DRUG DEVELOPMENT
`
`Selection of the First-Time Dose in Humans:
`
`Comparison of Different Approaches Based
`on Interspecies Scaling of Clearance
`
`Iftekhar Mahmood, PhD, Martin David Green, PhD, and ]. Edward Fisher, PhD
`
`
`
`The authors describefour approaches to selecting a safe start-
`ing dose for humans in clinical drug trials based on
`interspecies scaling of clearance. Human clearance was pre-
`dicted by scaling for 10 example drugs for which animal
`clearance values were available in the literature. The pre-
`dicted human clearance values were then used to select the
`
`estimated starting dose in humans. These doses were then
`compared with the actual doses given to humans during clin-
`ical trials. All four approaches used to estimate the first-time
`dose in humans provided values that were within the dose
`
`range given to humans from Phases I to III. This work demon-
`strates that animal pharmacokinetic data can be used to esti-
`mate a suitable human starting dose, provided the data have
`been obtainedfrom a dose thatproduces no adverse effects.
`
`Keywords: Interspecies scaling; first-time dose; drug devel-
`opment; allometric scaling; predicted human
`clearance; toxicity
`Journal of Clinical Pharmacology, 2003;43:692-697
`@2003 the American College of Clinical Pharmacology
`
`Interspecies scaling is frequently used to predict
`
`pharmacokinetic parameters from animals to hu-
`mans during drug development and is becoming a use-
`ful tool especially for the selection of the first-time dose
`in humans.1 Estimation of a starting dose for “first-in-
`human” clinical trials of new molecular entities in
`
`healthy volunteers is very important since a low start-
`ing dose will prolong dose optimization, and a high
`starting dose may cause serious toxicity. But despite
`the importance of this task from a drug development
`standpoint, there is no consensus regarding the best ap-
`proach for estimating the starting dose.2
`
`From the Division ofClinical Trial Design and Analysis, Centerfor Biologics
`Evaluation and Research (I. Mahmood, M. D. Green) and the Division of
`Neuropharmacological Drug Products, Center for Drug Evaluation and
`Research (J.E. Fisher), Food and Drug Administration (FDA), Rockville,
`Maryland. The views expressed in this article are those of the authors and
`do not reflect the official policy of the FDA. No official support or endorse-
`ment by the FDA is intended or should be inferred. Submitted for publica-
`tion December 22, 2002; revised version accepted April 12, 2003. Ad-
`dress for reprints: lftekhar Mahmood, Division of Clinical Trial Design and
`Analysis, Office of Therapeutic Research and Review, Clinical Pharmacol-
`ogy and Toxicology Branch (HFD-579), Center for Biologics Evaluation
`and Research, Food and Drug Administration, Woodmont Office Center I,
`Suite 200N, 1401 Rockville Pike, Rockville, MD 20852.
`DOI: 10.1177/0091270003254631
`
`One of the most commonly used methods for select-
`ing the starting dose has involved the conversion of an
`animal no observed adverse effect level (NOAEL) to the
`human equivalent dose (HED) using appropriate scal-
`ing factors, followed by application of a safety factor
`(SF). This procedure is described in detail in a recently
`issued Food and Drug Administration (FDA) draft
`guidance, which outlines an algorithmic process for se-
`lecting the maximum recommended starting dose
`(MRSD) for healthy adult volunteers based on animal
`toxicology data and administered doses? Although
`this method has proven generally adequate, it presents
`a number of problems for the toxicologist. Determina-
`tion of the appropriate animal NOAEL is a difficult and
`time-consuming task, depending on a number of key
`study variables, including duration of treatment, dose
`selection, and species. The choice of an appropriate
`scaling factor also involves considerable uncertainty.
`For most systemically administered drugs, conversion
`to the I-IED is typically based on the normalization of
`doses to body surface area.“'5 However, the applicabil-
`ity of the data used to derive this scaling factor (i.e., the
`two-thirds power ofbody weight) to risk assessment for
`noncytotoxic agents and the largely unknown impact
`of experimental conditions on dose scaling across spe-
`
`692 o I Clin Pharmacol 2003;43:692-697
`
`Apotex v. Novartis
`lPR2017-00854
`
`NOVARTIS 2102
`
`Apotex v. Novartis
`IPR2017-00854
`NOVARTIS 2102
`
`

`

`SELECTION OF THE FIRST-TIME DOSE IN HUMANS
`
`cies have been widely discussed.‘*'3 As has been
`pointed out repeatedly, the choice of an SF is almost
`entirely arbitrary and without scientific justification.
`In those cases for which adequate data exist, animal
`pharmacokinetics may aid in determining initial clini-
`cal doses. Although not without their own set of limita-
`tions and assumptions, pharmacokinetically guided
`approaches to dose extrapolation such as those de-
`scribed herein should provide a more rational basis for
`calculating an appropriate starting dose. In 1986, Col-
`lins et al'3 proposed a pharmacokinetically guided dose
`escalation scheme for anticancer drugs. To select the
`first-time dose in humans, Reigner and Blesch2 sug-
`gested the use of the lowest AUC at the NOAEL when a
`drug is given to several species as well as the predicted
`human clearance (dose in humans = AUC X predicted
`human clearance). Therefore, for this approach, it is es-
`sential that the clearance be predicted with reasonable
`accuracy in humans. Over the years, many different ap-
`proaches have been suggested to improve the predic-
`tion of clearance in humans.1°'14 With the introduction
`
`of the rule of exponents by Mahmood and Balian,10 the
`probability of predicting clearance with reasonable ac-
`curacy has vastly improved. The objective of this report
`is to propose four different approaches based on the
`predicted human clearance and their suitability to se-
`lect a safe first-time dose in humans for the Phase I
`
`study.
`
`METHODS
`
`A literature search was conducted to obtain clearance
`
`values for 10 drugs15'33 that have been studied in at least
`three animal species (mice, rat, rabbit, monkey, or dog).
`These drugs are eliminated renally or by extensive me-
`tabolism. Drugs used in this study were given either in-
`travenously or orally. The first step was to predict clear-
`ance in humans. Scaling of clearance was performed
`using the following three methods. Human data were
`not included in the scaling. In the case when several
`doses were given to animals, the lowest dose was se-
`lected for scaling.
`
`Prediction of Clearance
`
`Method I. Clearance of each drug was plotted against
`the body weight on a log-log scale, and the following
`allometric equation was used to predict clearance in
`humans.
`
`CL = a(W)b,
`
`(1)
`
`where W is the body weight and a and b are the coeffi-
`cient and exponent of the allometric equation,
`respectively.
`Method II. The observed clearance values in the dif-
`
`ferent animal species were multiplied by their respec-
`tive maximum life span potential (MLP) and plotted as
`a function of body weight on a log-log scale. From the
`allometric equation, clearance X MLP was estimated in
`humans and the result was then divided by the MLP of
`humans (8.18 x 105 h) to predict the clearance in man.
`
`CL = a(MLP x Clearance)b/8.18 x 105.
`
`(2)
`
`MLP in years was calculated from the following equa-
`tion, as described by Sacher“:
`
`MLP (years) = 135.4 (13W)°-83$ (W)‘°'225,
`
`(3)
`
`where both brain weight (BW) and body weight (W) are
`in kilograms.
`
`Method III. In this approach, clearance of animal
`species was multiplied by the brain weight of the spe-
`cies, and the product was plotted as a function of body
`weight on a log-log scale. The allometric equation
`(equation (4)) was then used to predict the clearance in
`humans using the human brain weight (1.53 kg).
`
`CL = a(CL x BW)b/1.53,
`
`(4)
`
`where BW is the brain weight in kilograms.
`
`Selection of Dose
`
`Based on the predicted human clearance, the following
`four approaches were used to select the dose in hu-
`mans. The predicted dose was then compared with the
`lowest and highest doses given to humans during the
`clinical trials (from Phase I to Phase III).
`
`Approach I. In this approach, the clearances of the
`species (used in the prediction of clearance for hu-
`mans) were plotted on linear scale against the dose
`given to the species. The resultant equation was then
`used to recommend the starting dose in humans as
`follows:
`
`Dose = a + b(x),
`
`(5)
`
`where a is the intercept, b is the slope, and x is the pre-
`dicted clearance in humans.
`
`Approach II. In this approach, the dose given to each
`species was multiplied by the HED and then plotted
`against clearance in each species on a linear scale. The
`resultant equation was then used to select the dose, as
`
`DRUG DEVELOPMENT
`
`693
`
`

`

`MAHZVIOOD ET AL
`
`Table I Names of the Studied Drugs and the Species Used in the Allometric Scaling
`
`
`
` Drugs Species Used Route of Administration“ Reference Animals Reference Humans
`
`
`
`
`
`
`
`17
`15, 16
`Oral
`m, r, (1
`Topiramate
`19
`18
`IV
`m, r, mk, d
`Moxifloxacin
`20
`20
`Oral
`r, d, mk
`Zonisamide
`21
`21
`Oral
`m, r, mk, (1
`Troglitazone
`23
`22
`Oral
`m, r, d, mk
`Venlafaxine
`27
`24-26
`IV
`r, rb, d
`Morphine
`29
`28
`Oral
`r, rb, d
`Felbamate
`30
`30
`Oral
`III, r, mk
`Bepridil
`32
`31
`IV
`m, r, rb, mk
`Stavudine
`
`Zenarestat 33 m, r, d Oral 33
`
`
`
`In, mouse; r, rat; rb, rabbit; mk, monkey; d, dog.
`a. The route of administration in animals and humans was same.
`
`described in equation (5). HED in a given species was
`estimated as followsa:
`
`Dose (mg) = (474 ugOh/mL X 1.26 L/h)/3.27.
`
`HED = animal dose in mg/kg
`>< (animal weight in kg/human weight in kg)°'33.
`
`Approach III. This approach is a slightly modified
`version of the pharmacokinetically guided approach.
`No safety factor was used in this dosing recommenda-
`tion based on the assumption that the dose given to ani-
`mals is much lower than the NOAEL. Since the AUC
`
`values for a given drug were available in more than one
`species, the recommended dose was estimated using
`the lowest AUC observed in a given species as follows:
`
`Dose (mg) = AUC in animal (pg-h/mL)
`>< predicted clearance in humans (L/h).
`
`Approach IV This mathematically manipulated ap-
`proach can be used to recommend the first-time dose in
`humans as follows.
`
`First, one selects the species whose clearance (per kg
`body weight) is nearest to the predicted human clear-
`ance (based on kg body weight). A correction factor is
`then obtained by dividing the clearance of the chosen
`species by the predicted human clearance. Then the
`recommended dose can be selected according to the
`following equation:
`
`Dose (mg) = (AUC in the chosen animal (ugOh/mL)
`X predicted clearance in humans (L/h))/correction factor.
`
`Example: For felbamate, the observed clearance in rat,
`rabbit, and dog was 2.54, 0.98, and 1.55 mL/min/kg, re-
`spectively. In man, the predicted clearance was 0.3 mL/
`min/kg. Based on the clearance value, rabbit was the
`closest species to man. The estimated correction factor
`was 3.27 (098/03), and the dose was calculated as
`follows:
`
`694 o I Clin Pharmacol 2003;43:692-697
`
`Thus, the recommended starting dose of felbamate in
`humans was 183 mg.
`
`RESULTS
`
`Table I is the summary of the drugs and species used in
`this study as well as the route of administration of these
`drugs to animals and humans. The exponents of the
`simple allometry and the correlation coefficient be-
`tween body weight and clearance are summarized in
`Table II. Table II also compares the predicted and ob-
`served clearance of the studied drugs.
`A good correlation between body weight and clear-
`ance was observed for most of the drugs tested. The re-
`sults of this study (as in the previous study”) indicate
`that there are specific conditions under which only one
`of the three methods can be used for improved predic-
`tion of clearance. When the exponents of the simple
`allometry ranged from 0.55 to 0.70, the simple
`allometry was considered suitable for the prediction of
`clearance in humans. When the exponents of the sim-
`ple allometry ranged from 0.70 to 0.99, the MLP ap-
`proach was found to be appropriate for the prediction
`of clearance in humans. The product of clearance and
`brain weight was considered suitable for the prediction
`of clearance in humans when the exponents of the sim-
`ple allometry were 2 1.0. It can be seen fiom Table II
`that simple allometry was not adequate for the predic-
`tion of clearance for all drugs, but the use of the rule of
`exponents vastly improved the prediction of these
`drugs.
`In this study, however, there were two drugs whose
`predicted clearances were many times higher than the
`observed clearances. Zenarestat’s exponent was 1.340,
`but as mentioned previously,10 if the exponent of the
`
`

`

`SELECTION OF THE FIRST-TIME DOSE IN HUMANS
`
`Table II Observed versus Predicted Clearance (mL/min) in Humans
`
`
`Drugs
`Exponent
`Correlation Coefficient (1')
`Observed CL
`Predicted CL (SA)
`Predicted CL (RE)
`
`Topiramate
`Moxifloxacin
`Zonisamide
`
`Troglitazone
`Venlafaxine
`
`Morphine
`Felbamate
`
`0.557
`0.603
`0.735
`
`0.633
`0.782
`
`0.777
`0.823
`
`0.864
`0.969
`0.987
`
`0.984
`0.866
`
`0.985
`0.979
`
`22-36
`154
`16
`
`751-891
`2240
`
`1300
`30
`
`51
`166
`22
`
`793
`4874
`
`2714
`59
`
`51
`166
`20
`
`793
`2268
`
`910
`21
`
`6535
`12,886
`1085
`0.995
`0.842
`Bepridil
`Stavudine
`0.901
`0.999
`572
`961
`400
`
` Zenarestat 1.340 0.994 47 688 155
`
`
`
`
`
`
`SA, simple allometry; RE, rule of exponents.
`
`allometry is > 1.30, the predicted clearance may be
`many times higher than the observed clearance, and in-
`deed this was the case with zenarestat. In a previous pa-
`per,35 Mahmood mentioned that the prediction of oral
`clearance would be erratic when the oral clearance of
`
`the species used in the scaling is either equal to or
`greater than the liver blood flow and the observed hu-
`man clearance is less than the liver blood flow. This ob-
`
`servation was found to be true with bepridil. Bepridil’s
`oral clearance in the mouse, rat, and monkey was
`greater than their respective blood flow, resulting in a
`much higher prediction of oral clearance (observed hu-
`man CL = 1085 mL/min, which is less than human liver
`
`blood flow) in humans (predicted CL = 6535 mL/min).
`The results obtained by the four approaches to se-
`lecting the first-time dose in humans are compared
`with the doses actually given to humans from Phase I to
`Phase III in Table III. The clinical dose range shown ex-
`tends from the first human dose to the highest dose
`used in the definitive efficacy trials, which can be con-
`sidered a well-tolerated but not necessarily a maximum-
`tolerated dose. The method of dose selection in the
`
`original trials is unknown. Approach I was the most ag-
`gressive approach and may not be suitable for the selec-
`tion of the first-time dose in humans. Approach III was
`the most conservative method and, like Approach I,
`may be unsuitable. Approaches II and IV were more
`moderate and could be useful in estimating a safe and
`efficient dose for the first-time administration of a drug
`to humans. It should be noted, however, that with the
`
`exception of zenarestat, even the dose arrived at by Ap-
`proach I did not exceed the highest dose given to hu-
`mans for any of the studied drugs.
`
`DISCUSSION
`
`Interspecies scaling is based on the assumption (a cor-
`rect assumption) that there are anatomical, physiologi-
`cal, and biochemical similarities among animals,
`which can be described by mathematical models. It is
`now a well-established fact that many physiological
`processes and organ sizes exhibit a power-law relation-
`ship with the body weight of species. This relationship
`is the scientific basis of allometry.
`Pharmacokinetics play an important role during
`drug development. Characterization of absorption,
`distribution, metabolism, and excretion (ADME) in
`animals is of fundamental importance. The pharma-
`cokinetic parameters, especially clearance, can be ex-
`trapolated to humans for the selection of a safe dose for
`the first-dosing trials in humans. Two of the widely
`used methods of first-time dose selection are based on
`
`the NOAEL or toxicokinetic studies of a drug in one to
`three species. There are several disadvantages to these
`approaches. In the NOAEL approach, it becomes nec-
`essary to select a NOAEL, which is a tedious and time-
`consuming process. In reality, one may never find an
`absolute NOAEL in a given species. In toxicokinetic
`studies, animals are given a very high dose chronically,
`and if the resultant toxicity does not kill the animal, it
`may alter the physiology of the animal. This change in
`the physiology of the animal may have an impact on
`the pharmacokinetics of a given drug. Overall, both the
`NOAEL and toxicokinetic approaches for the first-time
`selection of a dose to humans are slow and very con-
`servative approaches. In fact, there is no need to select
`the first dose in humans based on NOAEL or
`
`DRUG DEVELOPMENT
`
`695
`
`

`

`MAHIVIOOD ET AL
`
`Table III Recommended First-Time Dose in Humans by Different Approaches
`
`
`
` Drugs Approach I Approach II Approach III Approach IV Dose Given
`
`
`
`
`
`
`
`
`
`Topiramate
`Moxifloxacin
`Zonisamide
`
`Troglitazone
`Venlafaxine
`
`Morphine
`Felbamate
`
`175
`107
`595
`
`90
`60
`
`9.3
`145
`
`95
`86
`329
`
`37
`23
`
`5.3
`80
`
`36
`23
`265
`
`35
`22
`
`2.5
`122
`
`44
`80
`490
`
`85
`60
`
`3
`183
`
`100-1200
`84-400
`200-800
`
`100-600
`25-150
`
`7.5-10
`100-800
`
`200-400
`176
`161
`177
`400
`Bepl‘idil
`47-280a
`134
`170
`86
`243
`Stavudine
`
`
`
`
`
`Zenarestat 150-600 845 460 313 353
`
`Approach I = dose versus clearance (CL); Approach II = human equivalent dose (HED) dose versus CL; Approach III = no observed adverse effect level (NOAEL);
`Approach IV = correction factor.
`a. Given orally, but the recommended dose is based on the scaling of IV data as the absolute bioavailability is 1 in humans. The initial IV dose in humans was
`70 mg.
`
`toxicokinetics. One can achieve this goal by giving a
`safe low dose to animals and, based on the prediction of
`clearance, selecting a safe and suitable dose in humans
`as described in this study. Nevertheless, this does not
`necessarily mean that one should not try to determine
`the NOAEL or conduct a toxicokinetic study as these
`endpoints provide information that is often useful in
`characterizing toxicity.
`In this report, we have proposed several methods
`that can be used to administer a first-time dose to hu-
`
`mans that is not only safe but also avoids many unnec-
`essary low-dose schemes. Even when the predicted
`clearance is many times higher than the observed clear-
`ance, as seen with bepridil (sixfold higher) and
`zenarestat (threefold higher), the doses selected by Ap-
`proaches II, III, and IV were well below the highest
`well-tolerated doses given to humans. It is, however,
`not known that these clinical doses represent the maxi-
`mum tolerated dose or whether one could give even
`higher doses to humans without producing any signifi-
`cant toxicity. Approach III is, in fact, a modified version
`of NOAEL or toxicokinetic methods for the selection of
`
`the first-time dose in humans. In this approach, a safety
`factor was not used since the animal doses were low
`
`and without any side effects.
`An important caveat is that the above-mentioned ap-
`proaches ignore the fact that there may be a toxic me-
`tabolite of a drug in humans that is not formed in ani-
`mals. Considering that current standards of drug
`development require a thorough investigation of me-
`tabolites formed in animals and humans, such a case is
`highly unlikely to occur.2 Another important point
`worth considering is that genetic polymorphism, as in
`the case of 2D6, may occur. One may then require de-
`
`veloping two different starting doses, one for poor
`metabolizers and one for extensive metabolizers.z
`
`The main objective of this work is to describe differ-
`ent approaches that can help to select a first dose in hu-
`mans that is not only safe but also efficient (neither very
`low nor very high). These suggested approaches pro-
`vide rational alternatives to the somewhat arbitrary
`dose selection process often used. Having a clear set of
`methods rather than relying on some fuzzy approach
`should be an important advantage in the bigger context
`of drug development. Although various safety factors
`used in the first-time dose in humans have no scientific
`basis, one can tailor these factors based on the need or
`
`the characteristics of a given drug. It should also be
`noted that there is no right or wrong method, and one
`can select one of the many approaches available by us-
`ing scientific judgment and the ease of the method. Our
`proposed methods are one of the many approaches that
`may be used to select a first-time dose to humans. Other
`approaches may be considered in relationship to the
`severity and incidence of toxicity. For example, a con-
`servative approach is suitable for drugs likely to have a
`narrow therapeutic index.
`
`REFERENCES
`
`1. Boxenbaum H, Dilea C: First-time-in—human dose selection:
`allometric thoughts and perspectives. I Clin Pharmacol 1995;35:957-
`966.
`
`2. Reigner BG, Blesch KS: Estimating the starting dose for entry into
`humans: principles and practice. Eur] CIin Pharmacol 2002;57:835-
`845.
`
`3. US. Food and Drug Administration: Draft guidance for industry
`and reviewers: estimating the safe starting dose in clinical trials for
`
`696 o I Clin Pharmacol 2003;43:692-697
`
`

`

`SELECTION OF THE FIRST-TIME DOSE IN HUMANS
`
`therapeutics in adult healthy volunteers. Fed Reg 2003;68:2340-
`2341.
`
`4. Freireich EI, Gehan EA, Rall DP, Schmidt LH, Skipper HE: Quanti-
`tative comparison of toxicity of anticancer agents in mouse, rat, ham-
`ster, dog, monkey, and man. Cancer Chemother Rep 1966;50:219-
`244.
`
`5. Shein PS, Davis RD, Carter S, Newman I, Shein DR, Rall DP: The
`evaluation of anticancer drugs in dogs and monkeys for the predic-
`tion of qualitative toxicities in man. Clin Pharmacol Therap 1979;
`11:3-40.
`
`6. Travis CC, Bowers IC: Interspecies scaling of anesthetic potency.
`Toxicol Indust Health 1991;7:249-260.
`
`7. Watanabe K, Bois FY, Zeise L: Interspecies extrapolation: areexam-
`ination of acute toxicity data. Risk Anal 1992 ;2:301-310.
`8. Rhomberg LR, Wolff SK: Empirical scaling of single oral lethal
`doses across mammalian species based on a large database. HiskAnal
`1998;18:741-753.
`
`9. Collins IM, Zaharko S, Dedrick RL, Chabner BA: Potential roles for
`preclinical pharmacology in phase I clinical trials. Cancer Treat Rep
`1986;70:73-80.
`
`10. Mahmood I, Balian ID: Interspecies scaling: predicting clearance
`of drugs in humans. Three different approaches. Xenobiotica
`1996;26:887-895.
`
`11. Boxenbaum H, Fertig IB: Scaling of antipyrine intrinsic clearance
`of unbound drug in 15 mammalian species. Eur I Drug Metab
`Pharmacokin 1984;92177-183.
`
`12. Lave TH, Dupin S, Schmitt C, Chou RC, Iaeck D, Coassolo PH: In-
`tegration of in vitro data into allometric scaling to predict hepatic
`metabolic clearance in man: application to 10 extensively metabo-
`lized drugs. IPharm Sci 1997;86:584-590.
`13. Obach RS, Baxter IG, Liston TE, Silber BM, Iones C, Macintyre F,
`et al: The prediction of human pharmacokinetic parameters from
`pre-clinical and in vitro metabolism. I Pharmacol Exp Ther 1997;
`283:46-58.
`
`14. Boxenbaum H: Interspecies pharmacokinetic scaling and the
`evolutionary-comparative paradigm. Drug Metab Rev 1984;15:1071-
`1 1 2 1.
`
`15. Streeter AI, Stahle PL, Holland ML, Pritchard IF, Takacs AR:
`Pharmacokinetics and bioavailability oftopiramate in the beagle dog.
`Drug Metab Dispos 1995;23:90-93.
`16. Streeter AI, Stahle PL, Hills IF, Lucas PN, McKown LA, Nese SL,
`et al: Pharmacokinetics of topiramate in the rat. Pharm Res 1994;
`11(Suppl.):SS72.
`17. Rosenfeld WE: Topiramate: a review of preclinical, pharma-
`cokinetic, and clinical data. Clin Ther 1997;19:1294-1308.
`
`18. Siefert HM, Domdey—Bette A, Henninger K, Hucke F, Kohlsdorfer,
`Stass HH: Pharmacokinetics of the 8-methoxyquinolone,
`moxifloxacin: a comparison in humans and other mammalian spe-
`cies. I Antimicrob Chemotherap 1999;43(Suppl. B):69-76.
`19. Stass H, Kubitza D: Pharmacokinetics and elimination of
`moxifloxacin after oral and intravenous administration in man. I
`Antimicrob Chemotherap 1999;43(Suppl. B):83-90.
`20. Matsumoto K, Miyazaki H, Fuji T, Maeda T, Hashimoto M: Ab-
`sorption, distribution and excretion of 3-(sulfanoyl [14C] methyl)-
`
`1,2-benzisoxazole—(AD-810) in rats, dogs and monkeys and of AD-
`810 in men. Arzneim-Forsch 1983;33:961-967.
`21. Izami T, Enomoto S, Hosiyama K, Sasahara K, Shibukawa A,
`Nakagawa T, et al: Pharmacokinetics of troglitazone, a new and ex-
`tensively metabolized antidiabetic agent, after oral administration,
`with an animal scale-up approach. I Pharmacol Exp Ther 1996;
`277:1630-1641.
`
`22. Howell SR, Hicks DR, Scatina IA, Sisenwine SF: Pharma-
`cokinetics of venlafaxine and o-desmethylvenlafaxine in laboratory
`animals. Xenobiotica 1994;24:315-327.
`
`23. Klamerus KI, Maloney K, Rudolph RL, Sisenwine SF, Iusko WI,
`Chiang ST: Introduction of a composite parameter to the pharma-
`cokinetics of venlafaxine and its active o-desmethyl metabolite. I
`Clin Pharmaco]1992;32:716-724.
`
`24. Bhargava HN, Villar VM, Gulati A, Chari G: Analgesic and
`hyperthermic effects of intravenously administered morphine in the
`rat are related to its serum levels. I Pharmacol Exp Therap 1991;
`258:511-516.
`
`25. Matsuzawa T, Wada Y, Shimoyama M, Nakajima K, Seki T,
`Sugibayashi K, et al: The effect of different routes of administration
`on the metabolism of morphine: the disposition of morphine and its
`metabolites after topical application. Biopharm Drug Disp 1994;
`15:665-678.
`
`26. Merrell WI, Gordon L, Wood AI, Shay S, Iackson EK, Wood M:
`The effect of halothane on morphine disposition: relative contribu-
`tions of the liver and kidney to morphine glucuronidation in the dog.
`Anesthesiology 1990;72:308-314.
`27. Miline RW, Nation RL, Somogyi AA: The disposition of morphine
`and its 3-and 6-glucuronide metabolites in humans and animals and
`the importance of the metabolites to the pharmacological effects of
`morphine. Drug Metab Rev 1996;28:345-472.
`28. Adusumalli VE, Yang IT, Wong KK, Kucharczyk N, Sofia RD:
`Felbamate pharmacokinetics in the rat, rabbit, and dog. Drug Metab
`Disp 1991;19:1116-1125.
`29. Physician’s Desk Reference. 56th ed. Montvale, NI: Medical Eco-
`nomics Company, 2002.
`30. Wu WN, Pritchard IF, Hills IF, Uetz IA, Yorgey KA, McKown LA,
`et al: Disposition of bepridil in laboratory animals and man.
`Xenobiotica 1992;22:153-169.
`
`31. Kaul S, Dandekar KA, Schilling BE, Barbhaya RH: Toxicokinetics
`of 2’, 3’-didehydro-3’-deoxythymidine, stavudine (D4T). DrugMetab
`Disp 1999;27:1-12.
`32. Rana KZ, Dudley MN: Clinical pharmacokinetics of stavudine.
`Clin Pharmacokin 1997;33:276-284.
`
`33. Tanaka Y, Shimojyo H, Hata T, Hashimoto M, Noguchi H:
`Toxicokinetics of zenarestat, an aldose reductase inhibitor in animals
`and man. Xenobiotica 1994;24:461-471.
`
`34. Sacher G: Relation of lifespan to brain weight and body weight in
`mammals, in: Wolstenholme GEW, O’Connor M (eds.), CIBA Foun-
`dation Colloquia on Aging. London: Churchill, 1959;115-133.
`35. Mahmood I: Interspecies scaling: predicting oral clearance in hu-
`mans. AmericanITherap 2002;9z35-42.
`
`DRUG DEVELOPMENT
`
`697
`
`