VOLUME 3 NUMBER
`
`JDTAEH 3
`
`[6] 411-478(1996]
`
`ISSN tOét—TSéX _
`
`
`
`‘An International Forum on Drug Delivery and Targeting
`
`OURNALofDRUG ‘
`# TARG TIME:
`~
`
`EDITORS IN CHIEF: ALEXANDER T FLORENCE & VINCENT H
`
`l. LEE'
`
`M
`F???
`
`Editorial
`
`CONTENTS
`
`An Integrated Pharmacokinetic and Pharmacodynamic
`Approach to Controlled Drug Delivery
`DOUWE D. BREIMER
`.
`.
`
`Editorial
`
`Modulation of
`STANLEY l. on
`
`
`
`'. :
`
`_
`Review
`Nasal Delivery. The Use of Animal Models i
`
`Performance in Man
`
`LISBETH ILLUM
`
`
`T'.‘
`Antitumor Drugs:
`Evaluation of the Efficiency
`Simulation Analysis Based on Pharmacokinetic/
`Pharmacodynamic Considerations
`D. NAKAI, E. FUSE, H. SUZUKI, M. INABA
`and Y. SUGIYAMA
`
`Nasal Delivery of Vaccines
`AJ. ALMEIDA and HG. ALPAR
`
`41]
`
`417;
`
`427
`
`
`
`we magma meme-v
`
`annééié7an§
`
`REQEWE D
`
`b890
`
`
`
`AUG 2? 1995
`
`
`
`11311201500410
`'OCl‘ COVE”
`Petitioners' EX. 1030
`iiNWEREslTY LIBRARY
`horwood academic publishers Pagel
`_
`
`._
`
`
`
`IPR2015-00410
`Petitioners' Ex. 1030
`Page 1
`
`

`

`i This material may be protected by Copyright law (Title 17 U.S. Code)
`
`[62936
`
`7 ll
`i
`
`Journal of Drug 'liiigeting, 19%, Vol. 3, pp. 411-415
`Reprints available directly from the publisher
`Photocopying permitted by license only
`
`© 1996 OPA {Overseas Publishers Association)
`Amsterdam B.V. Published in The Netherlands
`by Harwood Academic Publishers Gmbl—l
`Printed in Panama
`
`Editorial
`
`An Integrated Pharmacokinetic and Pharmacodynamic
`Approach to Controlled Drug Delivery
`
`DOUWE D. BREIMER
`
`Division of Pllfll'lil'llt'uitlgy, Li’ideii/Iiiiistmiam Ci'iiici'fiir Drug Research, Leiden University, P. O. Box 9503, 2300 RA Leiden, The Netherlands
`
`INTRODUCTION
`
`The field of new drug delivery system research and
`development has come to great blossom over the
`past 15—20 years. Numerous systems have been
`designed and some have actually reached the phase
`of practical application. Technologically they repre-
`sent major feats characterized by rate— andXor time-
`controlled drug release, is delivery of active ingre-
`dient at a predetermined rate and/or time. These
`include polymeric devices and osmotic systems for
`the oral delivery of drugs and patches for delivery
`across the skin. Their release rate is most often of a
`zero-order nature that should lead to less fluctuat-
`
`ing drug levels than with conventional pharmaceu—
`tical formulations. The potential therapeutic advan~
`tage of such more or less constant delivery rates
`have been claimed to be severalfold: in vivo pre—
`dictability of release rate on the basis of in vitro
`data, minimized peak plasma levels and thereby
`reduced risk of adverse reactions, predictable and
`extended duration of action, reduced inconvenience
`
`of frequent redosing and thereby hence improve
`patient compliance. However,
`in relatively few
`cases such potential advantages have in fact proved
`to be of great therapeutic significance. Only too
`often major emphasis is placed on the relatively flat
`plasma level profile that is achieved (pharmacoki—
`netics), rather than on the improved drug effect
`profile (pharmacodynamics in disease state).
`In figure 1 the inter—relationship between drug
`delivery, pharmacokinetics, pharmacodynamics
`and pathophysiology is shown schematically.
`in this scheme it is clearly indicated that drug
`concentration in plasma is no more than a ”surro-
`
`gate” for pharmacological and clinical effects, the
`relevance of which can only be judged if the rela-
`tionship between pharmacokinetics and pharmaco-
`dynamics (PK/ PD)
`is well established. In other
`words, Only on the basis of quantitative information
`of this relationship the desired optimal drug c0n-
`centration time profile can be defined. This is then
`to be translated into desired characteristics of the
`
`drug release profile from the delivery system (feed-
`back, see figure 1).
`In fact, what
`is needed is
`pertinent information on the kinetics of drug effects
`and its (potential) dependence on the rate and time
`of drug input. This is what controlled drug delivery
`should be aiming at: optimal drug treatment
`through rate and time programmed drug delivery.
`Therefore in the design and development of such
`systems at least two fundamental questions should
`be asked and answered prior to their further devel—
`opment (Breimer, 1993a):
`
`1.
`
`2.
`
`a clinical pharmacological one in terms of the
`optimal rate and timing at which the drug
`should be delivered;
`this requires profound
`knowledge of the cencentration-effect rela tion—
`ship of the drug in man and its dependence on
`disease and rate and time profile of drug input
`(e.g. continuous versus pulsatile as extreme
`input profiles);
`a pliorrrioceiiticai technological one in terms of the
`most suitable system that can provide the
`required rate and time specifications via the
`desired route of administration,- this requires
`knowledge on the capacity, flexibility, rate and
`time programming possibilities.
`
`-l-|l
`
`IPR2015—00410
`
`Petitioners' EX. 1030
`
`Page 2
`
`IPR2015-00410
`Petitioners' Ex. 1030
`Page 2
`
`

`

`412
`
`D. D. BREIMER
`
`FEEDBACK
`
`
`
`PHARMACO-
`KINETICS
`
`PHARMACO-
`DYNAMICS
`
`PATHO-
`FYSIOLOGY
`
`DRUG
`
`DELIVERY
`
`input
`
`(dosing
`rate,
`
`DRUG 1%
`
`
`
`
`EUM'N’
`ATION
`
`
`(clearance)
`
`DRUG
`CONCEN-
`TRATION
`[N
`PLASMA
`
`(surrogate)
`
`
`surrogate)
`4—.- DRUG
`
`
`EFFECTS
`
`{pharm300'
`logical;f
`
`DISEASE
`PARAMETERS
`
`(clinical effects;
`(surrogate)
`endpoints)
`
`Currently these questions are often studied in the
`reverse order, i.e. new drug delivery systems look»
`ing for a suitable drug candidate. Ideally, however,
`relevant clinical pharmacological studies are under—
`taken in man with the drug candidate looking for a
`suitable delivery system. This requires PK/PD
`modelling experiments with rate and time con—
`trolled drug input as important variables to be
`studied. Rate-controlled delivery systems of a ge—
`neric type may be useful
`tools in this respect
`(Breimer et al., 1984; Soons et al., 1989).
`
`PKII’D MODELLING
`
`The relevance of PK/ PD modelling for drug re—
`search and development in general is beginning to
`be well appreciated (Peck et al., 1992). Its primary
`objective is to identify some key properties of a
`drug in viva, which allow the characterization and
`prediction of the time course of drug action under
`physiological and pathophysiological conditions.
`The modelling of direct pharmacological effects
`usually consists of three components: 1. a pharma—
`cokinetic model, characterizing the time course of
`drug (and metabolite) concentrations in blood or
`plasma,- 2. a pharmacodynamic model, characteriz—
`ing the relationship between concentration and
`intensity of effect; 3. often, a link model that serves
`to account for
`the frequently observed delay or
`
`the pharmacological effect
`time effects of
`other
`relative to the plasma concentration (Holford and
`Sheiner, 1981). This approach of ”effect compart—
`ment modelling” has proved to be very successful
`for quite a number of drugs; for several others no
`delay in drug distribution from plasma to the site of
`action has been observed and therefore a link model
`
`is not needed. Alternative, more physiologically
`and mechanistically based models have been pro—
`posed which are in particular relevant to indirect
`pharmacological effects,
`i.e. when the delay be—
`tween plasma—concentration and effect
`is
`largely
`determined by slowly developing or declining {sec—
`ondary) effects, rather than by slow distribution to
`the site of action (Jusko et a1., 1994).
`from
`The most important
`lesson to be learnt
`PK/PD modelling exercises for the field of drug
`delivery is that the time course of drug effects can
`be and generally is quite different from the time
`course of drug concentrations.
`In other words,
`without PK / PD information the effect time course
`
`cannot be predicted on the basis of pharmacokinet-
`ics alone. For example, a short plasma elimination
`half-life will not necessarily imply short duration of
`action. This has clearly been recognized as early as
`1966 in pioneering studies by G. Levy, showing that
`the decrease of pharmacological effect intensity of
`several reversibly acting drugs is a function of the
`slope of the drug’s intensity of effect versus log
`concentration relationship (in the linear part) and
`IPR2015—00410
`
`
`
`Petitioners' EX. 1030
`
`Page 3
`4—4
`
`IPR2015-00410
`Petitioners' Ex. 1030
`Page 3
`
`

`

`AN INTEGRATED l‘l-lARMACOKlNE'I'IC AND PHARMACODYNAMIC
`
`413
`
`the drug’s elimination rate constant (Levy, 1966).
`Two drugs with similar elimination half—lives but
`very different slopes in concentration—effect rela-
`tionships will have very different durations of ef-
`fect. In terms of drug delivery this implies that for
`the compound with the steeper slope a controlled
`delivery system may well be indicated, whereas
`this is not
`the case for the compoimd with an
`intrinsically long duration of action. Examples can
`be found among the B—blocking agents, e.g. pin-
`dolol versus propranolol (Carruthers et al., 1985).
`Of course, there are also cases where duration of
`
`action is much more dependent on plasma kinetics
`if effect slopes between two compounds are rela-
`tively steep and similar, e.g. nifedipine versus am-
`lodipine. Here, controlled delivery is indicated for
`nifedipine (elimination half-life 2—4. hours) but not
`for amlodipine (elimination half-life 35—50 hours).
`It should furthermore be understood that if the
`
`relationship between concentration and effect is of a
`sigmoidal nature, i.e. a maximum is reached with
`higher concentrations, there will be no decline of
`drug effect intensity in spite of decreasing plasma-
`concentrations until the range is reached where the
`slope will determine the time course of declining
`effect. An example of this is omeprazole, which
`seems to be absorbed and eliminated rapidly, but
`exhibits profound proton pump blocking effect for
`at least 24 hours.
`
`Observations in studies where pharmacological
`effect intensity only gradually increases after rapid
`iv. injection have triggered the concept of ”effect
`compartment” modelling. This often reflects a slow
`access of the drug to the site of action (effect
`compartment), but can also be caused by processes
`secondary to drug—receptor interaction which oper-
`ate at a different
`than instantaneous time scale.
`
`Several examples of drugs exhibiting ”distribution
`delays“ are known,
`like some benzodiazepines,
`neuromuscular blocking agents, digoxin. Examples
`of drugs exhibiting ”effect delays” include corticos-
`teroids, oral anticoagulants, growth factors, cytok-
`ines and probably several other regulatory protein
`and peptide drugs. Again, for such compounds
`controlled drug delivery cannot be based on their
`(often rapidly fluctuating) plasma kinetics; PK/PD
`information is essential to achieve optimal results.
`It should also be noted that with such compounds
`no rapidly changing effects with time can be
`achieved, which might occasionally be desirable for
`chronotherapeutic reasons.
`A clear example of the application of PK/PD
`modelling to optimize controlled drug delivery is
`represented by somatosta tin in suppressing growth
`
`hormone levels in acrornegaly (Mazer, 1990). it was
`shown that continuous subcutaneous infusion is
`the optimal delivery regimen. Once the PK/PD
`relationship is established and validated under
`different conditions, also simulation experiments of
`various controlled delivery regimen may be quite
`helpful for this purpose.
`
`TIME DEPENDENCE
`
`PK/PD relationships are usually established in
`clinical pharmacological studies in healthy subjects
`Lmder relatively standardized conditions. For this
`information to be relevant for controlled drug de—
`livery, it is very important that its potential rate and
`time dependence be elucidated. It is for example
`not true that the maintenance of constant concen-
`
`trations (stead y-sta te) through zero—0rd er input rate
`is always associated with a constant pharmacologi—
`cal effect intensity. Theory, as outlined under PK/
`PD modelling, dictates that this should in principle
`be the case. However, tolerance development and
`circadian variation in the (pathOJphysiological sys-
`tems to be influenced, are two important factors
`that may cause major deviations. A well—knewn
`example of tolerance development is that of con-
`tinuous delivery of nitroglycerine by the transder-
`mal route. On the other hand it is now reasonably
`well established that this can be prevented if deliv—
`ery takes place at 12 hours’ on—off cycles; in other
`words,
`time programming in drug delivery is
`needed to avoid tolerance development. Relatively
`simple PK/PD studies (iv.
`infusion studies with
`different duration and different intervals) with ni-
`troglycerine at an early stage of patch development
`could have provided this very relevant piece of
`information. Then transdermal nitroglycerine treat-
`ment could have been optimal from the beginning
`onwards. With respect to circadian variability, also
`well—documented examples are available. For ex—
`ample, continuous i.v.
`infusion of famotidine (a
`HZ—blocking a gent} leads to constant plasma levels,
`but constant effect levels (high pH in the stomach)
`are only reached between 2 am. and 1 pm. After 1
`pm. a very considerable decrease in pH is seen.
`This is most
`likely caused by food intake and
`intrinsic diurnal variation in H+—secretion. Similar
`
`results have been obtained during continuous infu—
`sion with ranitidine. In terms of drug delivery this
`clearly implies the need for time control of different
`delivery rates at different times of day, if indeed a
`continuously high pH should be desirable.
`Also the severity of disease may be quite different
`at different times during a 24 hours’ cycle. Dethlef-
`sen and Repges 0985) studied the incidence of
`IPR2015—00410
`
`Petitioners' EX. 1030
`
`Page 4
`
`r -
`
`s___,__
`
`IPR2015-00410
`Petitioners' Ex. 1030
`Page 4
`
`

`

`414
`
`D. D. BREIMER
`
`severe asthmafic attacks in more than 1500 patients
`and found that these occur predominantly in the
`early morning hours and not during day-time. It is
`therefore questionable whether drug treatment of
`asthma with theophylline controlled release prepa—
`rations, which aim at a plasma level profile as
`constant (flat) as possible for 24 hours,
`is most
`optimal. Indeed, Staudinger (1990) has shown that
`a time-controlled theophylline formulation aiming
`at maximal drug concentrations during early morn—
`ing hours exhibited better improvement of lung
`function than a zero-order release product. PK/PD
`is clearly time-dependent under such circum‘
`stances, and this should be investigated in the
`context of appropriate experimental protocols lead-
`ing to relevant information with respect to time
`specifications for controlled delivery. The fields of
`chronopharmacology and chrono(patho)physio]-
`ogy are rapidly emerging and should be taken
`seriously by those engaged in drug delivery re-
`search (Hrushesky et al., 1991).
`
`RATE DEPENDENCE
`
`The other very important variable that may influ-
`ence both pharmacokinetics and pharmacodynam-
`ics is the rate of drug input (absorption rate for
`conventional
`formulations). Differences
`in Such
`rates will usually result in differences in peak times
`(tmx) and peak concentrations (CHM) and thereby
`in intensity of drug effects. Since high peak levels
`are sometimes associated with too intensive effects,
`
`a potential advantage of controlled delivery is that
`these can be avoided. This may well be explained in
`the context of the same PK/PD relationship for an
`entire concentration range, independent of the rate
`at which concentrations are reached. H0wever,
`
`there may be instances where the rate of change of
`plasma concentration as determined by the rate of
`input, may be of major influence on the PK/PD
`relationship. in other words: at one plasma concen—
`tration differences in effect intensity may be ob—
`served, dependent on the rate at which they were
`reached. The best documented example in this
`respect is nifedipine. This calcium channel blocker
`was originally marketed in a capsule formulation,
`from which it is rapidly released and absorbed and
`later on also in a sustained release tablet formula—
`
`tion. Extent of bioavailability from the two prepa—
`rations is comparable (average value of about 50%),
`but there is a profound difference in plasma level
`versus time profile (Kleinbloesem et al., 1984a). The
`capsule preparation leads to rapid and relatively
`
`high peak concentrations, whereas the tablet gives a
`relatively flat plasma level profile. In fact the latter
`is an example of a ”flip-flop” situation: the rising
`part of
`the curve represents drug elimination,
`whereas the decreasing part is a reflection of the
`delayed release and absorption of nifedipine.
`Interestingly, it was observed that the increase in
`heart rate (side—effect) was far less with the tablet
`than with the capsule in all subjects, whereas with
`both preparations a slight blood pressure lowering
`effect was achieved in the normotensive subjects.
`This observation was further substantiated in a
`
`subsequent study in which nifedipine was admin—
`istered rectally by the OSMET—osmotic pumps to
`healthy subjects for 24 h (Kleinbloesem et a1.,
`1984b). Concentrations rose relatively slowly and
`steady—state was reached after 6—1 0 h,- there was no
`increase in heart rate at all and a smooth decrease in
`
`blood pressure was noted. This led to the hypoth-
`esis that the rate-of—increase of plasma concentra-
`tion nifedipine (rather than absolute concentration)
`is a determinant factor for the drug’s haemody-
`namic effects. Clear evidence for this was obtained
`
`in a study in which nifedipine was given by two i.v.
`regimens, each to produce the same steady—state
`concentration, but attained gradually (infusion pro-
`tocol) or rapidly (injection-infusion protocol). No
`increase in heart rate occurred with the slow regi-
`men, whereas a substantial and long—lasting in-
`crease was seen with the rapid regimen (Kleinb-
`loesern et al, 1987). In the former case a gradual
`decrease in blood pressure was observed, whereas
`in the latter case hardly any blood pressure lower-
`ing effect occurred. Clearly, control of the rate of
`drug input
`is an essential
`feature in nifedipine
`therapy: side—effects can be avoided and desirable
`effect enhanced in this way.
`It seems not unreasonable to claim that the enor-
`
`mous success of Procardia XL, which is nifedipine
`in an oral osmotic once—daily formulation, is based
`on the principle of this clear dissociation of unde—
`sirable (increase in heart—rate) and desirable (blood
`pressure lowering) effects. Current sales figures of
`this product of over one billion dollar per year on
`the American market, make this the most successful
`controlled drug delivery product ever. It clearly
`illustrates what competitive advantage can be
`achieved if the concept of a controlled delivery
`profile has a solid PK/PD basis which is further
`substantiated by clinical studies. This issue also
`gives
`rise
`to interesting questions concerning
`bioequivaience assessment of similar nifedipine for—
`mulations. Can this be based on pharmacokinetic
`
`
`
`
`
`IPR2015—00410
`
`*
`
`Petitioners' EX. 1030
`
`Page 54.4
`
`IPR2015-00410
`Petitioners' Ex. 1030
`Page 5
`
`

`

`r_—i
`
`AN INTEGRATED PHARMACOKINETIC AND PHARMACODYNAMIC
`
`415
`
`should pharmacodynamic
`comparison only, or
`measurements be made mandatory also (Breimer,
`1993b}?
`
`CONCLUSIONS
`
`The assessment of optimal drug delivery regimen
`requires integrated pharmacokinetic/ pharmaco—
`dynamic/ clinical research strategies, in which both
`rate and time control are important issues. This
`holds true for existing drugs for which therapeutic
`application could be improved by taking controlled
`drug delivery into perspective, as well as for new
`drugs. Optimal drug delivery is an important as-
`pect of achieving a maximal efficacy/safety ratio for
`any drug and sthuld therefore be taken into ac—
`count at an early stage of (new) drug development.
`
`References
`
`Breimer, DD. (1993a) The need for rate and time programming
`in future drug delivery. Pharmacokinelic, pharmacodynamie
`and clinical considerations. ln Pillsalile Drug Delleery—Carrel”
`Applications and Fulm‘e 'll'emls, eds. R. Corny, H.E. lunginger,
`NA. Peppas, Wiss. Verlagsgesellscha ft, Stuttgart, pp. 25—10.
`Breimer, DD.
`(1993b) Relevance of pharmacodynamics in
`bioequivalence studies.
`In
`l3iiilliterriatin:ml—Biurivirilnliilily,
`Bieeiiuivalena' and Plim’umceltinelics, eds. K.K. Midha, HH.
`Bltnne, Med. Pharm. Sci. Publishers, Stuttgart, pp. 259 260.
`Breimer, DD.
`(1984) New drug delivery systems as tools in
`clinical pharmacology. In Proceedings 2nd World Conference on
`Clinical Plul'l'l'liflCtJlLng and Therapeutics, eds. L. Lemherger,
`M.M. Reidenberg, American Society of Pharmacology 3: Ex-
`perimental Therapeutics, Bethesda, pp. 431—143
`Carruthers, S.W., Pacha, W.L., Aellig, W.|l.
`(1985) Contrasts
`between pindolol and propranolol concentrahon—response
`relationship. Br. J. cliu. l’lrarimrcal. 20, 417—420
`
`Dethlefsen, U., Repges, R. (1985) Ein neues Therapieprinzip bei
`ntichtlichem Asthma. Med. Kliuik 80, 44—4?
`Holford, N.H.C., Sheiner, LB. (1981) Understanding the dose-
`effect relationship: Clinical applications of pharmacokinetic—
`pharmacodynamic models. Clln. Plrm'lmicnltin.
`IS, 429—453
`Hrushesky, W.J.M., Langer, R., Theeuwes, F. (eds) (199]) li‘mpnral
`coulrril efrlrug delivery, The New York Academy of Sciences,
`New York
`
`Jusko, W.]_, Ko, H.C. (1994) Physiologic indirect response models
`characterize diverse types of pharmacodynamic effects. Clin.
`Pluirumcril. Then, 56, 406—119
`Kleinbloesem, C.H., Van Brummelen, P., Van de Linde, J.A.,
`Voogd, P.]., Breimer, DD.
`(1984a) Nifedipine: kinetics and
`dynamics in healthy subjects. Cliu. lermacel. 'l'lier. 35, ?42—749
`Kleinbloesem, C.fl., Van Harten, ]., De Leede, L.C.]., Van Brom—
`melen, Breimer, D.D.
`(1984b) Kinetic and haemodynamic
`effects of nifedipine during rectal infusion to steady—state with
`an osmotic system. Cliu. Plim'macol. Tllt’l'. 36, 396—101
`Kleinbloesem, C.I-i., Van Brummelen, P., Danhof, M., Faber, H_,
`Urquhart, J., Breimer, DD. (198?) Rate of increase in plasma
`concentration of nifedipine as a major determinant of its
`haemodynamie effects in humans. Cliu. lermacal. Tla'r. 4‘1,
`26—30
`
`Levy, (5. (1966) Kinetics of pharmacological effects. Cliu. Pharma—
`cal. Ther. 5", 362—372
`Mazer, NA.
`(1990} Pharmacokinetic and pharmacodynamic
`aspects of polypeptide delivery. l. Control. Rel. 11, 343—354
`Peck, CC. Barr, W.H., Benet, L.Z., Collins. ]., Desjardins, R.E.,
`Furst, D. E., Harter, LG, Levy, (3., Ludden, T_, Rodman, ]_H_,
`Sanathanan, L., Schentag, ].J., Shah, V. P., Sheiner, L.B., Skelly,
`J.P., Stanski, D.R., Temple, R.J., Viswanathan, Weissinger, ].,
`Yacobi, A. (1992) Opportunities for integration of pharmaco—
`kinetics, pharmacodynamics and toxicokinetics in rational
`drug development. Cllll. lermacel. Titer. 5], 465—474
`Soons, PA, De Boer, A.G., Van Bruinmelen, P., Breimer, DD.
`(1989) Oral absorption profile of nitrendipine in healthy
`subjects: a kinetic and dynamic study. Br. I. cliu. lermacel. 2'3",
`‘lT‘J—lBQ
`
`Standinger, H. (‘l 990) Chronopharmaeology as a rationale for the
`development of controlled release products. in Oral Controlled
`Release Producls—Tliernpeutic and l3iaplmrmnccalic Assessment,
`eds. U. Cundert—Remy, H. Moller, Wiss. Verlagsgesellschaft,
`Stuttgart, pp. 83—9?
`
`IPR2015—00410
`
`Petitioners' EX. 1030
`
`Page 6
`
`IPR2015-00410
`Petitioners' Ex. 1030
`Page 6
`
`