical in viva studies to test the relevance of
`in uitro findings. For instance, individuals known to be high or low in a
`particular P450 from the use of other noninvasive assays can be examined
`with regard to the pharmacokinetics of the new drug to see if there is a
`match. In some cases, inducers or inhibitors of a specific P450 can be given
`safely to people to verify that a P450 is involved in the oxidation of a drug.
`Also, the drug under consideration can be given to people to determine if
`it affects the pharmacokinetics of other drugs through enzyme induction.
`The acquisition of the in L/itro information about a new drug can be
`extremely useful. In many cases, the FDA now expects in uitro information
`on the P4505 involved in the oxidation of a drug early in the registration
`process. The in 1/itro information can be used to guide the more expensive
`and time—consuming in viva studies. In particular, potential adverse drug
`interactions due to pharmacokinetics can be predicted and the number of
`in uiz/0 interaction studies can be restricted, as some of those historically
`done with all new drugs may be found irrelevant. The in 1/[tro procedures,
`if used early in the drug development process, may be used to select from
`a series of potential candidates, in terms of which will be least likely to
`cause problems with drug—drug interactions. Another point is that the in
`uitro studies can be used as a guide in predicting bioavailability, simply by
`screening candidate drugs for resistance to oxidation by the major known
`human P4505.
`
`IV. Mechanisms of Drug Interactions Attributable
`
`to P4505
`
`A. Induction
`
`This is a phenomenon first identified a halfcentury ago in in viva studies
`with humans, primarily in the laboratories of Remmer and Brodie (Remmer,
`
`Vanda Exhibit 2043 - Page 7
`
`Vanda Exhibit 2043 - Page 7
`
`

`
`F. Peter Guengerich
`
`1959; Brodie et a/., 1958). Individuals who were administered certain drugs
`developed a certain “tolerance,” in that increasing doses were needed to
`produce the same effect. Work with experimental animals demonstrated
`that the effect could be reproduced. For instance, animals treated with
`barbiturates decreased their “sleeping time,” a parameter indicating how
`long a certain dose of a barbiturate would keep animals sedated (Burke,
`1981). Other studies on chemical carcinogenesis reinforced the concept of
`enzyme induction (Conney et ((1., 1956; Conney, 1967), particularly with
`what are now termed the P450 1A subfamily genes.
`The mechanism of P450 1A induction is perhaps the most well character-
`ized in this field (for reviews, see Hankinson, 1993; Denison and Wliitlock,
`1995). Although barbiturate induction was also discovered early, mechanis-
`tic studies on this phenomenon are not as well developed and there is not
`general agreement regarding observations in different laboratories (Rangara—
`jan and Padmanaban, 1989; Liang er a/., 1995; Ramsden er a/., 1993).
`Nevertheless, there seems to be a rather general agreement that most human
`P450 2C and 3A subfamily proteins are induced by barbiturates (Zilly et
`a/., 1975; Morel et al., 1990). Studies with experimental animals indicate that
`subfamily 2B proteins are induced by barbiturates (Burke, 1981; Guengerich,
`1987), but direct information on inducibility in humans is not available.
`Evidence indicates that human P450 2E1 is inducible by ethanol and isonia—
`zid, although the mechanism of the process is complex (Perrot et al., 1989;
`Kim et a/., 1994). Compounds (including drugs) that cause peroxisomal
`proliferation induce P4505 in the 4A subfamily in experimental animals
`(Muerhoff et (11,, 1992; Rao and Reddy, 1991; Gibson, 1993); presumably
`this can also happen in humans, although the system is suspected to be less
`responsive (Bell et a/., 1993). The mechanism involves the interactions of
`ligand—bound peroxisomal proliferation activation receptor (PPAR): retinoid
`X receptor (RXR) heterodimers with upstream recognition sequences (Lee
`et al., 1993), and compounds such as fatty acids and retinoids may be
`involved in this response.
`The effect of induction is simply to increase the amount of the P450
`present and make oxidation and clearance of a drug faster. As mentioned
`earlier, details of mechanisms remain to be understood. The overall situation
`is complicated because even in situations where a response element can be
`identified, there are probably interactions with other response systems that
`must be considered. However, knowledge of such phenomena will be useful
`in the further development of in uitro systems that can be used to screen new
`drug candidates for their potential as inducers of P4505 and other enzymes.
`
`B. Inhibition
`
`Vanda Exhibit 2043 - Page 8
`
`

`
`that can be attributed to inhibition rather than induction. There are different
`
`types of enzyme inhibition, and the clinical effects are influenced by the
`basic mechanisms.
`V
`The first type of inhibition is competitive, where the inhibitor and sub-
`strate compete for the same binding site on an enzyme. In the situations
`under consideration here, the inhibitor and the substrate would be drugs,
`competing for the binding site of a P450. Insofar as is currently known,
`P4505 are thought to have a single substrate—binding site (aside from some
`possible allosteric situations, z/ide infra) (Raag and Poulos, 1991; Cupp-
`Vickery and Poulos, 1995), although the sizes and flexibility of the micro-
`somal P4505 we are concerned with here are not known. The inhibitor may
`be a substrate itself. For examples, see the section on P450 ZD6 (uide infra).
`This type of inhibition is easily identified by the classic intersecting plots
`seen in in z/itro studies (Kuby, 1991).
`Another type of inhibition has precedent in the classical studies of
`enzymology. The two situations are called nozzcompetitiz/e inhibition, where
`the inhibitor binds at a site on the enzyme distinct from the substrate, and
`zmcompetitiz/e inhibition, where the inhibitor binds only to the enzyme-
`substrate complex (Kuby, 1991 ). Actually, neither of these have many clear
`examples in the literature of drug—drug interactions or in drug metabolism
`in general. (An example of a noncompetitive inhibitor would be a reagent
`that modifies sulfhydryl groups remote from the substrate—binding site to
`attenuate the activity of an enzyme.)
`A fairly common mechanism of inhibition related to drug—drug interac-
`tions is mec/mzzism-based, or suicide, inhibition (Silverman, 1988, 1995;
`Ortiz de Montellano and Correia, 1983; Ortiz de Montellano and Reich,
`1986; Halpert and Guengerich, 1997). In the strict definition of the mecha-
`nism, a substrate (the inhibitor) is transformed by the enzyme in the normal
`course used for other substrates and an intermediate is formed, which usually
`has a fleeting but finite half-life. This intermediate can partition between
`reaction with the enzyme (at the active site), to inactivate the enzyme, or
`undergo a different transformation (e.g., reaction with water or proton loss)
`to yield a stable product. The ratio of the two processes (latter : former) is
`termed the partition ratio and is used to compare the efficiencies of different
`_mechanism-based inactivators. Mechanism—based inactivators are character-
`ized in uitro by a number of properties, including time—dependent loss of
`enzyme activity, requirement for normal enzyme cofactors, blockage by
`noninhibitory substrates, saturation kinetics, and (usually) single irreversible
`modification of the protein or prosthetic group (Silverman, 1988). Examples
`of this type are seen in the P450 drug metabolism literature with compounds
`such as secobarbital (Levin et al., 1973), gestodene (Guengerich, 1990a),
`
`Vanda Exhibit 2043 - Page 9
`
`Vanda Exhibit 2043 - Page 9
`
`

`
`F. Peter Guengerich
`
`furafylline (Kunze and Trager, 1993), and disulfiram (Guengerich et a/.,
`1991; Brady et al., 1991), and many clinical interactions may be understood
`in these terms. The inhibition of specific enzymes by mechanism—based inacti-
`vators is an approach used in the design of new drugs. In principle, a
`substrate can be designed as a mechanism—based inactivator of a single
`enzyme. This approach has been used to attenuate monoamine oxidase
`(Thull and Testa, 1994). The only good examples of development of drugs
`to specifically inhibit P450s deal with P450 19, the steroid aromatase, which
`is a target in breast and ovarian tumors because of its role in estrogen
`production (Brodie, 1994). Nevertheless, there are many examples of experi-
`mental compounds that are selective inactivators of individual P450 enzymes
`in uitro and in experimental animals (Ortiz de Montellano and Reich, 1986).
`These can be used in a diagnostic manner (to help identify P4505 involved
`in various reactions) (Guengerich and Shimada, 1991) to label enzyme active
`sites (Roberts et al., 1994; Yun et al., 1992) and to identify a drug target
`in a complex mixture of proteins (Rando, 1984).
`Several other types of irreversible enzyme inhibition are related to
`mechanism—based inactivation but can be distinguished. In one case, there
`is time-dependent inhibition at the active site by reaction of a substrate (or
`analog) with the protein, unrelated to the normal enzyme mechanism. A
`good example is not available for P450, except perhaps a substrate such as
`acrylonitrile that reacts rather nonselectively with all protein sulfhydryls
`but is oxidized by P450 2E1. A slow—binding inhibitor of testosterone 50¢-
`reductase is the prostate growth inhibitor finasteride (Proscar) in which the
`enzyme bonds with the drugs at a slow rate, competitive with normal ste-
`roids, and irreversibly inactivates the enzyme (Tian et a/., 1995).
`Another case involves the conversion of a substrate to a product that
`is reactive enough to modify the protein. An example of this latter case is
`chloramphenicol, which is oxidized by P450 to an acyl chloride (Halpert
`et 111., 1985). The acyl chloride is not an enzyme intermediate in the strict
`sense. The product can be readily hydrolyzed by water. It would also leave the
`protein and modify other proteins; however, the similarity of the molecule to
`the substrate seems to keep it in the active site so that it will label groups
`there. Distinguishing inhibitors of this type from true mechanism—based
`inactivators may not be easy; one test is to determine if a scavenger such
`as glutathione (which would not enter the active site of the enzyme) can
`block inhibition. Another test is to find a certain P450 enzyme (“P450 1”)
`that is not inactivated when incubated with the drug (plus normal cofactors).
`This P450 (“P450 1”) can be mixed with the drug, cofactors, and another
`P450 known to be inactivated (“P450 2”). If P450 1 is now inactivated,
`then the most direct explanation is that a reactive product has migrated
`from P450 2 to P450 1. Although the mechanistic distinction may seem
`
`Vanda Exhibit 2043 - Page 10
`
`

`
`the heme iron instead of covalent modification of amino acid residues. For
`
`instance, many amines are oxidized to nitroso compounds that form spectral
`complexes with absorbance maxima at 455 nm (jonsson and Lindeke, 1992;
`Mansuy et a/., 1983). A classical case in pesticide biochemistry is the synergist
`piperonyl butoxide, which is oxidized to a carbene that binds the heme
`(Ortiz de Montellano and Reich, 1986). Evidence shows that mechanisms
`of this type may be important in inhibition under physiological conditions
`(Bensoussan et a/., 1995).
`
`C. Stimulation
`
`Enzyme stimulation refers to the process by which direct addition of
`one compound to an enzyme enhances the rate of reaction of the substrate.
`This phenomenon has been observed in a number of cases with P4505
`(Halpert and Guengerich, 1997; Huang et al., 1981).
`Distinguishing enzyme induction and stimulation in viva is not easy
`because some of the compounds that seem most effective in P450 stimulation
`are also enzyme inducers, e.g., flavonoids. One approach used was the
`treatment of rats with a substrate in which product formation was accompa-
`nied by the release of tritiated water, for a short period of time (15 min),
`in the absence or presence of flavone (Lasker er ((1., 1982). The increase in
`product formation observed (in total body radioactive water) in the presence
`of flavone provides evidence that stimulation occurred in a time frame before
`significant enzyme induction could have occurred.
`In our laboratory we have been studying the effect of a—naphthoflavone
`on the oxidation of the carcinogen aflatoxin B, by P450 3A4 (Raney et al.,
`1992; Ueng et a1., 1995). oz-Naphthoflavone has the interesting effect of
`inhibiting the 3oz—hydroxylation of aflatoxin B, but stimulating the 8,9-
`epoxidation, and our current working hypothesis is that an allosteric mecha-
`nism is involved (Ueng et al., 1995; Guengerich et al., 1994b). In line with
`this view, plots of rates of these reactions versus substrate concentration
`are sigmoidal
`in the absence of wnaphthoflavone but hyperbolic in the
`presence of oz-naphthoflavone (Ueng et a/., 1995). There is also evidence in
`the literature that sigmoidal kinetics are observed in the in uitro oxidation
`of drugs [e.g., carbamazepine (Kerr et a/., 1994) and possibly acetaminophen
`(Lee et al., 1991)] and steroids [e.g., progesterone and 17B—estradiol (Schwab
`et al., 1988)], usually with P450 3A subfamily enzymes.
`The in viz/o relevance of these phenomena to drug metabolism remains
`to be established, as does the mechanism(s).
`
`Vanda Exhibit 2043 - Page 11
`
`Vanda Exhibit 2043 - Page 11
`
`

`
`F. Peter Guengerich
`
`V. Examples of P450-Based Interactions
`
`A. Cimetidine
`
`Cimetidine (Tagemet, Fig. 2) is a drug that inhibits antihistamine H1
`receptor binding and is used in the treatment of gastric ulcers. There is
`considerable literature on the inhibition of drug metabolism by cimetidine
`in both animal and human models (Gerber er al., 1985). A similar H3
`receptor antagonist, ratinidine (Zantac), was developed by another company
`and was devoid of the inhibitory properties, a point that was exploited
`in marketing.
`Analysis of the scientific literature indicates thatcimetidine is a relatively
`weak P450 inhibitor (Knodell et £11., 1991). No serious acute episodes of
`adverse health have been attributed to cimetidine despite long use in many
`patients, many of whom are undoubtedly using other drugs.
`The mechanism of inhibition appears to involve the imidazole ring of
`cimetidine, which is not present in ranitidine. Cimetidine shows selectivity
`for inhibiting reactions catalyzed by P4505 ZD6 and 3A4 (Knodell et ((1.,
`1991). The inhibition has generally been regarded as due to competitive
`binding of cimetidine, possibly through interaction of the imidazole with
`the P450 heme. However, some evidence for mechanism—based inactivation
`of P450 has also been published (Coleman et a/., 1991), although a chemical
`basis has not been established.
`
`B. P450 2D6
`
`P450 2D6 inhibitors and substrates have attracted considerable concern.
`In the early 1970s Smith personally experienced an adverse response in a
`clinical trial of the antihypertensive agent debrisoquine. This episode led
`him to study the basis in more detail, and the work led to the identification
`of a subset of the population (~7% Caucasians) as “poor metabolizers,”
`who hydroxylated the drug at a much slower rate than the rest of the
`population (Mahgoub et al., 1977).
`Subsequent work led to characterization of this enzyme, P450 ZD6, by
`purification, CDNA cloning, and genetic analysis (Gonzalez et a/., 1988;
`Gonzalez and 1\/Ieyer, 1991). P450 ZD6 is now recognized to be involved
`in the oxidation of >30 drugs. Some of these show relatively narrow thera-
`
`CN
`
`H3C\N N/\/SWé<
`
`H
`
`H
`
`NH
`
`N§/
`
`Vanda Exhibit 2043 - Page 12
`
`

`
`(Shah et a/., 1982). However, P450 2D6—deficient individuals do not convert
`the pro-drug encainide to its active form as effectively as the rest of the
`population (\X/oosley ez‘ a/., 1981). A number of drugs are also potent
`inhibitors of P450 ZD6 (Fig. 3) (Strobl et a/., 1993). Prominent among these
`are alkaloids such as quinidine and the ajmalicine derivatives (Strobl er a/.,
`1993; Fonne—Pfister and Meyer, 1988).
`It is now relatively easy to identify P450 ZD6 substrates and inhibitors
`in 1/itro early in the development process. Strong P450 ZD6 inhibitors are
`generally avoided. An issue can be raised, though, as to how serious a P450
`ZD6 inhibitor really is. Because ~5% of the population (depending on the
`country) is already deficient in P450 ZD6, the effect of the inhibitor is to
`extend this group of individuals. The problem would be slow metabolism
`of P450 2D6 substrates, but this may not be a serious issue.
`
`
`
`H‘-,CO,C
`
`Caihanamina
`
`R0
`
`Sempervirlne
`
`R5
`
`
`
`lndoia derivatives
`
`R2
`
`H
`
`H
`
`HO
`
`N
`R3
`
`H
`
`Fl:
`
`/ \
`
`N
`Oulnldine/quinine and
`derivatives
`
`
`
`Ouinlnone
`
`R4
`
`\
`/
`N
`Ouinoiine and derivatives
`
`R
`
`/
`
`N
`H
`Harman and derivatives
`
`FIGURE 3
`
`Some inhibitors of P450 ZD6 (Strobl at a/., 1993).
`
`Vanda Exhibit 2043 - Page 13
`
`Vanda Exhibit 2043 - Page 13
`
`

`
`F. Peter Guengerich
`
`The issue of development of P450 ZD6 substrates has been a more
`serious matter, and some pharmaceutical companies had developed policies
`of dropping these from development. A realistic way of addressing the issue
`is to test candidate drugs in 1/itro to determine if they are substrates and
`then proceed to examine them in viva to establish the pharmacokinetics
`and the therapeutic window. The majority of P450 ZD6 substrates can
`probably be tolerated reasonably well even by P450 ZD6-deficient indi-
`
`The molecular basis of the P450 ZD6 polymorphism has been described
`in detail, and there are a number of alleles that contribute to cause both
`unusually slow and also unusually rapid oxidation (Broly et 41., 1991;
`johansson et (11,, 1993). The P450 ZD6 substrates and inhibitors all seem
`to share a basic nitrogen group, which is positioned 5-7 A away from the
`site of hydroxylation. The carboxylate anionic moiety of Asp 301 has been
`suggested to interact with the basic nitrogen of the substrate, on the basis
`of modeling and site-directed mutagenesis work (Ellis et al., 1995). Sub-
`strates and inhibitors have been used to develop pharmacophore models of
`P450 ZD6 (Strobl et al., 1993; Islam et (IL, 1991; Koymans et £11., 1992).
`The strong inhibitors of P450 ZD6 (e.g., quinidine) are not readily oxidized
`(Strobl et al., 1993; Guengerich et al., 1986b), and the conclusion has been
`reached that the basic nitrogen in these binds to the same protein anion as
`the substrates (Asp 301) but no atoms that can be oxidized are accessible
`to the FeO complex (Islam er al., 1991). The model does not explain the
`oxidation of deprenyl by P450 ZD6 (Grace et al., 1994). A modification
`involves the transient deprotonation of the amine and electron transfer
`(Grace et al., 1994; Guengerich, 1995).
`
`I701-Ethynylestradiol
`
`This is a classic example of a drug—drug interaction and one of the few
`attributed to induction, instead of inhibition. In the early 1970s several
`German reports indicated that women who were using oral contraceptives
`began spotting or became pregnant after using rifampicin or barbiturates
`(Reimers and jezek, 1971; Nocl<e—Fincl< et £11., 1973; Janz and Schmidt,
`1974). The major estrogen in oral contraceptives is 17a—ethynylestradiol
`(Fig. 4), which is metabolized via 2-hydroxylation, plus other pathways
`(Bolt et al., 1973; Guengerich, 1990b). Administration of rifampicin resulted
`in the faster elimination of 17oz—ethynylestradiol in volunteers (Bolt 61‘ al.,
`
`Subsequently, P450 3A4 was shown to be a major enzyme involved in
`the (2—)hydroxylation of 17oz—ethynylestradiol (Guengerich, 1988). P450
`3A4 can also be induced by rifampicin or barbiturates in cultured human
`
`Vanda Exhibit 2043 - Page 14
`
`

`
`} |
`
`OH
`§.“\\'_-E
`
`.
`
`S OH
`.\‘\\—-_
`
`170:-Ethynylestradiol
`
`D°3°9°5"°'
`
`0H
`| am“:
`
`Norethlsterone
`
`OH
`.\‘\\:_
`
`3-Ketodesogestrel
`
`OH
`
`r 39“:
`
`O
`
`O
`
`O
`
`HO
`
`0
`
`o
`
`,
`
`; oH
`“E
`
`Gestodene
`
`OH
`.\‘\\'——
`
`Levonorgestrel
`
`OH
`
`‘F
`
`Q4 3 E
`
`11.cH2.A15.N°ret|-fistefone
`11'CH2-N0l'ethlStef0ne
`Structures of 17a—ethynylestradiol and several progestins used in oral Contracep-
`FIGURE 4
`tives (Gucngerich, l990a,l)).
`
`The ineffectiveness of oral contraceptives due to P450 3A4 induction
`can be explained in these terms. There could also be contributions of induced
`conjugating enzymes (e.g., UDP-glucuronosyltransferases), but these have
`not been documented. This phenomenon of lack of efficacy of oral contracep-
`tives is still a problem because of the low doses 0f17oz-ethynylestradiol used
`(to prevent unwanted effects of estrogens) and the sensitivity to changes due
`to variations in P450 3A4.
`
`In the course of work with 17oz—ethynylestradiol and oral contraceptives,
`the progestin gestodene (Fig. 4) was found to be a relatively effective and
`
`Vanda Exhibit 2043 - Page 15
`
`Vanda Exhibit 2043 - Page 15
`
`

`
`F. Peter Guengerich
`
`selective mechanism—based inactivator of P450 3A4 in in z/itro experiments
`(Guengerich, 1990a). This inactivation is due in part to the presence of an
`ethynyl moiety, which is also a part of many P450 inactivators (Ortiz de
`Montellano et a/., 1979; Gan et al., 1984). However, most of the pro-
`gestins used in oral contraceptives have 1701-ethynyl groups (Fig. 4), and
`other features of gestodene are apparently responsible for the inactivation
`(Guengerich, 1990a). This inactivation phenomenon has been postulated to
`account for the increased levels of estradiol and cortisol in women using
`oral contraceptives (]ung—Hoffmann and Kuhl, 1990), although it
`is not
`clear that the dose of gestodene is sufficient to inhibit a large fraction of
`hepatic or intestinal P450 3A4 (Guengerich, 1990a).
`
`D. Terfenadine
`
`Terfenadine is a component of the antihistamine formulation Seldane.
`It is rapidly oxidized by P450 3A4 to two products, acyclinol and an alcohol
`derived from oxidation of a t-butyl methyl group (Fig. 5) (Yun et al., 1993).
`Acyclinol is inactive. The alcohol is further oxidized to a carboxylic acid
`by either P450 3A4 (Rodrigues ez‘ al., 1995) or by dehydrogenases; the
`relative contributions of the two enzyme systems are not known (Fig. 6). This
`carboxylic acid, like terfenadine itself, binds to the H, histamine receptor and
`should produce relief of allergy symptoms. However, the acid is a zwitterion